Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Estimating the preventable portion of lifestyle-related reproductive casualties Ross, Susan E. 1984

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1985_A6_7 R67.pdf [ 16.8MB ]
Metadata
JSON: 831-1.0096250.json
JSON-LD: 831-1.0096250-ld.json
RDF/XML (Pretty): 831-1.0096250-rdf.xml
RDF/JSON: 831-1.0096250-rdf.json
Turtle: 831-1.0096250-turtle.txt
N-Triples: 831-1.0096250-rdf-ntriples.txt
Original Record: 831-1.0096250-source.json
Full Text
831-1.0096250-fulltext.txt
Citation
831-1.0096250.ris

Full Text

ESTIMATING THE PREVENTABLE PORTION OF LIFESTYLE-RELATED REPRODUCTIVE CASUALTIES By SUSAN E. ROSS B.H.Ec, The University of B r i t i s h Columbia, 1967 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE HEALTH SERVICES PLANNING AND ADMINISTRATION i n THE FACULTY OF GRADUATE STUDIES Department of Health Care and Epidemiology We accept t h i s thesis as conforming to, the required .standard THE UNIVERSITY OF BRITISH COLUMBIA December 1984 © Susan E. Ross, 1984 In presenting t h i s thesis i n p a r t i a l f u l f i l l m e n t of the requirements for an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of t h i s thesis for s c h o l a r l y purposes may be granted by the Head of my Department or by h i s or her representative. I t i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of /J^L^i^dC ^^tg-^£*»wC^^ 60-^^-The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada, V6T 1W5 Date ^ /9fS ABSTRACT The purpose of t h i s study was to review the evidence l i n k i n g maternal and paternal l i f e s t y l e habits i n the preconception and prenatal period to adverse reproductive outcomes; to determine either the proportion of reproductive c a s u a l t i e s which could be at t r i b u t e d to l i f e s t y l e r i s k , thus be amenable to prevention, or the information required to estimate the preventable portion of l i f e s t y l e - r e l a t e d reproductive c a s u a l t i e s ; and to examine a method f or s u r v e i l l a n c e of reproductive health i n the community which would provide the basis for a comprehensive information system suited to the needs of the research, planning, preventive medicine and health promotion communities. As a means of managing the s i z e of the study report, only a representative set of l i f e s t y l e s (smoking, alcohol consumption and n u t r i t i o n ) and research l i t e r a t u r e (major cohort and case-control studies i n human populations) was reported i n d e t a i l . A method was developed to review and describe the degree to which the evidence meets established c r i t e r i a for causal association. The most recently a v a i l a b l e prevalence data for determining smoking, alcohol and n u t r i t i o n a l r i s k , and incidence data f o r seven reproductive outcomes ( i n f e r t i l i t y , spontaneous abortion, s t i l l b i r t h , i n f ant m o r t a l i t y , congenital anomalies, f e t a l growth and morbidity) i n the B r i t i s h Columbia population were used to c a l c u l a t e the preventable portion of reproductive casualties i n th i s community. A review of the va r i a b l e s required, compared with the data a v a i l a b l e , provided the i i i b asis for recommendations regarding a reproductive health information system to support community s u r v e i l l a n c e , evaluation and research. The study supports the conclusion that there i s evidence of a causal l i n k between exposure to l i f e s t y l e r i s k s and the majority of adverse reproductive outcomes selected as i n d i c a t o r s of reproductive health. The c a l c u l a t i o n of the preventable portion ( e t i o l o g i c f r a c t i o n ) of l i f e s t y l e - r e l a t e d reproductive c a s u a l t i e s i n B r i t i s h Columbia suggests the preventable portion associated with s i n g l e l i f e s t y l e r i s k v a r i a b l e s may be i n the range of 10-50 percent. A more extensive and up-to-date set of population data for B r i t i s h Columbia i s required to determine an accurate estimate. The benefits to be derived from an improved information system were d e t a i l e d i n the study. Reproductive health data c o l l e c t e d f or B r i t i s h Columbia i s p r i m a r i l y outcome oriented with very l i t t l e input data on which to base r a t i o n a l planning decisions f o r the improvement of reproductive health outcomes. The study recommends that a more comprehensive reproductive health information system, with an integrated, linked data base, be considered a high p r i o r i t y by government and a l l i n s t i t u t i o n s , agencies and i n d i v i d u a l s working to improve reproductive health outcomes i n B r i t i s h Columbia. The p o t e n t i a l to improve reproductive health i s s i g n i f i c a n t enough to warrant action at the c l i n i c a l and community l e v e l , but a d d i t i o n a l data are required to plan cost-e f f e c t i v e intervention s t r a t e g i e s , to monitor improvements i n reproductive health, and to support applied research i n i t i a t i v e s . i v TABLE OF CONTENTS Page CHAPTER I: INTRODUCTION TO THE STUDY 1 Introduction 1 Objectives 6 Rationale and Background 7 Limitations 11 Thesis Format 12 CHAPTER I I : MEASURING REPRODUCTIVE HEALTH IN A COMMUNITY . . . . 13 F e r t i l i t y and Impaired Fecundity 15 Spontaneous Abortion 25 S t i l l b i r t h 35 Infant M o r t a l i t y 40 F e t a l Growth and Growth Retardation 53 B i r t h Defects and Malformation 68 Infant Morbidity 78 CHAPTER I I I : THE EFFECT OF LIFESTYLE ON REPRODUCTIVE HEALTH . . . . . . . . . 87 Smoking - Tobacco 88 - Cannabis 122 Diet and N u t r i t i o n 133 Alcohol 182 V Page CHAPTER IV: THE PREVENTABLE PORTION OF REPRODUCTIVE CASUALTIES 204 The E f f e c t of L i f e s t y l e : Review of Evidence 205 The B r i t i s h Columbia Population at Risk 228 The Preventable Portion of Reproductive Casualties 231 B r i t i s h Columbia's Preventable Portion of Reproductive Casualties: The Problem of Missing Data 237 CHAPTER V: A REPRODUCTIVE HEALTH DATA BASE 245 The Benefits of a Reproductive Health Data Base 246 Reproductive Health Index Variables 249 The F e a s i b i l i t y of a B r i t i s h Columbia Index 260 Study Conclusions 266 BIBLIOGRAPHY 275 Chapter I 275 Chapter II 276 Chapter I II 283 Smoking 283 Cannabis 294 Diet and N u t r i t i o n 303 Alcohol 315 Chapter IV 320 Chapter V 323 APPENDIX 327 v i LIST OF TABLES Page TABLE 1: Number of Currently Married White Women 15-44 Years of Age and Percent D i s t r i b u t i o n by Number of Reported Spontaneous Pregnancy Losses, According to Age, P a r i t y , and Fecundity Status: United States, 1976 30 TABLE 2: T o t a l Number of S t i l l b i r t h s and Rate per 1,000 Li v e and S t i l l b i r t h s for B r i t i s h Columbia, 1978-1983 37 TABLE 3: Tot a l Number of S t i l l b i r t h s and Percent D i s t r i b u t i o n by Birthweight Category for B r i t i s h Columbia, 1979-1983 38 TABLE 4: Tot a l Number of S t i l l b i r t h s and Percent D i s t r i b u t i o n by Gestational Age for B r i t i s h Columbia, 1979-1983 39 TABLE 5: Tot a l L i v e b i r t h s , T o t a l Population and B i r t h r a t e per 1,000 Population; and Tot a l Infant Deaths by Six Age-Specific Categories ( F e t a l , Early Neonatal, Late Neonatal, P e r i n a t a l , Neonatal, and Infant) and Rate per 1,000 L i v e b i r t h s Cor per 1,000 Live and S t i l l b i r t h s ) i n B r i t i s h Columbia, 1978-1983 . 49 TABLE 6: T o t a l Number of L i v e b i r t h s and Infant Deaths: Ea r l y Neonatal, Late Neonatal, Post Neonatal, and Infant Death, and Rate per 1,000 L i v e b i r t h s by Birthweight Category i n B r i t i s h Columbia, 1979-1983 50 TABLE 7: T o t a l Number of L i v e b i r t h s and Tot a l Number of Congenital Anomalies (ICD 740-759) and Rate per 1,000 L i v e b i r t h s by Major Diagnostic Categories and for Male and Female Births i n B r i t i s h Columbia, 1971-1980, as Reported by Year End 1981 76 TABLE 8: Major Diagnostic Categories of Congenital Anomalies (ICD 740-759) by Incidence Rate per 10,000 L i v e b i r t h s and Proportion of a l l Congenital Anomalies Reported i n B r i t i s h Columbia L i v e b i r t h s for the period 1971-1980 77 v i i Page TABLE 9: Proportion of Female Population 15 and Over Reporting Daily Cigarette Smoking, by Age i n the Smoking Habits of Canadians Survey, December 1977 and December 1979, and Canada Health Survey, 1978-1979 120 TABLE 10: To t a l Female Population 15 Years and Over and Percent D i s t r i b u t i o n by Type of Cigarette Smoker and by Number of Cigarettes Smoked Daily, by Age Group, fo r Canada and for B r i t i s h Columbia, Canada Health Survey, 1978-1979 121 TABLE 11: Proportion of the Pregnant and Female Population 10-19 Years of Age i n B r i t i s h Columbia with an Inadequate Daily Intake of Nutrients and with Moderate to High Risk Biochemical Measures, Assessed by the N u t r i t i o n Canada Survey, 1973 . . 181 TABLE 12: Population 15 Years and Over by Type of Drinker and Weekly Volume of Alcohol Consumed, by Age and Sex, Canada and by To t a l Population, BC, 1978-79 203 TABLE 13: Smoking and Reproductive Health—Summary of Evidence f or Causal Ass o c i a t i o n Between Smoking Exposure and Seven Reproductive Health Outcomes, Based on Seven C r i t e r i a 221 TABLE 14: N u t r i t i o n and Reproductive Health—Summary of Evidence f or Causal Association Between Exposure to Poor N u t r i t i o n and Seven Reproductive Health Outcomes, Based on Seven C r i t e r i a 222 TABLE 15: Alcohol and Reproductive Health—Summary of Evidence for Causal Association Between Exposure to Alcohol and Seven Reproductive Health Outcomes, Based on Seven C r i t e r i a 223 v i i i Page TABLE 16: An Estimate of the Preventable Portion ( E t i o l o g i c Fraction) of Adverse Reproductive Health Outcomes i n B r i t i s h Columbia, 1983. A. C a l c u l a t i o n of the E t i o l o g i c a l F r a c t i o n f o r Single L i f e s t y l e Factors . . 235 TABLE 17: An Estimate of the Preventable Portion ( E t i o l o g i c Fraction) of Adverse Reproductive Health Outcomes i n B r i t i s h Columbia, 1983. B. Estimates of the Minimum Number of Preventable Reproductive Casualties Given the E t i o l o g i c a l F r a c t i o n Reported i n Table 16 236 TABLE 18: Alternate Methods to Estimate the Preventable Portion of Low Birthweight from Removing Exposure to Two L i f e s t y l e Risk Factors 238 TABLE 19: An Estimate of the Preventable Portion of Low Birthweight i n the Kansas Study Population 241 TABLE 20: An Estimate of the P o t e n t i a l Impact of a Successful L i f e s t y l e Intervention Program on Low Birthweight i n the Population when Risk Groups are S e l e c t i v e l y Targeted, Based on Data from the Kansas Study 242 TABLE 21: Current A v a i l a b i l i t y of Data f o r the Recommended Reproductive Health Index, BC Population, 1984 . . . 261 i x LIST OF FIGURES Page FIGURE 1: A Model of the Threshold of I n f e r t i l i t y Concept . . . 10 FIGURE 2: Age-Specific C l a s s i f i c a t i o n s f o r F e t a l and Infant M o r t a l i t y 42 X ACKNOWLEDGEMENTS I would l i k e to thank the members of my thesis committee f o r t h e i r continuous support and patience, and for t h e i r " j u s t r i g h t " blend of challenge and encouragement: Dr. Annette Stark, chairperson, for her s p e c i a l a b i l i t y to l i s t e n and ask the pertinent questions, f o r her expert guidance and for her frien d s h i p ; Dr. Sam Sheps f o r so w i l l i n g l y sharing h i s knowledge and expertise, and f o r his in s i s t e n c e that the data be r e a l , the d e t a i l s c l e a r and the r e s u l t s u s e f u l ; Dr. Gerald Bonham for h i s conceptual s k i l l s , for the i n s p i r a t i o n provided by his sincere and active i n t e r e s t i n reproductive health issues, and for teaching, by example, that optimal standards are r e a l i s t i c goals. I am appreciative of the support of Mr. A. H. Herson, Director of the D i v i s i o n of V i t a l S t a t i s t i c s , B r i t i s h Columbia, who r e a d i l y agreed to provide data needed for t h i s study and would p a r t i c u l a r l y l i k e to thank Mr. Jack Rowe, Manager of S t a t i s t i c a l and Health Records for h i s ongoing cooperation i n preparing the data and h i s guidance i n i t s i n t e r p r e t a t i o n . The s k i l l s and experience of Mary Vorvis who typed t h i s manuscript were much appreciated. S p e c i a l thanks are due my husband Peter, my family and my partners for the love, encouragement and understanding that sustained me throughout t h i s period of study. 1 CHAPTER I: INTRODUCTION TO THE STUDY Introduction Over the past ten years, considerable attention has been given to the r o l e that a p o s i t i v e change i n l i f e s t y l e habits could play i n improving the health status of the North American population. In Canada, a working document published by the nationa l government i n 1974, e n t i t l e d "A New Perspective on the Health of Canadians" (Canada, Department of National Health & Welfare, 1974), introduced "the health f i e l d concept" which proposed a balance of attention be paid to each of four elements—human biology, environment, l i f e s t y l e , and health care o r g a n i z a t i o n — a f f e c t i n g the population's health. The concept was designed to be comprehensive, i n that any health problem could be traced to one or a combination of the four elements, and to permit a system of analysis by which the re l a t i o n s h i p and r e l a t i v e s i g n i f i c a n c e of each of the four elements to a health problem could be assessed. This a n a l y s i s , i t was stated, would permit program planners to focus t h e i r attention on the most important contributing f a c t o r s . Such analyses also, the report recognizes, requires information not yet at hand, places a heavy burden on research i n each of the four areas, and c a l l s for unity of purpose i n a l l f i e l d s of research towards improving health status. In 1979, the US Government published the Surgeon General's report on health promotion and disease prevention e n t i t l e d "Healthy People" (US, Department of Health, Education, & Welfare, 1979), which also indicated the government's readiness to make unequivocal statements 2 regarding the l i n k between c e r t a i n l i f e s t y l e or environmental factors and poor health. Analysis of the r e l a t i v e contributions of the four health f i e l d concept elements to the ten leading causes of death i n the United States i n 1976 suggested that as much as h a l f the mortality was due to unhealthy behavior and l i f e s t y l e ; 20 percent to environmental f a c t o r s ; 20 percent to human biology; and 10 percent to inadequacies i n health care. The s i g n i f i c a n c e " o f the document i s i t s r e f l e c t i o n of the growing consensus i n the United States about the need f o r , and value of, disease prevention and health promotion. The Department of Health and Human Services i s now required to submit a nationa l disease prevention p r o f i l e to the US Congress every three years. A chapter of the Canadian working paper e n t i t l e d "Science versus Health Promotion" declares that the government i s prepared to take health promotion action i n some areas of the environment and l i f e s t y l e i n advance of s c i e n t i f i c p r o o f — g i v e n that a hypothesis appears " s u f f i c i e n t l y v a l i d to warrant taking p o s i t i v e a c t i o n " (p. 58). This i s based on the r a t i o n a l e that "messages designed to influence the publ i c must be loud, clear and unequivocal" (p. 57) while science i s f u l l of " i f s " and "maybe1s." I t i s expected, however, that the gathering of r e l i a b l e evidence w i l l remain a focus of the s c i e n t i f i c community and that " i n due course the v a l i d i t y of the . . . hypothesis w i l l l i k e l y be resolved i n a s c i e n t i f i c way, precise cause-and-effect r e l a t i o n s h i p s • w i l l be ascertained and measured, and the exact s i g n i f i c a n c e of each factor determined" Cp- 58). 3 The f i r s t n a t i o n a l disease prevention p r o f i l e submitted to the US Congress also made reference to the l i m i t s of knowledge i n the f i e l d of prevention. The approach taken was to e s t a b l i s h a continuum for p r e v e n t a b i l i t y and determine where diseases or conditions f i t within the continuum extremes of absolutely preventable and no known prevention. This type of assessment was to allow for more r e a l i s t i c expectations of the varying degrees of impact that might be gained from preventive measures given that r i s k factors associated with the target outcome, as w e l l as intervention s t r a t e g i e s , are understood i n varying degrees. "In short," the report s t a t e s , "the p o t e n t i a l scope of disease prevention and health promotion i s vast, the types of possible interventions e x t r a o r d i n a r i l y varied, and the knowledge base uneven. . . . Fortunately, the data . . . i n d i c a t e that many of the diseases and conditions that constitute the Nation's most pressing health problems are ones where knowledge of r i s k s and e f f e c t i v e r i s k reduction approaches are present and growing" (US, Department of Health & Human Services, 1980, p. 270). The d e l i b e r a t e decision to support or encourage health promotion and prevention s t r a t e g i e s despite l i m i t e d knowledge places the health planner, as w e l l as the c l i n i c i a n , i n an intercept p o s i t i o n between the s c i e n t i f i c community which i s required to be s k e p t i c a l and the health education community which i s required to provide c l e a r , p r a c t i c a l health guidelines for the p u b l i c . Once consensus to promote a p a r t i c u l a r health behavior or to regulate a p a r t i c u l a r element i n the environment has been reached, there i s i m p l i c i t need for a shared plan which 4 determines the complimentary r o l e of each area of the health-oriented community i n confirming or r e j e c t i n g the v a l i d i t y of the approach, and the best approach to reaching s p e c i f i c goals for health, status improvement. One of the f i v e n a t i o n a l goals established by the US Government for health promotion and disease prevention, was to continue to improve in f a n t health, and, by 1990, to reduce i n f a n t m o r t a l i t y by at l e a s t 35 percent, to fewer than nine deaths per 1,000 l i v e b i r t h s . In recognition of the two factors most associated with i n f a n t mortality, subgoals of reducing the number of low b i r t h weight infants and reducing the number of b i r t h defects have also been established along with measurable objectives aimed at s p e c i f i c outcome areas where the rates of infant death are i n excess of the current n a t i o n a l average or the prevention of b i r t h defects and severe mental retardation does not match expectation. The type of preventive services to be made a v a i l a b l e to pregnant women as a means of achieving the goal and objectives have been i d e n t i f i e d as a balance of prenatal medical, r i s k assessment and r e f e r r a l s e r vices; l i f e s t y l e counselling p r i o r to and during pregnancy; and adequate services for labor, d e l i v e r y and postpartum period. Unlike the United States, Canadian governments have not made a pu b l i c commitment to a s p e c i f i c reduction i n infant m o r t ality, low birthweight, or b i r t h defects. Despite t h i s , there are i m p l i c i t l y recognized goals of p e r i n a t a l health care across the country with c a r e f u l monitoring of these outcomes v i a analysis of v i t a l s t a t i s t i c s 5 records at the p r o v i n c i a l and national l e v e l , and by compiling area s t a t i s t i c s within p u b l i c health administrations and h o s p i t a l s . Health promotion has been a component of public health prenatal education i n a l l areas of Canada since the 1970s. The most recent focus has been i n the area of n u t r i t i o n , with the d i s t r i b u t i o n of the "Five-Year F e d e r a l - P r o v i n c i a l Plan on N u t r i t i o n i n Health Promotion for Pregnant Women" (Ottawa, Canada, 1984) to be implemented over the period 1984-1989. At the same time, the B r i t i s h Columbia [BC] government has published the f i r s t p r o v i n c i a l health promotion guidelines to focus on the need to improve n u t r i t i o n and other l i f e s t y l e habits p r i o r to conception as w e l l as during pregnancy (BC, M i n i s t r y of Health, 1984). This thesis i s concerned with the r o l e of the health planner i n f a c i l i t a t i n g improvement i n reproductive and infant health status i n the community given the r e a l i t y of a l i m i t e d knowledge base on which to act and the consensus (as implied by government reports and actions) that health promotion and prevention s t r a t e g i e s should be undertaken. The primary focus w i l l be on the l i f e s t y l e element. The health promotion message for pregnant women i s multifaceted, but c l e a r : i t i s important to eat a we l l balanced d i e t ; to be p h y s i c a l l y f i t ; not to smoke; not to take drugs unless prescribed and necessary; to drink very l i t t l e a l c o h o l , i f any at a l l ; to avoid environmental hazards; to avoid undue str e s s ; to seek medical care early; to p a r t i c i p a t e i n prenatal education; to gain weight during pregnancy; to take a l l precautions to prevent having t h e i r baby born too soon or too small; and to breastfeed. The general l e v e l of awareness i n the community suggests the message has been received and acted upon by the majority of women and t h e i r partners i n t h e i r childbearing years (BC, Minis t r y of Health, 1984; Ottawa, Canada, 1984; US, Department of Health, Education, & Welfare, 1979; US, Department of Health & Human Services, 1980). Given public acceptance, the l i k e l i h o o d that the message i s also believed to be true by those who promote i t , and the p r o b a b i l i t y that the underlying hypotheses w i l l be used by many who plan and implement preventive and intervention s t r a t e g i e s , what degree of impact can be expected when an appropriately planned and implemented l i f e s t y l e modification program i s undertaken i n an e f f o r t to prevent low birthweight, infant m o r t a l i t y , or b i r t h defects? Moreover, given that there i s considerable disagreement i n the s c i e n t i f i c community about the degree to which any of the l i f e s t y l e factors contribute to adverse infant health outcomes, based on an assessment of the evidence" at hand, what types of information are required to help c l a r i f y these issues? What outcomes are aff e c t e d , i n what way, for whom, with what consequences What plan of action directed toward both the goal of improving infant health and the goal of improving our knowledge base would e l i c i t the support of both the s c i e n t i f i c and the health promotion communities? Obj ectives The objectives of th i s thesis are: 1. To review the evidence l i n k i n g maternal or paternal l i f e s t y l e habits around the time of conception, and maternal l i f e s t y l e habits 7 during pregnancy to adverse reproductive and pregnancy outcomes. 2. To determine what proportion of reproductive c a s u a l t i e s can be a t t r i b u t e d to l i f e s t y l e and, therefore, are amenable to prevention, or what information i s required to determine i f there i s a preventable portion a t t r i b u t a b l e to l i f e s t y l e f a c t o r s . 3. To examine a method for s u r v e i l l a n c e of l i f e s t y l e - r e l a t e d reproductive health outcomes i n B r i t i s h Columbia. Rationale and Background The desire to understand the reproductive process, i t s pathologies, the e t i o l o g i c factors r e l a t e d to these pathologies and the manner i n which they can be treated or prevented i s shared by many. This i s demonstrated by the extraordinary number of publications on the subject and the d i v e r s i t y of approach that stems from m u l t i - d i s c i p l i n a r y involvement. There i s a r e a l need to draw t h i s information together so that a l l the options f o r improving reproductive outcome can be considered. This i s e s p e c i a l l y true f o r the l i t e r a t u r e r e l a t i n g l i f e s t y l e f a c t o r s to reproductive health and for l i t e r a t u r e examining the underlying mechanisms associated with the reproductive c a s u a l t i e s . Within t h i s volume of l i t e r a t u r e , there are purported l i n k s between each of the l i f e s t y l e factors ( n u t r i t i o n , smoking, alcohol, drugs, s t r e s s , etc.) and many of the general outcomes associated with reproductive health ( i n f e r t i l i t y , i n t r a u t e r i n e death, b i r t h defects, low birthweight, prematurity, infant death, e t c . ) . This would suggest that a proportion of each type of reproductive casualty i s , at l e a s t i n part, preventable 8 through l i f e s t y l e modification. Is t h i s a v a l i d suggestion? The f a c t that each study can only focus on a portion of the question produces a s i t u a t i o n where there are many "pieces of the puzzle" but no completed pi c t u r e . A working hypothesis of the o v e r a l l manner i n which l i f e s t y l e factors may adversely a f f e c t reproductive outcome has been proposed by Wynn and Wynn (1981) . This represents a comprehensive attempt to r a t i o n a l i z e the reproductive health and l i f e s t y l e l i t e r a t u r e and for th i s reason, i s used i n t h i s thesis as a guideline for determining the s e l e c t i o n of the reproductive outcome measures and the boundaries for in v e s t i g a t i n g evidence of the impact of l i f e s t y l e factors on these measures. The hypothesis proposed by Wynn and Wynn (1979, 1981) can be summarized as follows: The health of parents around the time of conception i s c r i t i c a l to the health of t h e i r i n f a n t . Poor health p r i o r to conception and during the f i r s t weeks of pregnancy i s associated with a continuum of reproductive c a s u a l t i e s which include impaired fecundity, spontaneous abortion, s t i l l b i r t h , preterm b i r t h , f e t a l growth, retardation, malformations, and infant death. A reduced rate of c e l l r e p l i c a t i o n and/or f a u l t y c e l l r e p l i c a t i o n are common a t t r i b u t e s of many of these reproductive c a s u a l t i e s . The cause of the r e s u l t i n g reproductive c a s u a l t i e s i s m u l t i f a c t o r i a l , but some of the factors are shared by a l l and some of these shared factors are exogenous i n nature and amenable to modification p r i o r to conception. This suggests a proportion of each type of reproductive casualty i s preventable. 9 Wynn and Wynn conceptualize t h i s hypothesis by means of a "threshold of i n f e r t i l i t y " (see Figure 1) which i s described as a l e v e l of marginal reproductive health, determined i n part by l i f e s t y l e f a c t o r s , at which conception i s possible, but the r i s k of poor reproductive outcome i s high. Based on t h i s i n t e r p r e t a t i o n , the aim of any int e r v e n t i o n would be to ensure a l e v e l of reproductive health for men and women that i s above t h i s threshold and the predeterminant of successful intervention would be the a b i l i t y to i d e n t i f y the population within the high r i s k threshold and to as c e r t a i n the s i g n i f i c a n t differences between those below and above the threshold l e v e l . Given the above parameters, a number of steps must be taken to f u l f i l l the objectives of the th e s i s . F i r s t , a review of the broad base of relevant l i t e r a t u r e must be completed: a review of the current l i t e r a t u r e by l i f e s t y l e f a ctor and reproductive outcome topic area; a c r i t i c a l review of studies which demonstrate a q u a n t i f i a b l e l i n k between a r l i f e s t y l e f a c t o r and adverse outcome; a review of the outcome measures which would provide the best means of measuring impact, the public health s i g n i f i c a n c e of the e f f e c t of l i f e s t y l e factors on reproductive health; and a review of the methods for quantifying t h i s e f f e c t . Second, a comprehensive scheme to describe, assess, and/or monitor the impact of the l i f e s t y l e element on reproductive health i n a community must be determined. Third, an attempt must be made to determine how f e a s i b l e such an approach would be, given the resources of a representative community, and to judge the d i r e c t and i n d i r e c t benefits of the proposal Low* REPRODUCTIVE HEALTH STATUS •High (* The threshold of i n f e r t i l i t y ) Based on Wynn & Wynn, 1981 pl4. FIGURE 1 A Model of the Threshold of I n f e r t i l i t y Concept O 11 to the p u b l i c , and the s c i e n t i f i c , planning and health promotion communities. Limitations Many factors are involved i n r e a l i z i n g the successful b i r t h of a healthy i n f a n t to a healthy mother and father. The q u a l i t y of maternal health and health care p r i o r to and throughout the prenatal period; the presence or absence of maternal disease or trauma; the q u a l i t y and use of health f a c i l i t i e s and manpower resources f o r labour, b i r t h and the post partum period; the q u a l i t y of the environment; and family h i s t o r y and genetic propensity are a l l known to play a r o l e in.the f i n a l determination of maternal and infant health. These f a c t o r s , with the i n c l u s i o n of l i f e s t y l e , cover a l l four of the elements of the health f i e l d model. The a b i l i t y to discriminate between e f f e c t s of any of these four elements, or the e f f e c t s of s p e c i f i c components with i n an element i s c l e a r l y dependent on the information a v a i l a b l e and the i n t e r p r e t i v e ' a b i l i t y of the interviewer. Where possib l e , reviews of the l i t e r a t u r e by s p e c i a l i s t s i n a p a r t i c u l a r f i e l d or topic area w i l l be accepted as a r e l i a b l e c r i t i q u e of the relevant research i n that f i e l d . Studies selected as evidence from t h i s l i t e r a t u r e w i l l be reviewed according to established research c r i t e r i a (Andersen,et a l . , 1979; Gehlbach, 1982; H i l l , 1971; McMasters Series, 1981; Schlesselman, 1982; Susser, 1973). Some animal and tiss u e or c e l l culture studies w i l l be referenced, since they provide much of the basis for speculation i n human studies, but research selected f o r discussion of issues w i l l p e r t a i n to human population only. 12 Thesis Format The chapters which follow have been organized i n the following manner. Chapter two provides a d e f i n i t i o n of the reproductive outcome measures included i n the study, a review of the extent to which adverse outcomes are perceived to occur generally i n the population, the most recent B r i t i s h Columbia outcome data where a v a i l a b l e , and the problems associated with determining accurate incidence or prevalence rates for these outcomes. Chapter three reviews the evidence that l i f e s t y l e f actors contribute to the occurrence of reproductive c a s u a l t i e s and reports on the prevalence of adverse l i f e s t y l e habits i n the community where a v a i l a b l e . Chapter four examines an appropriate method for measuring the impact of the l i f e s t y l e element and provides estimates of the preventable portion of l i f e s t y l e - r e l a t e d reproductive c a s u a l t i e s from the study and B r i t i s h Columbia data. Chapter f i v e examines the r a t i o n a l e and design of the Reproductive Health Index, a measure to describe and monitor the reproductive health status of a community, and the change i n reproductive outcomes that might be a t t r i b u t e d to the l i f e s t y l e element. D i s p a r i t i e s between the type of data r o u t i n e l y c o l l e c t e d i n the community and the type of data seen to be important for evaluation and s u r v e i l l a n c e w i l l be i d e n t i f i e d and discussed i n terms of t h e i r f e a s i b i l i t y and p o t e n t i a l b e n e f i t s . Chapter f i v e concludes with an o v e r a l l evaluation of the extent to which study objectives are met, and the i m p l i c a t i o n of the study r e s u l t s f or the s c i e n t i f i c , health promotion and prevention, and planning communities i n improving reproductive health outcomes. 13 CHAPTER I I : MEASURING REPRODUCTIVE HEALTH IN A COMMUNITY The extent to which a population encounters problems i n producing healthy, l i v i n g infants i s influenced to varying degrees by each of the four elements which comprise the health f i e l d concept—human biology, environment, health care organization, and l i f e s t y l e . The extent to which the impact of any one of these elements on reproductive outcome can be assessed, and p o t e n t i a l l y influenced, i s dependent on (a) the a b i l i t y to adequately describe and accurately document the type and number of reproductive impairments or casualties occurring i n the population of childbearing years, and (b) on the a b i l i t y to adequately describe the population i n which reproductive c a s u a l t i e s do and do not occur. The purpose of t h i s chapter i s to review the manner i n which each type of reproductive problem i s described and q u a n t i f i e d , and the problems associated with determining accurate incidence or prevalence rates. The reproductive problems to be considered are: lack of f e r t i l i t y and impaired fecundity; spontaneous abortion; s t i l l b i r t h ; neonatal mortality; b i r t h defects and malformations; f e t a l growth ret a r d a t i o n and i n f a n t morbidity associated with impaired growth and development. The ultimate aim i n t h i s approach i s to develop a reproductive health index which w i l l adequately describe a l l the reproductive problems occurring i n a community, and help to i d e n t i f y r e l a t i o n s h i p s between the types of problems. This aim i s based on the premise that there 14 are common causal or mediating factors that a f f e c t reproductive capacity per se, and where these factors are amenable to modification and are al t e r e d , benefits to reproductive health should be evident across the continuum of reproductive outcome measures. A comprehensive index of reproductive health could provide a means of examining the community-specific r e l a t i o n s h i p between the occurrence of d i f f e r e n t types of reproductive problems as w e l l as a means of assessing the impact of delib e r a t e change i n environmental, health care, l i f e s t y l e , or b i o l o g i c a l factors important to reproductive outcome. However, the information needed to carry out t h i s type of analysis i s not currently a v a i l a b l e , f o r understandable r e a s o n s — g i v e n that our health care system i s extremely compartmentalized. S p e c i a l i s t s who are involved i n treatment of i n f e r t i l i t y , for example, end t h e i r involvement when pregnancy occurs; o b s t e t r i c i a n s , when a b i r t h occurs; g e n e t i c i s t s may begin t h e i r involvement around the period of conception and early pregnancy or a f t e r b i r t h ; p a e d i a t r i c i a n s do not become involved u n t i l a b i r t h occurs. Medical and h o s p i t a l care i s p r i m a r i l y oriented to cases and i n d i v i d u a l c l i e n t e l e ; public health, to populations; occupational health to worksites and the environment, etc. Reproductive health care i s the end r e s u l t of many d i s t i n c t l y independent a c t i v i t i e s and areas of focus, and the corresponding reporting and data c o l l e c t i o n system r e f l e c t s t h i s independent approach. I f , for example, an excess of pregnancies r e s u l t i n g from successful treatment of i n f e r t i l i t y subsequently end i n spontaneous abortion, 15 s t i l l b i r t h and/or f e t a l growth retardation, t h i s a s s o c i a t i o n (unless extreme) would go unnoticed at the community l e v e l f o r lack of access to linked records. And, i f these pregnancies were associated with a high proportion of women who smoked or were inadequately nourished, t h i s would also go unnoticed since (a) record linkage i s not established and, more importantly, (b) l i f e s t y l e habits are not c o n s i s t e n t l y assessed or recorded. I t i s apparent that the development of a comprehensive index of reproductive health would require some changes be made to the present reporting system. In order to assess the f e a s i b i l i t y of e s t a b l i s h i n g an index, incidence or prevalence data a v a i l a b l e from the BC community (or a representative community) w i l l be reported along with a discussion of the type of data needed to assess the impact of the various factors important to reproductive outcome. Since the focus of the study i s on the p o t e n t i a l impact of l i f e s t y l e f a c t o r s , discussion w i l l be l i m i t e d to l i f e s t y l e — b u t the process of review for t h i s chapter i s equally applicable to environmental, health care, and b i o l o g i c a l f a c t o r s . F e r t i l i t y and Impaired Fecundity The manner i n which va r i a b l e s concerned with the f e r t i l i t y or fecundity status of the population are c l a s s i f i e d and defined i s generally determined by the way the information i s to be used ( i . e . , f o r family planning and demography studies, for c l i n i c a l research on i n f e r t i l i t y treatment, e t c . ) . Apart from reporting on annual general f e r t i l i t y rate, one which r e l a t e s the t o t a l number of l i v e b i r t h s f or 16 the year to an approximation of the population exposed to the r i s k of pregnancy (midyear population of women 15-44 y e a r s ) — o r s p e c i f i c f e r t i l i t y rates f o r selected subgroups of the population (e.g., age-specific rates) — w h i c h can be r e a d i l y calculated from v i t a l s t a t i s t i c s data, information about f e r t i l i t y and impaired fecundity i s not r e g u l a r l y c o l l e c t e d nor reported f o r any community. A v a i l a b l e information i s most l i k e l y to be generated as a function of population growth studies, demographic surveys associated with family planning issues, and c l i n i c a l i n v e s t i g a t i o n s of the diagnosis and treatment of i n f e r t i l i t y . 'Questions of primary i n t e r e s t to t h i s study are: what proportion of the population who try to conceive do not become pregnant? How i s th i s proportion calculated? What types of problems are associated with i n f e r t i l i t y , how i s t h i s a s s e s s e d a n d how many are affected? Are reproductive outcomes of i n d i v i d u a l s with a h i s t o r y of i n f e r t i l i t y or impaired fecundity who become pregnant d i f f e r e n t from those who have no h i s t o r y of i n f e r t i l i t y ? In p a r t i c u l a r , i s there a r e l a t i o n s h i p between infant health and a h i s t o r y of impaired fecundity? Are the l i f e s t y l e c h a r a c t e r i s t i c s associated with i n d i v i d u a l s who have d i f f i c u l t y conceiving any d i f f e r e n t from those who do not? A v a r i e t y of reproductive models ( i . e . , macrosimulation and microsimulation) (Menken, 1975; Sheps, 1971) have been developed as a means to study the r e l a t i o n s h i p between measures of f e r t i l i t y and determinants of f e r t i l i t y . F e r t i l i t y i s influenced by seven key v a r i a b l e s : f e c u n d a b i l i t y (the p h y s i o l o g i c a l capacity to produce a 17 l i v e c h i l d ) , exposure (marriage/cohabitation and marriage d i s r u p t i o n ) , the period of postpartum i n f e c u n d a b i l i t y , contraception, onset of permanent s t e r i l i t y , spontaneous i n t r a u t e r i n e mortality and induced abortion. Bongaarts and Jones (1982) r e f e r to these v a r i a b l e s as proximate determinants—the b i o l o g i c a l and behavioral factors through which s o c i a l , economic, psychological, and environmental v a r i a b l e s a f f e c t f e r t i l i t y . Reproductive models are important tools for those involved i n population studies (Menken, Stein, & Susser, 1982) and they also help to c l a r i f y l i m i t a t i o n s i n the study of health e f f e c t s on f e r t i l i t y . For example, while there i s evidence of an i n t e r a c t i o n of s p e c i f i c health f a c t o r s , such as n u t r i t i o n , with both fecundity and f e r t i l i t y (Stein & Susser, 1982), population models in d i c a t e that the e f f e c t of health factors on fecundity, except i n the extremes, i s far l e s s powerful than the e f f e c t of the s o c i a l v a r i a b l e s such as marriage, and behavioral v a r i a b l e s such as contraceptive use, which promote or i n h i b i t f e r t i l i t y . F e r t i l i t y d i f f e r e n t i a l s cannot be a t t r i b u t e d conclusively to these health factors because the confounding behavioral va r i a b l e s are d i f f i c u l t , i f not impossible, to co n t r o l adequately i n the a n a l y s i s . These models may not be s u f f i c i e n t l y s e n s i t i v e , therefore, to assess health e f f e c t s (as compared to s o c i a l e f f e c t s ) at the community l e v e l . In smaller populations the weight of factors associated with f e r t i l i t y are l i k e l y to s h i f t . A s i m i l a r conclusion was reached by Bonham and Placek (1978) i n 18 th e i r attempt to examine the r e l a t i o n s h i p of-maternal health, i n f a n t health and sociodemographic factors to f e r t i l i t y by means of cross-s e c t i o n a l data from the 1973 US National Survey of Family Growth and the 1972 National N a t a l i t y Survey. Data have been reported on each of the proximal determinants of f e r t i l i t y by the US National Center for Health S t a t i s t i c s as c o l l e c t e d i n seven n a t i o n a l , c r o s s - s e c t i o n a l surveys of f e r t i l i t y and family planning c a r r i e d out between 1955 and 1976 (Mosher, 1982). Results from 1982 data c o l l e c t e d during interviews with a n a t i o n a l l y representative sample of 7,600 US women between 15-44 years i n 1982 have yet to be reported. These surveys represent the only substantive North American data on fecundity status of the population. The data are informative to the extent they describe the prevalence of impaired fecundity reported by married women of d i f f e r e n t age groups, p a r i t y , r a c i a l o r i g i n , region, r e l i g i o n , educational attainment, and household occupation—they are not appropriate for examining cause and e f f e c t r e l a t i o n s h i p s , but may i d e n t i f y trends or v a r i a t i o n s which deserve further study. The sample a v a i l a b l e f o r interview may underrepresent some groups of women (for example, those without c h i l d r e n may be more l i k e l y to be away from home) and the information reported i s subject to r e c a l l b i a s — e i t h e r s i t u a t i o n could r e s u l t i n conservative estimates. For these reasons, the value of carrying out these surveys has been debated (Ryder, 1973, 1975; Westoff, 1975). Of i n t e r e s t to t h i s study i s the report of reproductive impairments 19 among 6,482 married couples who were interviewed for the 1976 National Survey of Family Growth [NSFG] (Mosher & P r a t t , 1982) which describes fecundity and f e r t i l i t y status, and reproductive h i s t o r y v a r i a b l e s f or t h i s population. Fecundity was defined i n the NSFG survey as the p h y s i o l o g i c a l capacity of a couple to produce a l i v e c h i l d and information was co l l e c t e d according to s i x mutually exclusive categories: the s u r g i c a l l y s t e r i l e f o r contraceptive reasons comprising 19 percent of the sample population; the s u r g i c a l l y s t e r i l e for noncontraceptive reasons which described 10 percent; the nonsurgically s t e r i l e , one percent; the subfecund, 10 percent; those with long i n t e r v a l i n f e r t i l i t y , 4 percent; and those who were fecund, the remaining 56 percent of the survey population. The term "impaired fecundity" was used to describe a l l couples for whom i t was p h y s i c a l l y d i f f i c u l t or impossible to conceive or carry a baby (or another baby) to term. This group consisted of the nonsurgically s t e r i l e couples ( i . e . , s t e r i l e f o r reasons such as accident or i l l n e s s ) , the subfecund couples (for whom i t i s d i f f i c u l t but may be possible to conceive and carry a pregnancy to term), and those who did not achieve a pregnancy a f t e r three continuous years of unprotected exposure (long i n t e r v a l i n f e r t i l i t y , presumed s t e r i l e ) . Couples with impaired fecundity represented 16 percent of the survey population. Of t h i s group, the subfecund couples (14 percent) represent the group to be i d e n t i f i e d i n a reproductive health index on the assumption that t h e i r i n f e r t i l i t y 20 i s not permanent and t h e i r reproductive health could be improved with treatment. About 47 percent of a l l couples with impaired fecundity wanted to have a baby, with the majority of c h i l d l e s s couples (75 percent) and couples with only one c h i l d (57 percent) wanting a c h i l d . This suggests that the proportion of married couples i n the 15-44 year old age group who t r y to conceive but are unable to become pregnant i s around 7 percent of t h i s population. In the United States a 7 percent estimate represents close to 2 m i l l i o n couples i n 1976 terms. In general, prevalence data can help to e s t a b l i s h estimates of the "at r i s k " population l i k e l y to seek medical treatment for i n f e r t i l i t y or other reproductive problems, such as spontaneous i n t r a u t e r i n e l o s s , or can be used to derive estimates of norms, such as age of natural menopause, for evaluating research on health factors and reproductive performance ( K r a i l o & Pike, 1983). Menken and Sheps (1970), however, have demonstrated that prevalence cases are a biased sample without f u l l knowledge of d u r a t i o n — i n c r o s s - s e c t i o n a l studies, the cases with longer duration are more l i k e l y to be included while i n a cohort study, the bias i s i n the opposite d i r e c t i o n unless a l l cases are followed to t h e i r conclusion. For a recurring condition, the e f f e c t of time of measurement xs even greater. In the case of impaired fecundity, the duration of the problem can range from one year (by d e f i n i t i o n ) to a l i f e t i m e , and i t may also occur more than once for the same or d i f f e r e n t reasons (i.e., i n f e c t i o n , severe weight l o s s ) . These problems of 21 measurement make accurate estimates of community prevalence d i f f i c u l t , and t h i s d i f f i c u l t y c a r r i e s over to attempts to measure the r i s k of occurrence of impaired fecundity re l a t e d to l i f e s t y l e f a c t o r s . I n f e r t i l i t y , the l a r g e s t component of impaired fecundity, i s more commonly reported i n the medical l i t e r a t u r e . I n f e r t i l i t y i s defined as "the i n a b i l i t y to conceive a f t e r one year of unprotected intercourse" and i n f e r s a condition which warrants c l i n i c a l i n v e s t i g a t i o n . By t h i s d e f i n i t i o n , 10 percent of the 1976 NSFG survey couples were i n f e r t i l e . Incidence data for i n f e r t i l i t y could be established i f physicians were to report a l l f i r s t consultations for treatment of i n f e r t i l i t y to the Health Ministry, and would be preferable to prevalence data for impaired fecundity as a component of the reproductive health index. The common use of contraception means that presentation for medical treatment may coincide with a couple's awareness of the problem, but not n e c e s s a r i l y with the onset or occurrence of i n f e r t i l i t y ; however, the same s i t u a t i o n occurs i n many diseases (e.g., cancer) for which incidence data are reported. The p o t e n t i a l delay between onset and recognition of i n f e r t i l i t y i s again l i k e l y to be a confounding factor i n assessing the impact of l i f e s t y l e . The assumption that one year's f a i l u r e to conceive i s evidence of i n f e r t i l i t y i s based on studies of healthy women (Cooke et a l . , 1981; Cramer, Walker, & S c h i f f , 1979; Vessey et a l . , 1978) where 90 percent report a pregnancy wi t h i n the f i r s t year of exposure: 95-99 percent achieve a pregnancy a f t e r 2 years of exposure and the monthly p r o b a b i l i t y of conception i s estimated at 12-20 percent. 22 Such estimates of the fecundity of the normal population do provide a basis for defining an abnormality and assessing the effectiveness of s p e c i f i c treatment (Cooke et a l . , 1981). L i f e - t a b l e analysis can be used to further r e f i n e expected rates according to v a r i a b l e s of i n t e r e s t such as age, duration of i n f e r t i l i t y , or the type and s e v e r i t y of condition (Katayama et a l . , 1979; Lamb & Cruz, 1972). In the medical l i t e r a t u r e , i n f e r t i l i t y data are generally reviewed i n terms of diagnostic and treatment techniques rather than e t i o l o g i c f a c t o r s . With standardized methods, s p e c i f i c male disorders (e.g., azoospermis, oliogospermia, disorders of m o t i l i t y or morphology) and/or s p e c i f i c female disorders (e.g., tubal damage, ovulation disorders, endometriosis, infections) are l i k e l y to be detected i n 3 out of 4 couples (Cooke et a l . , 1981). However, for approximately 15-24 percent of couples, no d e f i n i t e "cause" or condition can be detected and t h e i r i n f e r t i l i t y i s described as "unexplained" (Moghissi & Wallach, 1983; Templeton & Penny, 1982). The expected incidence of s p e c i f i c conditions rel a t e d to i n f e r t i l i t y i s d i f f i c u l t to determine because c l i n i c - s p e c i f i c reports show a wide range of rates. This may r e f l e c t the lack of major centres with large enough populations to provide a representative sample, or d i s p a r i t y i n methods of i n v e s t i g a t i o n and diagnosis, or r e f e r r a l bias. A study of i n f e r t i l i t y treatment i n Canada i s currently being c a r r i e d out by the f e d e r a l M i n i s t r y of Health and Welfare Canada. I t i s hoped that these data w i l l help to provide s u i t a b l e incidence rates for conditions associated with i n f e r t i l i t y treatment: unfortunately, 23 l i f e s t y l e c h a r a c t e r i s t i c s of the i n f e r t i l e population are seldom reported and are not being c o l l e c t e d f or the Canadian study. There i s obviously i n s u f f i c i e n t data to describe the population affected by/or treated for impaired fecundity or i n f e r t i l i t y i n BC, or to ascer t a i n the determinants and consequences of the problem. I t may be acceptable i n the absence of any information to assume an i n f e r t i l i t y or impaired fecundity rate approximating those determined for the US p o p u l a t i o n — i n t h i s case, a rate of 10 percent of married couples would represent 41,117 couples i n BC based on S t a t i s t i c s Canada data for mid-June, 1983. If physicians were to report the number of women and men treated for i n f e r t i l i t y / i m p a i r e d fecundity as one measure of the reproductive health i n the community, would t h i s be s u f f i c i e n t ? Could the type of problem be estimated from smaller c l i n i c samples? The following study suggests not. Petterson, F r i e s , and N i l l i u s (.1973) reported on the incidence and prevalence of secondary amenorrhea i n Uppsala County, Sweden as derived from a cro s s - s e c t i o n a l r etrospective study (mailed questionnaire survey) of 2,000 women—previous to t h i s , only estimates from very s e l e c t groups of women had been reported with a range of 1.9-100 percent (Drew, 1961). Sampling from the county population r e g i s t r y consisted of a l l women between the ages of 18-45 years born on the 10th and 20th days of each month, and represented one i n f i f t e e n of the female population. A response rate of 93 percent was reported. Non-respondents (n = 138) were more l i k e l y than respondents to be unmarried 24 or i n s o c i a l c l a s s I I I , but were s i m i l a r i n age d i s t r i b u t i o n (the most s i g n i f i c a n t factor) and place of residence. A h i s t o r y of amenorrhea for more than 3 months during the previous year was given by 258 women (13.8 percent) of the 1,862 respondents—183 (9.8 percent) were due to pregnancy, 13 (0.7 percent) were secondary to s u r g i c a l treatment and 62 (3.3 percent) were judged to have secondary amenorrhea. No primary amenorrhea was reported. -The one year incidence rate of secondary amenorrhea of more than 3 months duration was 3.3 percent; the one year prevalence rate was 4.4 percent and the point prevalence rate at the time of the interview was 1.8 percent based on the population at r i s k who were not pregnant or on contraceptive p i l l s . If these rates were to apply to the BC population, some 18,000 women ( S t a t i s t i c s Canada, mid-June, 1983) would be affected. Prevalence rates were lowest i n women aged 25-39 years, i n married women, i n those l i v i n g i n r u r a l areas rather than c i t i e s , i n nonsmokers, i n those who had had ch i l d r e n , and i n those who had experienced early rather than l a t e menarche. In multiple regression analysis only the age factor was found to be s i g n i f i c a n t , but i t i s possible that other factors may be shown to be s i g n i f i c a n t given a larger sample s i z e . The extreme v a r i a t i o n among se l e c t groups from c l i n i c a l studies i l l u s t r a t e s the importance of e s t a b l i s h i n g incidence and prevalence rates f or types of reproductive problems from a representative sample of the t o t a l population. I f enough women were affected by secondary 25 amenorrhea to enable measurement of the impact of various l i f e s t y l e f a c t o r s , t h i s would be a s u i t a b l e outcome measure of impaired fecundity. The value of documenting the l e v e l s and trends of various types of fecundity impairment i s generally discussed i n the context of demographic and s o c i a l planning or medical treatment. How pertinent are these data to planning for improved neonatal health? While there i s evidence, for example, that secondary amenorrhea occurs i n conditions of inadequate n u t r i t i o n , severe weight l o s s , and extremes of a c t i v i t y and s t r e s s ; that endocrine function i s altered by l i f e s t y l e factors and that an increased incidence of spontaneous abortion i s associated with i n f e r t i l i t y treatment; i t i s not possible to r e l a t e these r e s u l t s from s e l e c t populations to occurrences i n the community unless data are c o l l e c t e d on the t o t a l population or a representative sample. Thus, data for the i n f e r t i l i t y / i m p a i r e d fecundity component of the reproductive health index should include: (a) the incidence of i n f e r t i l i t y - r e l a t e d reproductive disorders of males and females of childbearing years, determined as the i n i t i a l contact for treatment of each separate occurrence; (b) the diagnosis given for the disorder; and Cc) other variables consistent with the use of the index to measure the e f f e c t of factors such as l i f e s t y l e . Spontaneous Abortion Abortion re f e r s to the spontaneous or therapeutic termination of pregnancy before the fetus has attained v i a b i l i t y . The current determination of v i a b i l i t y — t h e time at which a fetus i s capable of independent extra-uterine l i f e — i s a f e t a l age of 20 weeks corresponding 26 to a f e t a l weight of approximately 400-500 grams. Beyond 20 weeks of gestation, spontaneous i n t r a u t e r i n e mortality i s defined as s t i l l b i r t h i n most j u r i s d i c t i o n s . An annual general abortion rate or r a t i o may be calculated r e l a t i n g the number of abortions (or spontaneous abortions, i f data are available) to an approximation of the population exposed to the r i s k (midyear population of women 15-44 years), to t o t a l l i v e b i r t h s , d e l i v e r i e s , or known pregnancies. S p e c i f i c abortion r a t i o s can also be calculated f o r selected subgroups of the population over varying time periods. For age-specific abortion rates i t i s important to use maternal age at time of conception i n order to prevent the d i s t o r t i o n which can occur when age at time of abortion i s compared to age at time of l i v e b i r t h — t h i s i s s p e c i f i c a l l y relevant to.the teenage population where large numbers are involved or where age at conception occurs at the upper l i m i t of an age group category (World Health Organization [WHO], 1970). With adequate data, weekly or monthly abortion r a t i o s for successive periods of gestation can be calculated using l i f e - t a b l e procedures. In t h i s case, the sum of the net rates for spontaneous abortions at each time period would equal the spontaneous abortion rate per 1,000 pregnancies for the aggregate of a l l spontaneous abortions or time periods. Information about the occurrence of spontaneous abortion i n a community i s not, however, r e a d i l y a v a i l a b l e . No country reports a complete r e g i s t r a t i o n of abortions as part of v i t a l s t a t i s t i c s — a l t h o u g h some, including Canada, mandate reporting of a l l induced abortions. An 27 accurate estimate of spontaneous abortion i s d i f f i c u l t since early spontaneous abortions are often not recognized by women, or i f detected, may not require medical treatment and, thus, would not be reported. H o s p i t a l discharge data can provide some incidence data but because the proportion of a l l cases of spontaneous abortion that are admitted to h o s p i t a l varies between i n s t i t u t i o n s and communities and the cases are not always c l e a r l y i d e n t i f i e d , t h i s i s not an accurate means of estimating the occurrence of spontaneous abortions i n the population. Estimates of the incidence of spontaneous abortion i n the population are p r i m a r i l y derived from retrospective surveys although some prospective studies have been c a r r i e d out. Confounding v a r i a b l e s that a f f e c t these estimates include: previous h i s t o r y of spontaneous abortion and maternal age; problems of delayed observation/recording, such as s u b c l i n i c a l abortion, induced abortion or delayed menses; and a r t i f i c i a l v a r i a b l e s such as r e c a l l bias i n retro s p e c t i v e studies, s e l e c t i o n b i a s , compensation for pregnancy l o s s , d e f i n i t i o n a l problems and possibly a recurrence a r t i f a c t (Jansen, 1982; Leridon, 1976). S p e c i f i c problems associated with studies of spontaneous abortion are as follows: problems of observation are a factor i n a l l studies; memory a r t i f a c t i s a fac t o r i n a l l r e t r o s p e c t i v e studies; s e l e c t i o n bias i s a factor i n studies of pregnant women but not i n studies of non-pregnant women. In retrospective studies where pregnancy i s a method of recruitment, bias i s minimized by excluding the current pregnancy from analysis (Naylor, 1974; Naylor & Warburton, 1979). 28 Prospective studies of pregnant women are often biased by the exclusion of early spontaneous abortions not re q u i r i n g medical care and can be corrected by using l i f e - t a b l e methodology and excluding spontaneous abortions which occur w i t h i n the f i r s t week of entry to a study (Shapiro, Levine, & Abramowicz, 1971). Abortion-only sequences w i l l be missed when a pregnant sample i s selected, r e s u l t i n g i n some degree of under reporting. The c l i n i c s e t t i n g associated with case s e l e c t i o n may lead to over or under reporting of spontaneous abortion (WHO, 1970). According to embryological (Hertig, 1967; Hertig et a l . , 1959; Shepard & Fantal, 1979; Short, 1979) and endocrinological (Bloch, 1978; Braunstein.et a l . , 1977; Chartier et a l . , 1979; Rosal, Saxena, & Landesman, 1975) studies, the greatest loss of f e r t i l i z e d ova occurs p r i o r to implantation and during the week following implantation. S u b c l i n i c a l spontaneous abortion i s not included i n abortion s t a t i s t i c s but the incidence of s u b c l i n i c a l and c l i n i c a l spontaneous abortion combined (less than 20 weeks gestation) i s estimated at 20-24 percent for a l l pregnancies and at 10 percent for primigravid women 20-29 years of age ( i . e . , optimal reproductive conditions). The a p p l i c a t i o n of l i f e — t a b l e methods to prospective studies of spontaneous abortion (French & Bierman, 1962; Shapiro, Levine, & Abramowicz, 1971) produces a s i m i l a r range of estimates. Studies reporting the e f f e c t of a previous spontaneous abortion on subsequent abortion incidence show that r e s u l t s vary with sampling methods. Where sampling methods selected against the i n c l u s i o n of abortion-only sequences, the incidence of abortion a f t e r one, two and 29 three previous abortions was 22.7-23.7 percent, 26.2-28.4 percent, and 32.2-33.3 percent, respectively.(Naylor & Warburton, 1979; Warburton & Fraser, 1964). In a study of women with a p r i o r h i s t o r y of spontaneous abortion, the incidence was reported to be 20.3 percent a f t e r one abortion, 44.4 percent a f t e r two, and 58.0 percent a f t e r three (Macnaughton, 1964). Jansen (1982) has c r i t i c a l l y reviewed the fourteen major abortion studies; c l a s s i f i e d the sample populations i n terms of t h e i r being pregnant or non-pregnant, with or without known reproductive problems ( i . e . , i n f e r t i l e , " h a b i t u a l " aborters), whether they were studied r e t r o s p e c t i v e l y or prospectively; has i d e n t i f i e d the inherent biases of the studies and suggested ways to correct for these biases. Close agreement ex i s t s on empi r i c a l l y derived incidences of spontaneous abortion i n North American populations, i r r e s p e c t i v e of the method of data c o l l e c t i o n , provided that age, previous abortion h i s t o r y , and gr a v i d i t y are co n t r o l l e d (Jansen, 1982). The normal incidence of c l i n i c a l l y apparent abortion among f i r s t pregnancies i n women under 30 years i s determined to' be i n the range of 8.3-11.0 percent. The o v e r a l l spontaneous abortion incidence for a l l ages was found to be between 12-15 percent. The cumulative percentage for married women with one or more completed pregnancies i s reported to be 25.9 percent (Mosher & Pr a t t , 1982). Table 1 reports incidence data derived from the North American studies of spontaneous abortion for d i f f e r e n t age groups and reproductive h i s t o r i e s . TABLE 1 Number of Currently Married White Women 15-44 Years of Age and Percent Distribution by Number of Reported Spontaneous Pregnancy Losses, According to Age, Parity, and Fecundity Status: United States, 1976 A e arit and Number of No reported 1 or more reported spontaneous pregnancy losses §e> P^ r i y» ^ n women in Total spontaneous fecundity status , .. ... . . . thousands pregnancy loss A l l losses 1 2 3 or more Percent distribution A l l women 24,795 100.0 78.4 21.6 .15.0 4.1 2.5 Age 5,412 100.0 89.0 11.0 9.3 0.9 0.8 918 100.0 88.9 11.1 9.0 0.1 1.9 4,493 100.0 89.0 11.0 9.4 1.0 0.6 10,993 100.0 80.4 19.6 14.1 3.8 1.7 25-29 years 5,806 100.0 85.0 15.0 11.9 2.2 0.9 5,187 100.0 75.2 24.8 16.6 5.5 2.6 8,390 100.0 69.0 31.0 19.8 6.7 4.6 4,339 100.0 71.2 28.8 19.4 5.2 4.1 4,051 100.0 66.5 33.5 20.2 8.2 5.0 Parity 0 4,874 100.0 89.9 10.1 7.0 1.9 1.2 1 4,923 100.0 82.5 17.5 12.3 3.0 2.2 2 6,939 100.0 79.2 20.8 15.5 3.1 2.2 8,059 100.0 68.2 31.8 21.1 7.0 3.7 Fecundity Status Surgically s t e r i l e : 4,781 100.0 75.2 24.8 16.4 5.3 3.1 Noncontraceptive.. 2,404 100.0 61.7 38.3 21.8 10.1 6.4 Impaired fecundity.. 3,701 100.0 67.6 32.4 18.4 7.6 6.5 13,909 100.0 85.2 14.8 12.4 1.8 0.5 Source: Mosher, W. D. & Pratt, W. F. (.1982). o 31 Abortion incidence i s proposed to be a s e n s i t i v e and objective parameter with which to assess d i s t o r t i o n s i n human reproductive physiology. Abortion incidence accompanying a l l but three modes of i n f e r t i l i t y treatment are higher than those found i n the baseline population, and are not explained simply by heightened c l i n i c a l awareness. The increased r i s k may r e s u l t from a residue of incompletely treated reproductive abnormality, may be introduced by therapy, or .may be both. This proposed use of spontaneous abortion as an i n d i c a t o r of reproductive health problems may be equally applicable to assessing the impact of the l i f e s t y l e element on reproductive health, as long as data are derived from the general rather than the c l i n i c a l population. The methodology used i n determining treatment e f f e c t i s reported and evaluated by Jansen (1982). Methodology s p e c i f i c a l l y used to examine l i f e s t y l e influences on spontaneous abortion i s reported by K l i n e and colleagues (Kline et a l . , 1980; Strobino et a l . , 1980; Warburton et a l . , 1980). In attempting to determine the data that are important to an index of reproductive health and to the assessment of l i f e s t y l e e f f e c t , the following issues are considered: Studies which include morphological examinations of abortuses report an incidence of abnormalities ranging from 30-60 percent,(Poland et a l . , 1981; Warburton et a l . , 1980). Abnormalities which lead to i n e v i t a b l e abortion (usually cytogenetic) are more l i k e l y to occur i n the early weeks of a pregnancy and can represent 80-90 percent of very early abortions. On the other hand, the proportion of abortuses without abnormalities seem to be higher i n the 12th-18th week of gestation and an excess of normal karyotype 32 f e t a l abortions are purportedly associated with some l i f e s t y l e f a c t o r s . Could a rough estimate of cytogenetic involvement be deduced from reports of spontaneous abortion by gestational age? Abnormalities can also r e s u l t from non-cytogenetic factors a f f e c t i n g development at the time of the i n s u l t . Is i t possible to i d e n t i f y these and the factors associated with them? A d e s c r i p t i v e study of a selected BC population designed to i d e n t i f y the embryos and fetuses that spontaneously abort, as normal or abnormal; to document the abnormalities; and to c o r r e l a t e these findings with factors i n the parental reproductive, contraceptive, medical and genetic h i s t o r y has been reported by Poland and colleagues (1981). Although the sample may be biased i n favor of high r i s k patients who were admitted to h o s p i t a l i n the early stages of threatened abortion, and i s a selected rather than representative population, the study provides the most comprehensive data a v a i l a b l e for BC. The sample included a l l conceptuses spontaneously aborted before 20 weeks of pregnancy at the Vancouver General H o s p i t a l from 1966-1976 and t o t a l l e d 2,020 abortuses from 1,961 women. Of these, 1,126 (56 percent) were embryos (defined as a crown-heel length of <30 mm); 813 (40 percent) were fetuses (defined as 30-180 mm i n length); 31 (2 percent) were hydatiform moles and the remaining 50 could not be categorized. Retrospective data r e l a t e d to parental .history and drug use, i n f e c t i o n s , disease or trauma during the study pregnancy were obtained from personal interview of the mother during the h o s p i t a l period or medical records. 33 Of the 1,939 embryos and fetuses, 1,153 (59 percent) were abnormal. The frequency of abnormalities i n embyros was 84 percent and i n fetuses was 26 percent. A comparison by the authors of the frequencies of defects i n the study fetuses (1966-1976) with those i n terms b i r t h s ( l i v e b i r t h s and s t i l l b i r t h s , 1966-1973) recorded i n the Registry of Handicapped Children and Adults, BC Department of Health, showed that defects occurred i n any given system at l e a s t 10 times more frequently i n the fetus. Tissue culture and chromosome analysis became a v a i l a b l e toward the end of the study and was completed for 228 embryos and 59 fetuses. In t h i s sample abnormal chromosomes were associated with 58 percent (.109/188) of emhryos with growth disorganization and 67 percent (16/24) of other abnormal embryos; with a smaller proportion of normal embryos (.25 percent, 4/16) and abnormal fetuses (22 percent, 4/18) and with only 2 percent (1/41) of normal fetuses. Associated chromosomal anomalies appear to present an adequate explanation for defective development. Ready explanations were not apparent i n cases of normal karyotype despite the f a c t that embryos with growth disorganization have obviously received a severe i n s u l t at a very early stage of development. The authors review the d i f f i c u l t y of e s t a b l i s h i n g or i n t e r p r e t i n g associations between maternal factors and morphology of the abortus given (a) r e c a l l bias, (b) the m u l t i p l i c i t y of factors involved, and (c) the common experience of multiple events occurring during the c r i t i c a l period. Determining causal associations under these conditions requires large numbers complete with d e t a i l e d maternal h i s t o r y and 34 c a r e f u l morphological examination, even with supporting evidence from animal or case-control studies. To examine the asso c i a t i o n between selected maternal factors and conceptus outcome, abortuses were c l a s s i f i e d into 5 groups (normal and abnormal embryo, embryo with growth disorganization, normal and abnormal fetus) and temporarily c l a s s i f i e d as early (embryos) versus l a t e (fetuses) abortions. L i f e s t y l e f a c t o rs with the exception of drug use and alcoholism were not examined. Results suggest i t i s possible to i d e n t i f y ( i n a general fashion) factors of importance to spontaneous abortion through within-group di f f e r e n c e s i n the proportion of l a t e to early abortions ( i - e . , ethnic groups, blood groups-RH f a c t o r , o b s t e t r i c h i s t o r y c l a s s i f i c a t i o n s ) . For example, r e s u l t s reported for o b s t e t r i c h i s t o r y show that women who had had no previous pregnancy or no previous successful pregnancy had a s i g n i f i c a n t l y increased proportion of early abortions ( s p e c i f i c a l l y of embryos with growth di s o r g a n i z a t i o n ) , and women who had had l i v e b i r t h s and pregnancy loss (either spontaneous abortion or s t i l l b i r t h ) had a s i g n i f i c a n t l y increased proportion of l a t e abortions of normal fetuses. Ethnic group v a r i a t i o n i n incidence of both l a t e and early, and normal and abnormal abortus, appear s i m i l a r to the s o c i a l class v a r i a t i o n s reported by Alberman and colleagues (1976). I t i s not clear whether these v a r i a t i o n s are r e a l or due to sample s e l e c t i o n . The following i s concluded from t h i s review: Data for the spontaneous abortion component of an index of reproductive health should include (a) the incidence of c l i n i c a l l y recognized spontaneous 35 abortions by gesta t i o n a l age, categorized as embryos (crown-heel length <30 mm) or fetuses (crown-heel length of 30-180 mm); (b) data describing p r i o r obstetric/reproductive h i s t o r y ; and (c) other v a r i a b l e s consistent with the use of the index as a t o o l for measuring e f f e c t of f a c t o r s , such as l i f e s t y l e , on outcome. S t i l l b i r t h S t i l l b i r t h r e f e r s to an infant born without any sign of l i f e at d e l i v e r y and i s distinguished from other i n t r a u t e r i n e mortality, such as l a t e f e t a l abortion, by gestational age or i n some cases, birthweight. Variations i n the d e f i n i t i o n of s t i l l b i r t h s for v i t a l s t a t i s t i c s reporting include r e s t r i c t i n g the d e f i n i t i o n to infants of 20 weeks gestation or more, to infants of 28 weeks gestation or more; to infants of 500 g or more, or to infants of 1,000 g or more. In the 1960s, S t a t i s t i c s Canada revised i t s d e f i n i t i o n to conform to recommendations by the World Health Organization that s t i l l b i r t h r e f e r to i n t r a u t e r i n e death " a f t e r at l e a s t 20 weeks" g e s t a t i o n — p r i o r to t h i s , the d e f i n i t i o n i n use made reference to the 28th week of pregnancy. To allow f o r i n t e r n a t i o n a l comparison and "change over time" trend analyses, s t i l l b i r t h s are reported for both 20 and 28 weeks gestation or more. Regi s t r a t i o n of s t i l l b i r t h s i s mandatory i n a l l developed countries and the r e s u l t i n g v i t a l s t a t i s t i c s are the primary data source. Supplementary data, such as clinico-pathology s e r i e s comparing c h a r a c t e r i s t i c s of s t i l l b o r n infants to those of infants dying i n the early neonatal period, are occa s i o n a l l y reported i n the l i t e r a t u r e . 36 S t i l l b i r t h r e g i s t r a t i o n data provide a composite of the information required f o r l i v e b i r t h and for infant death r e g i s t r a t i o n . Problems of measurement concerning c r i t e r i a for cause of death are s i m i l a r to those encountered with a l l infant death reports and are discussed i n the following section on neonatal m o r t a l i t y . A s t i l l b i r t h rate can be determined by r e l a t i n g the annual number of f e t a l deaths of a s p e c i f i c period of gestation to the number of l i v e b i r t h s plus f e t a l deaths as follows: annual number of f e t a l deaths of s p e c i f i c period of gestation Annual S t i l l b i r t h Rate = • X 1,000 annual number of l i v e b i r t h s plus annual number of f e t a l deaths of s p e c i f i c period of gestation Both the numerator and denominator are affected i f f e t a l deaths are inadequately reported and i f f e t a l age i s misjudged at the 20 week boundary between l a t e abortion (which are not required to be registered) and s t i l l b i r t h . A s t i l l b i r t h or f e t a l m o r tality r a t i o can be used to r e l a t e f e t a l deaths of a stated g e s t a t i o n a l period to l i v e b i r t h s alone. Questions of i n t e r e s t to t h i s study include: What proportion of c l i n i c a l l y recognized pregnancies are reported as s t i l l b i r t h s ? At what gestational age and birthweight do s t i l l b i r t h s occur? What factors are .associated with s t i l l b i r t h s and do these vary with gestational age? What proportion of s t i l l b i r t h s are associated with f e t a l growth retardation, b i r t h defects, or p l a c e n t a l pathologies? Are the reproductive outcomes of i n d i v i d u a l s with a h i s t o r y of a s t i l l b i r t h 37 d i f f e r e n t from those who have no h i s t o r y of reproductive problems, or from those who have a h i s t o r y of l a t e spontaneous abortion or early neonatal mortality? Are the l i f e s t y l e c h a r a c t e r i s t i c s associated with i n d i v i d u a l s who have experienced a s t i l l b i r t h d i f f e r e n t from those who have no h i s t o r y of reproductive problems, or from those who have a h i s t o r y of spontaneous abortions or of neonatal mortality? A number of these questions can be answered by a v a i l a b l e data. In BC the average s t i l l b i r t h rate for the period 1978-1983 was 7.8 with an annual rate ranging from 7.1-8.8 as seen i n Table 2 below. TABLE 2 T o t a l Number of S t i l l b i r t h s * and Rate per 1,000 Live and S t i l l b i r t h s f o r B r i t i s h Columbia, 1978-1983 Year S t i l l b i r t h s L i v e b i r t h s Live & S t i l l b i r t h s No. No. Rate/1,000 1978 330 37,231 8.8 1979 298 38,432 7.8 1980 297 40,104 7.4 1981 368 41,679 8.8 1982 309 42,942 7.1 1983 312 43,047 7.3 Source: Province of B r i t i s h Columbia, D i v i s i o n of V i t a l S t a t i s t i c s . Twenty weeks gestation and over. 38 Table 3 reports percentage of t o t a l s t i l l b i r t h s f o r each of nine birthweight categories for 1979-1983. On average over ..the five-year period, 70 percent of s t i l l b o r n i n fants weighed less than 2,500 g and ha l f of these infants weighed 1,500 g or l e s s . TABLE 3 To t a l Number of S t i l l b i r t h s * and Percent D i s t r i b u t i o n by Birthweight Category for B r i t i s h Columbia, 1979-1983 Birthweight Categories i n D i s t r i b u t i o n of S t i l l b i r t h s by Weight 1979 1980 1981 1982 1983 Grams % % % % % <500 15.0 17.2 17.6 17.3 23.0 501-1,000 20.1 21.8 16.5 17.0 20.4 1,001-1,500 11.8 11.9 12.8 12.5 10.1 1,501-2,000 6.7 8.6 9.2 11.5 11.7 2,001-2,500 10.2 6.6 9.2 6.7 6.3 2,501-3,000 7.7 11.9 8.9 11.5 11.6 3,001-3,500 11.2 8.2 12.6 10.3 9.1 3,501-4,000 7.4 6.6 4.5 7.4 2.5 >4,001 3.8 2.3 3.3 3.2 2.2 Unknown 6.1 4.9 5.3 2.6 3.1 S t i l l b i r t h s % 100.0 100.0 99.9 100.0 100.0 N 313 303 358 312 318 Source: Province of B r i t i s h Columbia, D i v i s i o n of V i t a l S t a t i s t i c s . Defined as 20 weeks gestation or more. 39 Table 4 represents percentage of t o t a l s t i l l b i r t h s born by category of g e s t a t i o n a l age, for 1979-1983. Two peaks con s i s t e n t l y occurred around 20-25 weeks gestation and around term (38 weeks and over), each representing approximately 30 percent of the t o t a l s t i l l b i r t h s . One-h a l f to two-thirds of the f e t a l deaths a f t e r 37 weeks were to infants weighing 3J500 g and above, which suggests a d i f f e r e n t pathology than i n f e t a l death of these very small i n f a n t s . A possible reporting bias was observed i n the choice of even- over odd-numbered weeks of gestation and numbers reported for the 20th and 40th week of gestation. TABLE 4 Total Number of S t i l l b i r t h s * and Percent D i s t r i b u t i o n by Gestational Age for B r i t i s h Columbia, 1979-1983 Gestational Age Groups i n D i s t r i b u t i o n of S t i l l b i r t h s by Weight 1979 1980 1981 1982 1983 Weeks % % % % % 20-25 26.2 33.0 27.4 25.6 33.6 26-28 12.1 7.6 9.5 10.3 11.0 29-31 5.8 7.6 7.8 6.4 7.6 32-34 9.9 8.3 10.9 10.9 9.4 35-37 11.8 16.5 15.1 11.9 10.4 38-41 27.5 23.1 25.6 29.8 24.8 42- 4.8 3.6 3.4 5.1 3.1 Unknown 1.9 0.3 0.3 - -S t i l l b i r t h s % 100.00 100.0 100.0 100.0 99.9 N 313 303 358 312 318 Source: Province of B r i t i s h Columbia, D i v i s i o n of V i t a l S t a t i s t i c s Defined as 20 weeks gestation or more. 40 Of the 224 s t i l l b o r n infants born l a t e r than 24 weeks gestation i n BC i n 1983, 72 (32 percent) were growth retarded (defined as -2SD or less) when compared to the f e t a l growth standard developed by Usher and McLean (1969). Review of the s t i l l b i r t h l i t e r a t u r e suggests the following information would be appropriate to the s t i l l b i r t h component of a reproductive health index: (a) numbers and rate of s t i l l b i r t h s ; (b) birthweight by gestational age; (c) the proportion of s t i l l b o r n infants affected by growth re t a r d a t i o n (defined by weight for gestational age at -2SD or below a designated growth standard, or by weight and crown-heel length below the 5th or 10th p e r c e n t i l e for gestational age; (d) the proportion of s t i l l b o r n infants with b i r t h defects, and the type of defect; (e) the proportion and type of major p l a c e n t a l pathologies associated with s t i l l b i r t h s ; (f) data describing p r i o r obstetric/reproductive h i s t o r y ; and (g) other v a r i a b l e s consistent with the use of the index as a t o o l for measuring e f f e c t of f a c t o r s , such as l i f e s t y l e , on outcome. With only minor change to the current manner of reporting or analyzing s t i l l b i r t h data, information for a l l categories except the l a s t , could be a v a i l a b l e for the BC community. Infant M o r t a l i t y The established v i t a l s t a t i s t i c s system i n developed countries provides information on the absolute frequency of deaths within the f i r s t year of l i f e and provides the appropriate numerators and denominators for comparison of deaths within or between a p a r t i c u l a r population, area or time period. 41 The simplest method of making comparisons i s to compute rates. The most d i r e c t r e l a t i o n s h i p i s expressed by r e l a t i n g deaths associated with pregnancy and b i r t h s to the population exposed to the r i s k of death. Since infant deaths account for most of the deaths occurring i n childhood, mortality rates are also reported for age-specific periods wi t h i n t h i s f i r s t year a f t e r b i r t h . Four mutually exclusive rates are r o u t i n e l y reported: f e t a l deaths ( s t i l l b i r t h s , as described p r e v i o u s l y ) ; early neonatal; l a t e neonatal; and post-neonatal deaths. The early neonatal death rate measures the rate of death during the f i r s t 7 days of l i f e and i s computed by r e l a t i n g deaths at ages under 7 days to l i v e b i r t h s as follows: annual number of deaths under 7 days of age Annual early neonatal = X 1,000 mo r t a l i t y rate annual number of l i v e b i r t h s F e t a l and early neonatal deaths are often grouped together and the r e s u l t i n g p e r i n a t a l m o r t a l i t y rate measures the r i s k of death from the 20th week of gestation to 7 days a f t e r b i r t h and i s computed as follows: Annual number of f e t a l deaths of s p e c i f i c period of gestation plus annual number of deaths under 7 days of age Annual p e r i n a t a l = X 1,000 mortality rate Annual number of l i v e b i r t h s plus annual number of f e t a l deaths of s p e c i f i c period of gestation Beyond the p e r i n a t a l period, the l a t e neonatal m o r t a l i t y rate measures the r i s k of death between the age of 7 and 27, days of l i f e and the postneonatal mortality rate measures t h i s r i s k from 28 days of l i f e to one year of age. Both these rates use the annual number of l i v e b i r t h s as a denominator. Figure 2 summarizes the age-specific d i v i s i o n s commonly used to analyze infant death trends. S t i l l b i r t h s F e t a l Deaths P e r i n a t a l Deaths { f e t a l period-6 days) Ea r l y Neonatal Deaths' (0-6 days) Late Neonatal Deaths (7-27 days) Post-neonatal Deaths (28-365 days) Pos t - p e r i n a t a l Deaths (7-365 days) Neonatal Deaths (0-27 days) Post-neonatal deaths (28-365 days) FIGURE 2 Age-Specific C l a s s i f i c a t i o n s for F e t a l and Infant M o r t a l i t y S p e c i f i c rates, defined i n terms of one or more c h a r a c t e r i s t i c s of the population ( i . e . , cause-of-death), and adjusted ( i . e . , age-adjusted) rates are frequently used i n de t a i l e d analysis of infant deaths. The maternal mortality rate measures the r i s k of death from d e l i v e r i e s and complications of pregnancy, c h i l d b i r t h and puerperium i n the group exposed to r i s k , which should consist of a l l women who have been pregnant at some time during the period. However, for lack of accurate data, t h i s rate (more accurately termed a r a t i o ) i s usually based on the t o t a l number of l i v e b i r t h s and does not account f o r multi p l e b i r t h s or f e t a l deaths. The r a t i o , commonly expressed as per 100,000 b i r t h s , i s computed as follows: Annual number of deaths a t t r i b u t e d to maternal conditions Annual maternal = X 100,000 mortality rate Annual number of l i v e b i r t h s The maternal mortality rate for BC has been reported at the l e v e l of 0.1 per 1,000 l i v e b i r t h s each year since 1972. A recent study which examined the extent to which v a r i a t i o n s i n the d e f i n i t i o n of maternal death a f f e c t the number of deaths reported i n the US na t i o n a l s t a t i s t i c s (Smith et a l . , 1984) found the actual incidence of maternal mo r t a l i t y i n the US for 1974-1978 to be 20-30 percent higher than published s t a t i s t i c s . For 1978, the incidence rate based on study r e s u l t s was 12.1 per 100,000 b i r t h s compared to the o f f i c i a l report of 9.6 per 100,000 b i r t h . The accuracy of the number of maternal deaths reported was seen to be affected by (a) whether reporting of pregnancy status was required for a death c e r t i f i c a t e ; (b) the coding c r i t e r i a used to assign the underlying cause of death; and (c) d e f i n i t i o n s regarding which underlying causes of death are considered maternal causes. Coding according to ICD-9-CM d e f i n i t i o n s of maternal deaths 44 increased the number by 10 percent above coding according to ICDA-9 (World Health Organization [WHO]., 1979). In general, however, records of mortality as reported by Divisions of V i t a l Statistics are more reliable than morbidity records, since death is easier to define than illness. Reports of death "from a l l causes" are more reliable than disease-specific mortality reports since the latter can be subject to change in diagnostic c r i t e r i a . Problems encountered in interpreting observed differences in mortality rates or ratios are similar to a l l v i t a l s t a t i s t i c s analysis and include: incomplete reporting, underregistration, ambiguous or overlapping classifications. Valid conclusions about differences must be based on comparable data that differs only with regard to the attribute(s) purported to be associated with the difference. The differences in v i t a l rates and ratios are usually related to social, economic, biologic or medical factors, therefore, valid conclusions can only stem from data which are descriptive enough to allow proper assessment of comparability. Grove and Hetzel (1968) provide an excellent review of the analysis and interpretation of v i t a l s t a t i s t i c s and in the a r t i c l e state, "imperfections in data and inconsistencies between numerator and denominator do not necessarily make a rate useless. It i s important that the imperfections be recognized and evaluated . . . a l l v i t a l s t a t i s t i c s measures should be used with f u l l knowledge of their limitations in respect to the primary interest to be served" (p. 17). A recent editorial in the American Journal of Public Health (Zemach, 1984) reviewed "what the v i t a l s t a t i s t i c s system can and 45 cannotido." V i t a l records of b i r t h s and deaths are a l e g a l requirement and need only state the time and place of a v i t a l s event along with enough information to uniquely i d e n t i f y the i n d i v i d u a l who dies or i s born. The content of b i r t h and death records has expanded f a r beyond these l e g a l requirements to support the needs of researchers, demographers, and epidemiologists seeking s t a t i s t i c a l information and, as such, provides an invaluable resource for assessing population health status. But given the l i m i t a t i o n s of the r e g i s t r a t i o n process o v e r a l l , most research that goes beyond broad trends or major categories must look to a d d i t i o n a l sources of data. Af t e r commenting on a report (Smith et a l . , 1984) of the underestimation of n a t i o n a l maternal mortality, Zemach (1984) contends that "a s i m i l a r issue of how much to expect from the v i t a l records system extends to the study of p e r i n a t a l m o r t a l i t y . " She suggests that with a t t e n t i o n now focused on the core problems (preterm d e l i v e r y and low birthweight) re l a t e d to the small numbers of remaining deaths, i t i s perhaps time to reconsider the h i s t o r i c a l categories of the v i t a l records system ( i . e . , include f e t a l death reporting with infant deaths; l i n k maternal deaths with pregnancy outcome, etc.) and to enhance the r o l e of the system as a multipurpose source of health-related data. In any attempt to investigate change i n mortality over time and to assess what diseases have contributed to change, problems r e l a t e d to the categorization of diseases by ICD coding format must be recognized. For example, the disease category reported may depend on any one of the 46 following: (a) the thoroughness of the i n v e s t i g a t i o n , (b) the diagnostic habits of the patient's physician, (c) the medical knowledge and opinion at any given time or place, (d) the ranking of importance between morbid anatomy and aetiology-focussed reporting,(,e) the way i n which borderline conditions are enumerated between two possible diagnoses, and (f) the way i n which nomenclature and c l a s s i f i c a t i o n changes are handled following any ICD r e v i s i o n (Davis & Dobbing, 1974 , Zemach, 1 9 8 4 ) . Leek (1974) states "the r u l e that analysis of deaths by cause should generally be based on categories of the ICD must be questioned as f a r as p e r i n a t a l deaths are concerned" (p. 7 0 9 ) . The ICD represents a compromise between the anatomical (or h i s t o l o g i c a l ) and a e t i o l o g i c a l concepts of diagnosis. Both concepts are accommodated although i n a majority of cases preference i s given to anatomical reporting. The problem of reporting only the morbid anatomy at death,or aetiology, i s greatest for p e r i n a t a l deaths and creates a r t i f i c i a l d i s t i n c t i o n s between causes that can be one and the same. For example, a death c e r t i f i c a t e may describe an infant's condition at death as immaturity, growth retardation, .or b i r t h asphyxia, or may describe the maternal, p l a c e n t a l , or cord conditions which may have led to the premature labor, growth retardation, or asphyxia, or both. Leek (1974) goes on to suggest that "more may be achieved by c o l l e c t i n g c l i n i c a l and pathological summaries of representative s e r i e s of p e r i n a t a l deaths, or standard forms, and arranging for each case to be a l l o c a t e d to one c a r e f u l l y defined a e t i o l o g i c a l and (one c a r e f u l l y defined) p a t h o l o g i c a l category by a s i n g l e assessor 47 or team" (p. 709). A e t i o l o g i c a l c l a s s i f i c a t i o n which distinguishes between o b s t e t r i c a l causes and environmental causes has proved valuable i n i d e n t i f y i n g means for preventing p e r i n a t a l death (Butler & Bonham, 1963). In general, routine v i t a l s t a t i s t i c s , as currently reported, have the advantage of being based on larger numbers than s p e c i a l surveys, but the disadvantages of including fewer c l i n i c a l and p a t h o l o g i c a l d e t a i l s and being more susceptible to bias by the diagnostic discrepancies already noted. Where.disorders are caused by a combination of many influences, c l a r i f i c a t i o n of the r e l a t i v e a e t i o l o g i c a l importance of the contributing influences i s key to the preventive approach and cannot be achieved i n the absence of d e s c r i p t i v e reporting and i n v e s t i g a t i o n . I t has been w e l l established that the two factors most c l o s e l y associated with the death of an infant i n the f i r s t weeks and months of l i f e are the g e s t a t i o n a l age and weight of the infant at b i r t h . In addition, infants with serious b i r t h defects are at greater r i s k of dying. The questions of i n t e r e s t to t h i s study r e l a t e to these f a c t o r s , to the v a r i a b l e s that are associated with an i n f a n t being born too small or too soon, and to the manner i n which growth retardation i s distinguished from birthweight and any differences i n the v a r i a b l e s associated with e i t h e r . Further to t h i s , the maternal c h a r a c t e r i s t i c s associated with f e t a l death that are of i n t e r e s t r e l a t e to whether the reproductive h i s t o r y of a mother whose infant dies i s d i f f e r e n t 48 from those whose in f a n t s do not die and are perceived to be healthy; whether the l i f e s t y l e c h a r a c t e r i s t i c s of mothers whose infants die and are growth retarded, premature, or malformed are d i f f e r e n t ; and whether the l i f e s t y l e of mothers whose infants die are d i f f e r e n t from mothers whose infants are a l i v e and healthy or from mothers who experienced other reproductive problems. What questions can be answered from the current v i t a l records system i n BC? Data from the v i t a l s t a t i s t i c s reports c e r t a i n l y confirm the ass o c i a t i o n between infant and f e t a l deaths, birthweight and gestational age. Table 5 summarized the BC l i v e b i r t h , f e t a l , and neonatal m o r t a l i t y s t a t i s t i c s for 1978 and 1983, and Table 6 reports b i r t h w e i g h t - s p e c i f i c mortality rates by age at death, f o r the years 1979-1983. Although a small gradual decline i n infant m o r tality i s seen over t h i s time period, the r i s k of dying remains d i r e c t l y proportionate to an infant's weight category at b i r t h . The 5 percent of infants weighing 2,500 g or less at b i r t h [LBW] account f o r 60-70 percent of a l l p e r i n a t a l deaths and 45-50 percent of infant deaths (RR:17.6) within the f i r s t year. Less than 2 percent of a l l b i r t h s are reported i n the 1,500 g and under birthweight category [VLBW], yet these infants account f o r 50-60 percent of p e r i n a t a l deaths and 25-35 percent of a l l i n f a n t deaths (RR:40.0) i n BC over these s i x years. One i n 5 VLBW and 1 i n 10 LBW infants die compared to l e s s than 1 i n 100 infants who weigh-more than 2,500 g at b i r t h — e v e n i n the l a t e r group, infants i n the 2,501-3,000 g birthweight category have a mortality incidence which i s twice that of heavier = infants (RR:'2.0). TABLE 5 Total Livebirths, Total Population and Birthrate per 1,000 Population; and Total Infant Deaths by Six Age-Specific Categories (Fetal, Early Neonatal, Late Neonatal, Perinatal, Neonatal, and Infant) and Rate per 1,000 Livebirths (or per 1,000 Live and Stillbirths) in British Columbia, 1978-1983 Numbers and Rates of Birth, Fetal and Infant Deaths Age-Specific Mortality Categories 1978 1979 1980 1981 1982 1983 N Rate N Rate N Rate N Rate N Rate N Rate St i l l b i r t h s (Fetal Deaths) 330 8.8 298 7.8 297 7.4 368 8.8 309 7.1 312 7.3 Early Neonatal Deaths (0-6 days) 248 6.7 214 5.6 199 5.0 242 5.8 229 5.3 199 4.6 Perinatal Deaths (Stillbirths + 0-6 days) 578 15.4 512 13.2 496 12.3 610 14.5 538 12.4 511 11.8 Late Neonatal Deaths (7-27 days) 44 1.2 44 1.1 48 1.2 29 0.7 37 0.9 20 0.5 Neonatal Deaths (0-27 days) 292 7.8 258 6.7 247 6.2 271 6.5 266 6.2 219 5.1 Infant Deaths (0-365 days) 472 12.6 434 11.3 442 11.0 424 10.2 423 9.9 374 8.7 Total Livebirths 37,231 14.7 38,432 14.9 40,104 15.2 41,679 15.2 42,942 15.4 43,047 15.3 Total Population 2,530,100 2,571,200 2,640,100 2,744,467 2,783,200 2,812 ,900 Source: Province of British Columbia, Division of Vital Statistics. TABLE 6 Total Number of Livebirths and Infant Deaths: Early Neonatal, Late Neonatal, Post Neonatal, and Infant Death, and Rate per 1,000 Livebirths by Birthweight Category i n B r i t i s h Columbia, 1979-1983 Age Birthweight Categories i n Grams Total at tear <500 501-1000 1001-1500 1501-2000 2001-2500 2501-3000 3001-3500 3501-4000 >4001 Unknown Deaths Death M Rate M Rate M Rate M Rate M Rate M Rate M Rate M Rate M Rate M Rate N 1979 944.4 541.0 162.0 40.3 14.0 4.6 0.9 1.1 0.4 8 209 Early 1980 956.5 440.4 136.8 28.4 16.4 2.6 1.0 1.4 1.3 14 200 Neonatal 1981 900.0 553.2 118.2 54.4 8.7 3.2 1.3 1.4 0.9 14 241. (0-6 days) 1982 1,000.00 550.3 107.3 41.9 9.7 2.0 1.5 1.1 0.9 10 229 1983 777.8 474.4 117.3 26.1 9.4 3.2 1.4 1.3 0.8 11 197 1979 _ 49.2 33.5 5.0 3.7 2.3 0.5 0.3 0.4 1 46 Late 1980 43.5 82.6 31.6 14.2 2.8 1.0 0.8 0.2 0.4 0 48 Neonatal 1981 - 42.6 4.9 7.1 1.5 0.5 0.2 0.5 - 3 28 (7-27 days) 1982 - 11.8 4.9 4.7 2.1 1.5 0.4 0.7 0.5 1 37 1983 - 7.3 10.2 7.1 0.7 0.5 0.1 0.5 0.3 1 21 1979 55.6 123.0 27.9 12.6 14.0 4.9 3.5 3.3 3.5 1 180 Post 1980 - 45.9 52.6 21.3 15.7 6.5 4.2 2.5 2.1 8 195 Neonatal 1981 66.7 35.5 39.4 16.5 8.0 3.0 3.0 2.2 3.9 9 156 (28-365 days) 1982 - 65.1 - 14.0 12.4 5.1 3.3 2.3 0.9 3 157 1983 111.1 29.2 40.8 11.8 6.5 4.3 3.1 2.4 2.5 7 156 1979 1,000.0 713.2 223.4 57.9 31.7 11.8 4.9 4.7 4.3 10 435 Total Infant 1980 1,000.0 568.9 221.0 63.9 34.9 10.1 6.0 4.1 3.8 22 443 Death 1981 966.7 631.3 162.5 78.0 18.2 6.7 4.5 4.1 4.8 26 425 (0-365 days) 1982 1,000.0 627.2 112.2 60.6 24.2 8.6 5.2 4.1 2.3 14 423 1983 888.9 510.9 168.3 45.0 16.6 8.0 4.6 4.2 3.1 19 374 N N N N N N Total N N N N Births 1979 19 122 179 397 1,354 5,663 13,988 12,062 4,838 38,644 1980 22 109 190 422 1,401 5,701 14,380 12,772 5,151 40,170 Livebirths 1981 30 141 203 423 1,374 5,960 14,791 13,369 5,378 41,679 1982 17 169 205 429 1,446 5,901 15,504 13,634 5,628 42,942 1983 18 137 196 422 1,385 5,990 15,350 13,868 5,673 43,047 Source: Province of B r i t i s h Columbia, Division of V i t a l S t a t i s t i c s . o 51 Thirty-three percent of a l l preterm liveborn infants (= <37 weeks gestation) are low birthweight, and h a l f of these are very low birthweight. Less than 2 percent of liveborn term infants weigh under 2,500 g. These proportions increase for f e t a l deaths: 20 percent of s t i l l b o r n infants of 38 weeks gestation or more are low birthweight; 85-95 percent of preterm s t i l l b i r t h s are low birthweight and 70-85 percent are very low birthweight i n BC. In addition, the a v a i l a b l e BC data can (and does) confirm other w e l l known associations with low birthweight and with infant m o r tality — i . e . , a p o s i t i v e a s s o c i a t i o n with younger and older parents, with s i n g l e mothers and/or " i l l e g i t i m a t e " b i r t h s , with multiple b i r t h s , with f i r s t b i r t h s and with a h i s t o r y of previous pregnancy loss or s t i l l b i r t h . A l l information reported in.the r e g i s t r a t i o n of l i v e b i r t h , s t i l l b i r t h , and infant deaths (see Appendix) i s a v a i l a b l e for an a l y s i s , given the constraints already mentioned. Using these data to determine v a r i a b l e - s p e c i f i c m o r tality rates or associations with low birthweight help to i d e n t i f y factors which describe high r i s k groups i n the population, but the type of information currently reported does not help to explain why the p a r t i c u l a r group (e.g., mothers born i n India) i s at r i s k or why t h e i r infants are born too small or too soon. An excess of low birthweight or mortality may be associated with l i f e s t y l e factors such as smoking and poor d i e t ; with poor pregnancy gain and low pregravid weight; with p a r t i c u l a r medical and o b s t e t r i c problems; etc., but documenting t h i s type of information of p o t e n t i a l e t i o l o g i c a l importance i s not required at,present. Without such 52 information, however, community resources cannot be knowledgably al l o c a t e d to areas of intervention which could be e f f e c t i v e i n improving reproductive and neonatal health. An e t i o l o g i c a l perspective could also provide valuable support i n assessing "cause of death" data. To recognize cause of death as short gestation and low birthweight as reported i n some 15 percent of cases; or i n t r a u t e r i n e hypoxia and b i r t h asphyxia, i n about 5 percent of cases; or congenital anomalies of major systems, i n 35-50 percent of deaths; provides no i n s i g h t into factors which contribute to t h i s outcome. Recognizing maternal conditions or p l a c e n t a l , cord or membrane conditions as the cause of death, recognizes contributing factors that are, i n themselves, the outcome of e a r l i e r problems. Pneumonia was amongst the four leading causes of infant death i n 1980 i n BC. Do infants die of i n f e c t i o n because of increased exposure or decreased resistance? Sudden Infant Death [SID] Syndrome now represents 40-50 percent of the postnatal deaths i n B C — i s the postnatal growth ret a r d a t i o n associated with SIDs unique to the postnatal period or an extended p r o f i l e of prenatal growth? We cannot say i f records do not include other growth parameters besides birthweight. With smaller numbers of deaths and improved technological support f o r data analysis, i t seems the appropriate time to suggest a r o l e for the v i t a l records system i n enhancing e t i o l o g i c a l research e f f o r t s . In t h i s way a data base representative of the t o t a l community could be supplemented with a selected population data base assembled from in-depth case studies c a r r i e d out by l o c a l health agencies and independent researchers. 53 Zemach'.(.1984) concludes her e d i t o r i a l : "The challenge to p u b l i c health and other researchers i s to document the types of a p p l i c a t i o n that are and are not s u i t a b l e for the v i t a l records system, and to move toward the establishment of other supplementary uniform n a t i o n a l data sets, instead of posing u n r e a l i s t i c expectations for b i r t h and death r e g i s t r a t i o n . " The index of reproductive health proposed i n t h i s thesis suggests the type of supplementary data which could provide d i r e c t i o n to those planning intervention programs and valuable s t a t i s t i c s for in-depth studies. The following i s concluded from t h i s review: Data for the i n f a n t m o r tality component of a reproductive health index should include (a) the incidence of early, l a t e , and postneonatal mortality; (b) birthweight-s p e c i f i c m o rtality; (c) mortality by birthweight for gestational age; and (d) v a r i a b l e s consistent with the use of the index as a t o o l for measuring e f f e c t of f a c t o r s , such as maternal weight, weight gain, and l i f e s t y l e h a bits, on both birthweight and mortality outcome. F e t a l Growth and F e t a l Growth Retardation The recognition of f e t a l growth, as measured by birthweight, as the most s i g n i f i c a n t predictor of f e t a l and infant m o r t a l i t y , and of ges t a t i o n a l age as the most c r i t i c a l determinant of birthweight has ensured a legitimate focus on reducing the incidence of low birthweight and premature b i r t h i n the population. Studies i n the early 1970s demonstrated that infants who weighed les s than 2,500 g at b i r t h , regardless of t h e i r gestational age, accounted for over two thirds of the infants who died, and for t h i s reason, low birthweight infants 54 (defined as <2,500 g) were singled out as a high r i s k group. Therefore, data regarding birthweight, gestational age, and the incidence of low birthweight and prematurity (<38 weeks) for the majority of l i v e b i r t h s , s t i l l b i r t h s , and neonatal deaths i n the community are r e a d i l y a v a i l a b l e . Subsequently, i t has been recognized that the birthweight of some infants i s very low r e l a t i v e to the birthweight of other infants of the same gesta t i o n a l age and that t h i s represents a f a i l u r e to t h r i v e . These small-for-gestational-age [SGA] or growth retarded [IGR] infants are shown to have an increased p e r i n a t a l m o r t a l i t y and suboptimal postnatal development (Abdul-Karim & Sunderji, 1978; Usher & McLean, 1974; Walther & Ramaekers, 1982). The recognition of growth retardation i s dependent upon a standard of normal growth for infants at each gestational age. Most standards for f e t a l growth have been based on d e t a i l e d c l i n i c a l studies of perceived-to-be healthy mothers and infants but are derived, of necessity, from small populations. The accuracy of the p e r c e n t i l e curves, i n p a r t i c u l a r the c a l c u l a t i o n of gestational age, i s an important a t t r i b u t e of these c l i n i c a l l y derived standards, but t h i s i s o f f s e t by the concern that the population, being small, might not be representative of the general population. For example, the high geographic e l e v a t i o n and low socioeconomic status of the Denver sample (Lubchenco et a l . , 1963), which was for many years the most commonly used growth standard, i s suggested as part of the reason for i t s now apparent d i f f e r e n c e from the majority of comparable standards ( i . e . , 100-300 g lower from 34 weeks gestation). In a review of various growth standards, U l r i c h 55 (1982) found the le a s t v a r i a t i o n i n mean b i r t h length (about 2 percent, generally not exceeding 1 cm i n i n d i v i d u a l g e s t a t i o n a l weeks) and head circumference measurements (approximately 3 percent v a r i a t i o n ) , whereas differences i n birthweight exceed 10 percent from one curve to another, p a r t i c u l a r l y i n the l a t e r weeks. Thus, mean values of length/head circumference r a t i o s show general agreement but ponderal index curves, the r a t i o of length to weight, are l i k e l y to vary by 4-6 percent. Postnatal measurements of b i p a r i e t a l diameter are effected by accommodation of the head during d e l i v e r y ( p a r t i c u l a r l y f i r s t boms, and male infants) and, thus, i n t r a u t e r i n e ultrasound measures are considered to more accurately represent growth a f t e r 36 weeks g e s t a t i o n — f o r example, Campbell and Newman (1971) report a growth rate of 1.2 mm/week a f t e r 36 weeks, whereas postnatal measurements show 1.0 mm/week growth. The c h a r a c t e r i s t i c s of the population that should be used i n developing growth standards has also been a topic of controversy. Should growth standards depict average f e t a l growth i n the population or optimal growth i n the sense that the growth p r o f i l e i s associated with the lowest r i s k of an infant dying and/or the pregnancies perceived to be unaffected by any growth retarding c h a r a c t e r i s t i c s ? Both types of growth standards have been developed. The type of standard used, the degree to which the standard population i s representative of the community and the c r i t e r i a used to assess f e t a l growth retardation w i l l determine the way i n which infant health i s described. The assessment of inf a n t health based s o l e l y on weight for gest a t i o n a l age i s now suggested to be an inadequate measure ( M i l l e r 56 et a l . , 1977; Neligan et a l . , 1976; Ulrich, 1982). This is of particular concern when trying to establish the origins and consequences of fetal growth retardation. The SGA infants are not a homogeneous group in terms of growth dimensions, and their individual prognosis varies widely—some infants are proportionately stunted; others are of expected length, but wasted; and some exhibit a combination of the two types of retardation. It is generally thought that infants with a small ponderal index—ponderal index being weight in grams x 100/body length in centimetres^ —have probably been growth-retarded over a relatively short period of time before delivery. Many SGA infants have appropriate ponderal indices, indicating expected weight for length proportions, but are symmetrically undergrown—the result, i t i s thought, of prolonged growth retardation of early pregnancy origin. Greater severity of growth deprivation is associated with a reduction in both head circumference and crown-heel length. It seems that infants who are both stunted and have a small ponderal index may be the most severely growth-retarded since they appear to have suffered a recent insult superimposed on prolonged growth retardation. Thus, measures of crown-heel length and head circumference (in addition to birthweight and gestational age), and an appropriate multi-measure standard of normal growth are required to identify and describe different types of fetal growth retardation. These data are not available for the majority of communities, including BC, and to date have been generated by a small number of large scale studies. Problems are also seen i n the a v a i l a b l e growth standards for premature b i r t h s . Naeye and Dixon C1978) found evidence that most f e t a l growth standards based on body measurements of neonates born at various gestational ages ( i . e . , c r o s s - s e c t i o n a l data) included inaccuracies i n gestational age assessment. Standard deviation [SD] from mean b i r t h weight values are r e l a t i v e l y larger i n preterm than i n term infants i n a l l of the major growth standards (Babson, Behrman, & L e s s e l , 1970; Brenner, Edelmann, & Hendricks, 1976; Gruenwald, 1966; Lubchenco et a l . , 1963; Tanner & Thomson, 1970; Usher & McLean, 1974). Because v a r i a t i o n s i n f e t a l growth are greatest near term, r e l a t i v e values for standard deviations from mean b i r t h weight values should progressively increase rather than decrease near term. Five of the si x frequently used standards combined measurements from infants who died i n the neonatal period with measurements from those who survived. Since f e t a l growth values are s i g n i f i c a n t l y larger i n preterm infants who survive than i n those who die, Naeye and Dixon (1978) suggest the standards are l i k e l y to be influenced by the i n c l u s i o n of data from growth-retarded neonates and neonates that were older than t h e i r calculated ages. Although the corrected preterm growth standards conform to the l o g i c a l (smaller) standard deviations, the authors suggest there i s no completely v a l i d method presently a v a i l a b l e to determine normal f e t a l growth using measurements from neonates because there i s no assurance that prematurely born neonates are normally grown. Williams and colleagues (1982) present a case for developing growth standards from v i t a l s t a t i s t i c s records i n order to provide 58 accurate information on contemporary patterns of f e t a l growth and v i a b i l i t y i n a community. The populations used for c l i n i c a l l y based growth standards are too small to allow the development of accurate weight- and age-specific v i a b i l i t y standards. Large population-based studies can provide more stable m o r t a l i t y rates but t h i s advantage may be o f f s e t by q u a l i t y c o n t r o l problems with respect to the c a l c u l a t i o n of gestational age and the r e s u l t i n g d i s t o r t i o n of p e r c e n t i l e curves ( p a r t i c u l a r l y those below 37 weeks gestation). Population-based studies have the advantage of being broad based and, thereby, more representative; t h e i r large numbers make i t possible to c o n t r o l for confounding factors such as sex, race, and b i r t h m u l t i p l i c i t y and because they are based on v i t a l records, they can be r e p l i c a t e d p e r i o d i c a l l y and economically to r e f l e c t ongoing improvement or change i n mortality and birthweights. The C a l i f o r n i a study (Williams et a l . , 1982), which developed methodology to minimize q u a l i t y c o n t r o l problems, i l l u s t r a t e s the c l i n i c a l support that can be derived from the comparison of age-specific p e r i n a t a l , neonatal,.and f e t a l m o r t a l i t y rates across selected parameters of i n t e r e s t (e.g., sex, race, multiple b i r t h s , f e t a l growth r e t a r d a t i o n ) . The a b i l i t y to use the same information as both a f e t a l growth standard (which r e f l e c t s average f e t a l growth i n the community) and a v i a b i l i t y standard (which r e f l e c t s optimal f e t a l growth for lowest mortality i n the community) resolves the problems rel a t e d to the exclusive use of e i t h e r average or optimal growth standards and provides the opportunity to monitor growth-related improvements i n infant health 59 outcomes. For example, the optimal birthweight-for-gestational age combinations for the C a l i f o r n i a population i s presently 500 g heavier and 2 weeks longer than the mean birthweight-for-gestational age. Thus, a target for improved infant m o r t a l i t y outcome would be to increase the mean birthweight for a l l gestational age groups i n t h i s community i n addition to the reduction of LBW and preterm b i r t h s . Most recently, population studies of f e t a l growth have developed optimal standards as a means to examine the impact of maternal, infant and health care c h a r a c t e r i s t i c s of pregnancy on f e t a l growth and v i a b i l i t y outcomes. An example of t h i s type of study was c a r r i e d out i n Odense, Denmark, and reported by U l r i c h (1982). The object of the i n v e s t i g a t i o n was to study normal, as f a r as possible optimal, f e t a l growth i n a representative population; to r e l a t e a number of maternal and pregnancy factors to f e t a l growth i n preterm and term d e l i v e r i e s ; and by measurement of several body dimensions, to contribute to the d e s c r i p t i o n of v a r i a t i o n s i n the r e l a t i v e growth of body dimensions i n normal and growth retarded i n f a n t s . Three groups of mothers and i n f a n t s ( a l l singleton b i r t h s ) were studied: (a) 358 normally grown newborns of healthy mothers with uncomplicated pregnancies, (b) 109 preterm infants delivered p r i o r to 38 weeks gestation of "moderately" healthy mothers, and (c) 222 preterm and 130 low weight mature infants and t h e i r mothers. C r i t e r i a f or i n c l u s i o n i n the study was p r i m a r i l y determined by maternal c h a r a c t e r i s t i c s for groups one and two, and by infant c h a r a c t e r i s t i c s for group three. The study was c a r r i e d out at the Odense University Hospital which 60 covered 97 percent of a l l prenatal c l i n i c care and subsequent d e l i v e r i e s i n the municipality for the study period 1972-1974. Mothers to be included i n group one were selected prospectively from the prenatal c l i n i c records according to the following c r i t e r i a : - 17.5 to 35 years at time of d e l i v e r y , - menstrual cycles with <6 days v a r i a t i o n , not exceeding 21-35 days, - >3 regular menstrual periods a f t e r discontinuing o r a l contraception, - date of l a s t menstrual period known, - normal gynaecological examination, - uncomplicated pregnancy, - absence of maternal chronic disease, - complications of e a r l i e r pregnancies l i m i t e d to one abortion or one premature d e l i v e r y or mild preeclamptic toxemia, - smoking habits of <10 cigarettes/day, - both parents born i n Denmark. Records were checked at each subsequent prenatal v i s i t to ensure continued compliance with t h i s c r i t e r i a and at time of d e l i v e r y randomization was c a r r i e d out i n term pregnancies. About 30 percent of a l l mothers met the above i n c l u s i o n c r i t e r i a and 3 out of 7 (13 percent of a l l d e l i v e r i e s ) were included i n the study. Preterm i n f a n t s of mothers meeting t h i s c r i t e r i a were so few i n number (47/358) that these b i r t h s were not randomized and were supplemented by the 109 infants i n group two (whose mothers met a le s s r e s t r i c t i v e i n c l u s i o n c r i t e r i a ) i n order to obtain an adequate study sample. Exclusion c r i t e r i a for p o t e n t i a l group one and two i n f a n t s — w h i c h included the presence of malformation, serious p l a c e n t a l pathology or c o n t r o v e r s i a l g e s t a t i o n a l age (>3 weeks d i s p a r i t y between reported menstrual data and c l i n i c a l assessment)—effected 34 i n f a n t s . Group three was composed of a l l singleton i n f a n t s , born to mothers l i v i n g i n the municipality, 61 with birthweights less than 2,750 g and/or a ges t a t i o n a l age under 38 weeks (266 days). For a l l groups, data were c o l l e c t e d prospectively from the prenatal through the immediate postpartum period. Results from t h i s and s i m i l a r studies emphasize the importance of d i s t i n g u i s h i n g between "normal" infant growth, r e s u l t i n g from healthy pregnancies, and infant growth patterns associated with l e s s optimal prenatal conditions. For example, the fa c t that so few healthy mothers give b i r t h prematurely and the f a c t that most f e t a l growth curves are c r o s s - s e c t i o n a l curves, means that with decreasing g e s t a t i o n a l age such curves represent increasing proportions of abnormal, mostly undergrown, i n f a n t s . U l r i c h (1982) found the following c h a r a c t e r i s t i c s of "normal" growth: a l l body dimensions showed r i s i n g values with advancing gest a t i o n a l age (body length and head circumference measures s i m i l a r to other f e t a l growth studies, birthweights above a l l but Swedish and Norwegian stu d i e s ) ; ponderal index values declined a f t e r week 40 i n f i r s t b o r n infants but continued to r i s e i n laterborn infants; laterborn infants were larger than f i r s t b o r n s with respect to a l l body dimensions, and r a t i o s were s i m i l a r except i n the l a t e s t gestational weeks. A p a r t i c u l a r l y i n t e r e s t i n g f i n d i n g , that i s of relevance i n monitoring the adequacy of intervention programs, concerns the d i f f e r e n t growth patterns seen i n boys and g i r l s of s i m i l a r weight—boys exhibited f a s t e r growth i n body length and head circumference and r a t i o s of body dimensions were consequently d i f f e r e n t . The r e s u l t s suggest that, with normal pregnancies and healthy mothers, g i r l s may a t t a i n an 62-average weight s i m i l a r to that of boys i n f e t a l l i f e , but tend to be plump while boys show f a s t e r s k e l e t a l and head growth. When f e t a l growth patterns of the normal mature newborn infants were examined i n r e l a t i o n to s o c i a l status (upper/lower), maternal weight (heavy/light), employment outside the home (home working/outside employment), and smoking habits (25 percent smoked under 10 ci g a r e t t e s / day versus non-smokers), the following r e s u l t s were reported: infants of l i g h t mothers were smaller than infants of heavy mothers except f o r head circumference and b i p a r i e t a l diameter; infants of upper and lower s o c i a l status were of s i m i l a r s i z e ; smoking reduced average birthweight by 145 g; boys and g i r l s showed d i f f e r e n t growth patterns i n the two maternal weight groups (infants of l i g h t mothers did not exhibit the growth v a r i a t i o n s between sexes as described above); and birthweight of f i r s t b o r n infants varied with maternal pregravid weight. With lower s o c i a l status, employment, and moderate smoking, a s i g n i f i c a n t weight reduction (120 g) was observed i n infants of l i g h t mothers, but not i n infants of heavy mothers. The authors state that t h i s marked d i f f e r e n c e between the two maternal weight groups, i n the e f f e c t of maternal employment, s o c i a l status, and smoking i n pregnancy, indicates an i n t e r a c t i o n between maternal c o n s t i t u t i o n and external influence rather than a mere summation of adverse f a c t o r s — t h i s i n t e r a c t i o n may, with i n the l i m i t s of normal uncomplicated pregnancy, r e s u l t i n newborns who are well-fed and others who show signs of undernourishment. In other words, one p o t e n t i a l l y adverse factor may produce r e s u l t s varying from 63 a considerable reduction i n birthweight to no measureable e f f e c t depending on other circumstances of the pregnancy or c h a r a c t e r i s t i c s of the mother. In group three infants who were preterm or low-weight mature, growth patterns previously described i n normally grown mature infants had l a r g e l y disappeared and new growth patterns were evident: f a s t e r head growth i n boys compared to g i r l s remained for a l l groups and subgroups but fa s t e r l i n e a r growth was no longer observed; premature infants were shorter, and had higher ponderal index and smaller heads than low-weight mature infants of s i m i l a r weight; both groups of infants of smoking mothers had disproportionate retardation of s k e l e t a l growth, compared to infants of non-smoking mothers and sex r a t i o was lowered; and s p e c i f i c types of p l a c e n t a l complications were associated with s p e c i f i c types of f e t a l growth retardation i n both groups of i n f a n t s . Based on the r e s u l t s , the authors suggest that male fetuses may be more s e n s i t i v e to abnormal immunological and n u t r i t i o n a l f a c t ors whereas female fetuses may be more s e n s i t i v e to hypoxia—thus, the impact of a given adverse pregnancy factor may vary according to f e t a l sex. A s i m i l a r but r e t r o s p e c t i v e study of a selected population was c a r r i e d out by M i l l e r and M e r r i t t (1979) who developed normal f e t a l growth standards from a f i v e year study of over 6,000 consecutive b i r t h s at the Kansas Medical Center ( t e r t i a r y care, high r i s k r e f e r r a l center), 1973-1978. Head circumference and b i r t h length growth patterns compared favorably to the Danish study ( U l r i c h , 1982), but birthweight 64 and, thus, ponderal Index dimensions were lower. The purpose of the Kansas study was to provide anthropometric data on newborn infants free from a l l known growth retarding influences i n utero; to construct from these data standards of normal f e t a l growth for comparative use i n determining newborn infants with a t y p i c a l f e t a l growth; and to describe the causes and consequences of a t y p i c a l growth. A unique component of the study was the c r i t e r i a established f or i d e n t i f y i n g growth retarding f a c t o r s . Four groups of growth retarding factors were outlined: f e t a l f a c t o r s (including i n t r a u t e r i n e i n f e c t i o n , chromosomal abnormalities, malformations, multiple b i r t h s ) ; medical complications of pregnancy (including hypertension, pre-eclampsia, chronic disease, i n f e c t i o n , anemia, medications, abnormalities of the placenta); environmental factors (including exposure to toxic substances); and maternal behavior (including abnormal low prepregnancy weight for height, low pregnancy gain, lack of prenatal care, de l i v e r y before 17 or a f t e r 35 years of age, smoking, and the use of addicting drugs or large quantities of alcohol during pregnancy). The presence of any of these factors excluded the mother-infant pair from the "normal" group — two out of three f u l l t e r m i n f a n t s and 85 percent of a l l premature infants born during the study were so excluded. Within the group free of known growth retarding f a c t o r s , pregnancy outcome was excellent: l e s s than 2 percent of infants weighed under 2,500 g and of the 2.4 percent born prematurely, 85 percent weighed more than 2,000 g. Within t h i s group, babies born to women i n the lowest socioeconomic group were at no greater r i s k of being low birthweight 65 or f e t a l growth retarded than those born to mothers i n the highest socioeconomic group. S i m i l a r l y , a mother's age, her m a r i t a l status, race or l e v e l of education were not, per se, related to a higher r i s k of her baby being born too small or too soon. Infants born to mothers with one or more growth retarding behavioral conditions were more l i k e l y to be premature, of low birthweight and growth retarded by each of the measures used (low ponderal index, short-for-dates, and small head circumference)—over h a l f the low birthweight infants were born to women with behavioral conditions and no other known complications of pregnancy. Both premature and f u l l t e r m i nfants with low ponderal index and short-for-dates types of growth re t a r d a t i o n were more l i k e l y to be low birthweight i n t h i s group compared to the group with no known growth retarding factors associated with pregnancy. S p e c i f i c behavioral conditions were associated with s p e c i f i c types of f e t a l growth retardation. Pregnancy outcome was s i g n i f i c a n t l y Worse i f multiple behavior conditions were present. For example, mothers whose only known growth retarding factor was low weight gain had a s i g n i f i c a n t l y increased proportion of preterm b i r t h s (RR:2.9), short-for-dates growth retardation (RR.-1.8) and f u l l - t e r m i nfants who weighed less than 2,500 g (RR:2.8) — i f low weight gain and low pregravid weight-for-height occurred together, the proportion of preterm and short-for-dates infants was s i g n i f i c a n t l y greater than f o r low weight gain alone (RR:12.1 and 4.5 r e s p e c t i v e l y ) . S i m i l a r l y , increase i n preterm b i r t h s (RR:4.1) and short-for-dates growth retardation (RR:3.5) occurred when smoking and 66 low weight gain occurred together. When low weight gain occurred with two or more other behavioral r i s k s , the r e l a t i v e r i s k f o r preterm b i r t h was 8.8 and for short-for-dates growth retardation was 8.2. Mothers with a s i n g l e medical complication had a s i g n i f i c a n t l y increased proportion of premature b i r t h and growth retarded f u l l t e r m infants ( a l l types) when compared to mothers with no complication or adverse behavioral conditions. Many were taking prescribed drugs for t h e i r medical conditions. Mothers with s i n g l e o b s t e t r i c a l problems had a s i g n i f i c a n t l y increased proportion of premature b i r t h s (50 percent were medically induced) and of low ponderal index i n f a n t s — a n d these infants were^more l i k e l y to be low birthweight. Infants born to mothers with both o b s t e t r i c problems and behavioral conditions were twice as l i k e l y to be short-for-dates and three times as- l i k e l y to have a small head circumference as those born to mothers with no such problems. M i l l e r and M e r r i t t (1979) concluded from the study r e s u l t s that i n t r a u t e r i n e growth retardation was notwuncommon and i s generally underreported; that i t i s important to d i f f e r e n t i a t e between the two main types of f e t a l malgrowth (low ponderal index and short-for-dates) since t h e i r pathogenesis are probably d i f f e r e n t and t h e i r prenatal and postnatal courses are d e f i n i t e l y d i f f e r e n t . The authors support previous proposals that low ponderal index infants have a milder form of growth retardation (most exhibited "catch-up" growth within 6 months) of l a t e r pregnancy o r i g i n than that of short-for-dates infants—many of whom remained growth retarded at the end of t h e i r f i r s t year. Short-for-dates growth retardation was most c l o s e l y associated with 67 low weight gain, short maternal stature, and smoking, and the proportion was s i g n i f i c a n t l y higher when any of these conditions occurred together. F e t a l growth retardation i s , of course, not a disease e n t i t y , but the f i n a l r e s u l t of a number of adverse conditions which determine a less than optimal growth and/or development and/or nourishment of the fetus. With the use of multi-measures of f e t a l growth, these studies delineate heterogeneous growth p r o f i l e s i n both the "normal" healthy i n f a n t s , those "at r i s k " of dying, and those who demonstrate adverse and longterm associations with t h e i r f e t a l growth experience. As w e l l , they can describe more p r e c i s e l y the associations between the d i f f e r e n t types of f e t a l growth retardation and the numerous va r i a b l e s a f f e c t i n g growth. In comparison to such studies, there are major gaps i n the currently a v a i l a b l e data with respect to assessment of f e t a l growth. Data for BC are l i m i t e d to birthweight for gestational age. Of the 2,090 l i v e b i r t h s born i n BC i n 1983, at 25 weeks gestation or l a t e r and weighing 2,500 g or l e s s , 608 (29.1 percent) were growth retarded (defined as -2SD) using Usher's and McLean's (1969) f e t a l growth standards; and of the approximately 6,000 l i v e b i r t h s weighing 2,501-3,000 g, 1.8 percent were growth retarded. O v e r a l l , 8.8 percent (715/8,080) of infants weighing 3,000 g or l e s s , and born at 25 weeks gestation or l a t e r , were growth retarded. Both the Odense and Kansas studies found that only 30 percent of t h e i r population could meet c r i t e r i a for the development of a "normal" f e t a l growth standard. 68 At present, there i s no way of i d e n t i f y i n g the "normally" grown infants i n the BC community. The following i s concluded from t h i s review of f e t a l growth retar d a t i o n : Data f o r the f e t a l growth component of a reproductive health index should include (a) the incidence of d i f f e r e n t types of f e t a l growth r e t a r d a t i o n — l o w ponderal index, short-for-dates and small head circumference; (b) f e t a l growth r e t a r d a t i o n - s p e c i f i c m o r t a l i t y and morbidity incidence rates; (c) variables which w i l l enable the development of a population growth standard for normal infants born to healthy mothers with uncomplicated pregnancies; and (d) variables consistent with the use of the index as a t o o l for measuring e f f e c t of fa c t o r s , such as maternal weight, weight gain, and l i f e s t y l e habits, on the various dimensions of f e t a l growth. B i r t h Defects and Malformations The study of b i r t h defects has been characterized by a dichotomy between knowledge of many of the fundamental aspects of genetic regulation and developmental processes, and l i m i t e d understanding of the mechanisms which underlie many of the developmental abnormalities that are c l i n i c a l l y recognized. Emphasis has been given to the need to delineate s p e c i f i c disorders and develop appropriate diagnostic terminology as a basis for c l i n i c a l management and counselling, and for the determination of incidence and prevalence between and with i n populations, and as a pr e r e q u i s i t e to the study of the mechanisms of abnormal development and to the c o l l e c t i o n of epidemiological information. 69 This emphasis i s r e f l e c t e d by the f a c t that i n most developed countries annual n a t a l i t y s t a t i s t i c s include incidence data for s p e c i f i c , as w e l l as general, categories of b i r t h defects based on the system of c l a s s i f i c a t i o n recommended by WHO (1979). There i s often, i n addition, an ongoing s u r v e i l l a n c e system established to monitor the occurrence and prognosis of b i r t h defects (regardless of when diagnosed) which provide accessible supplementary data. Although information about b i r t h defect incidence i n a community i s u sually r e a d i l y a v a i l a b l e , there are recognized problems i n ensuring the data are accurate and complete. The primary data source i s determined by the routine c l i n i c a l examination of the newborn and i s susceptible to errors of misdiagnosis, inadequate reporting, and the f a i l u r e to screen the whole population. The most common problems are r e l a t e d to diagnostic d i f f i c u l t i e s which most often r e s u l t i n underreporting. Very conservative estimates are derived from h o s p i t a l records and b i r t h notices, while more accurate estimates are associated with a system of prospective standardized data c o l l e c t i o n which can allow for delays i n diagnoses. Results of the Collaborative P e r i n a t a l Project (Myrianthopoulos & Chung, 1974) supported e a r l i e r observations that less than h a l f of a l l major or minor malformations present at b i r t h and diagnosed by age 1-2, were detected at b i r t h . To be complete, the incidence of defects and malformations i n a community should cover early and l a t e f e t a l deaths as w e l l as l i v e b i r t h s — f o r reasons discussed e a r l i e r , t h i s composite p i c t u r e i s not generally a v a i l a b l e and reporting becomes progressively le s s complete 70 from l i v e b i r t h s to s t i l l b i r t h s and from l a t e to early f e t a l deaths. Estimates are often assumed from the r e s u l t s of large scale studies of spontaneous abortion. In terms of the a b i l i t y to i n v e s t i g a t e causes and the p o t e n t i a l for prevention of b i r t h defects, the manner i n which these defects are reported i s of p a r t i c u l a r l y importance. Within the framework of established terminology b i r t h defects can be reported according to c l a s s i f i c a t i o n systems which emphasize the anatomical c h a r a c t e r i s t i c s of the defects, the period of development of the defects, or the c l i n i c a l s i g n i f i c a n c e of the defect. This r e s u l t s i n a lack of standardization and d e f i n i t i o n of the terms and nomenclature used i n the f i e l d and diminishes the comparative value of research and study fin d i n g s . An example, given by Polani (1978), r e f e r s to anomalies—congenital defects can present as "anomalies" or i n other ways such as the presence of f o e t a l disease, p e r i n a t a l i l l n e s s , b i r t h trauma, or biochemical derangement. Anomalies i n turn can be "malformations" or embryopathies i f they are defects o r i g i n a t i n g during the period of organogenesis, or "deformations" or foetopathies i f they are defects a r i s i n g a f t e r completion of major organogenesis—both represent alternations i n form or structure but with very d i f f e r e n t outcomes. The incidence of deformities i s higher i n e a r l i e r i n t r a u t e r i n e l i f e (approximately one-third of congenital anomalies are musculo-skeletal deformities), p e r i n a t a l m o r t a l i t y of infants with malformations i s about 40 percent compared to 5 percent with deformations, and i n the l a t t e r , spontaneous 71 recovery or recovery with treatment i s common. Recovery i s rare with malformations. Other terminology may be needed to d i s t i n g u i s h s t r u c t u r a l abnormalities which are due to delay (or error) of t r a n s i t i o n from f o e t a l to postnatal l i f e (e.g., undescended t e s t i s ) or unusual v a r i a t i o n i n s t r u c t u r a l s i z e or function. " A l l i n a l l , " Polani (1978) concludes, "there are 3 d i f f e r e n t methods of c l a s s i f y i n g b i r t h defects. The f i r s t i s based on the c l i n i c a l s i g n i f i c a n c e of the defect i n terms of s u r v i v a l , handicap and c o r r e c t a b i l i t y . A second method depends on the prenatal time at which the defect i s l i k e l y to have arisen. The t h i r d depends on the cause of the defect. The simultaneous use of some or a l l of these guidelines of c l a s s i f i c a t i o n s may help to c l a r i f y studies on b i r t h defects" (p. 423). Smith (1982) recommends a method of c l a s s i f i c a t i o n for patterns of malformations which describes infants according to three c l i n i c a l sub-categories: deformations; malformations (the consequence of a s i n g l e l o c a l i z e d defect i n morphogenesis) which present as an anomaly (a s i n g l e defect at b i r t h ) or an anomalad (multiple defects but s t i l l r e l a t e d to a s i n g l e l o c i ) ; and malformations syndromes (anomalads which are the consequence of multiple defects i n morphogenesis). This type of c l a s s i f i c a t i o n i s of benefit to the c l i n i c i a n i n counselling since the recurrence r i s k s are f a i r l y consistent for deformities (low) and anomalies of a l l kinds (1-5 percent), while for syndromes the r i s k v a r i e s according to the s p e c i f i c aetiology of the p a r t i c u l a r syndrome. I t also has p o t e n t i a l as a c l a s s i f i c a t i o n system to examine causation, i f d i fferences and s i m i l a r i t i e s i n f a ctors associated with anomalies/ anomalads as a syndrome. When data from the Collaborative Perinatal Project (prospective study of 54,452 pregnancies) was re-examined according to this classification system (Smith, 1982), 587 of the 1,488 infants with malformations were considered to have significant multiple malformations (incidence rate, 10.7/1,000 births). Of these, the conditions in 251 infants (43 percent) could be interpreted as the result of a single l o c i producing multiple defects (anomalad); in 143 (24 percent) a specific syndrome was identified, caused either by chromosomal abnormalities (74 infants, 59 with Down's Syndrome), single gene defects (.24 infants) , environmental (34 infants) or unknown factors (11 cases); and in 193 infants (33 percent) the pattern of malformation could not be readily interpreted. The fact that anomalads as a category were responsible for almost half of a l l multiple defect conditions underscores the importance of a developmental approach to classification which can support epidemiological and aetiological study. Classification systems supporting studies of causation and prevention must also be designed to help distinguish between the role of genetic factors and environmental/lifestyle factors since few diseases or defects can be attributed entirely to one or the other, but the contributing influence can vary greatly between and within specific conditions. Most cases of severe genetic disorders are not familial. Phenomena like genetic heterogeneity, where different genes can produce apparently identical outcomes; and phenocopies, where environmental factors can produce abnormalities in the fetus which resemble known hereditary 73 conditions, cannot be distinguished without s k i l l e d c l i n i c a l i n v e s t i g a t i o n (Emery, 1979). There are many recognized causes of congenital abnormalities that include both environmental/lifestyle factors (e.g., i n f e c t i o n s such as r u b e l l a , toxoplasmosis, and cytomegalovirus; drugs such as thalidomide and alcohol; exposure to r a d i a t i o n ; maternal disease) and genetic factors (e.g., s i n g l e gene defects such as the autosomal recessive form of microcephaly; chromosome abnormalities such as the autosomal trisomy syndromes). However, the causes of 65-70 percent of developmental defects are s t i l l unknown—Childs (1978) reports the proportion of defects associated with known causes by categories: 20 percent caused by genetic transmission, 3-5 percent of chromosomal o r i g i n , and 6-9 percent caused p r i m a r i l y by environmental factors ( r a d i a t i o n , <1 percent; i n f e c t i o n s , 2-3 percent; maternal metabolic conditions, 1-2 percent; 2-3 percent). Ultimately, i t i s believed that the cause of developmental aberrations w i l l be understood i n terms of t h e i r a s s o c i a t i o n with abnormal molecular processes (the processes which operate through the medium of macromolecules and r e l a t e d substances to control development) since both.epigenetic and environmental factors are presumedly mediated by molecular events (Epstein, 1978) . Knowledge of normal biochemical processes and of t h e i r genetic c o n t r o l has permitted an understanding of how a genetic abnormality r e s u l t s i n metabolic abnormalities, and of how metabolic processes (the b i o l o g i c a l element of the health f i e l d 74 concept) may be adversely influenced by environmental and l i f e s t y l e elements. Recent publications underscore the importance of being able to document the r e l a t i o n s h i p between environmental and l i f e s t y l e r i s k exposure during the preconception period and the o v e r a l l occurrence of anomalies ( i . e . , associated with spontaneous abortion, s t i l l b i r t h , and l i v e b i r t h ) i n the population (International Commission for Protection against Environmental Mutagens and Carcinogens [ICPEMC], 1983; Obe, 1984; Wynn & Wynn, 1984). Questions of i n t e r e s t to th i s study r e l a t e to the t o t a l incidence of b i r t h defects i n the community, the proportion associated with l i v e b i r t h s and the timing of the i n s u l t ; whether there i s a d i f f e r e n c e between the population who abort or do not abort abnormal karyotypes; whether the reproductive h i s t o r y of women are d i f f e r e n t i f t h e i r pregnancies are/are not associated with b i r t h defects, whether a v a i l a b l e data are d e s c r i p t i v e enough to discriminate between the influence of environmental/lifestyle element versus the bi o l o g i c / g e n e t i c element on abnormal development and i f so, whether there i s an ass o c i a t i o n between abnormal f e t a l development and the l i f e s t y l e of women or men. B r i t i s h Columbia has established a s u r v e i l l a n c e system i n the province which, through a multiple source reporting procedure and an extended ascertainment period, enables a reasonably comprehensive coverage of the occurrence of b i r t h defects i n the community. For the period 1966-1980, the average annual rate among t o t a l b i r t h s for both 75 sexes combined was 47.3 cases per 1,000 (4.73 percent of t o t a l b i r t h s ) ; f o r male b i r t h s was 53.1 cases per 1,000 (5.3 percent); and for females was 41.2 per 1,000 (4.1 percent). B i r t h defects are defined by the ICD-9 codes 740-759. The average rates among l i v e b i r t h s are s l i g h t l y lower than among t o t a l b i r t h s (52.5 per 1,000 male l i v e b i r t h s and 40.1 per 1,000 female l i v e b i r t h s ) due to the high rate (125.4 per 1,000) of congenital anomalies reported i n s t i l l b i r t h s (Province of B.C., 1981). The s u r v e i l l a n c e system has developed the p o t e n t i a l for assessing factors which could contribute to the occurrence of b i r t h defects, but lacks comparative data from a reference population and standardized assessment measures for many of the variables (e.g., l i f e s t y l e f a c t o r s ) . Tables 7 and 8 describe the incidence of congenital anomalies i n BC for the period 1971-1980 as reported by the D i v i s i o n of V i t a l S t a t i s t i c s and the Health Surveillance Registry. 76 TABLE 7 Tota l Number of L i v e b i r t h s and To t a l Number of Congenital Anomalies (ICD 740-759) and Rate per 1,000 L i v e b i r t h s by Major Diagnostic Categories and for Male and Female Bir t h s i n BC, 1971-1980, as Reported by Year End 1981 Major Diagnostic Categories of Reported Congenital Anomalies Congenital Male Female T o t a l Anomalies N N N R 1. Nervous System 517 461 978 26 .9± 2 .2 2. Eye 253 246 499 13 .7±1.5 3. Ear, Face & Neck 382 340 722 19 .9±2.2 4. Heart & C i r c u l a t o r y System 2,280 2,021 4,301 118 • 3±7.3 5. Respiratory System 194 157 351 9 .7±3.3 6. C l e f t Palate & C l e f t L ip 406 299 705 19 .4±2.2 7. Upper Alimentary Tract & Digestive System 1,436 664 2,100 57 • 8±3.8 8. Genital Organs 2,476 167 2,643 72 .7±6.0 9. Urinary System 506 418 924 25 • 4±6.4 10. Musculo-Skeletal System 5,235 3,794 7,029 193 .3±14.3 11. Integument 404 418 822 22 ,6±6.7 12. Chromosomal Anomalies 254 251 505 13 , 9 ± 1 . 4 13. Other & Unspecified Congenital Anomalies 152 139 291 8 .0±1.3 T o t a l Congenital Anomalies i n L i v e b i r t h s 12,495 9,376 21,871 601 .5±20.1 Number of L i v e b i r t h s i n B r i t i s h Columbia 186,703 176,930 363,633 Source: Province of B r i t i s h Columbia, D i v i s i o n of V i t a l S t a t i s t i c s , Health Surveillance Registry Annual Report, 1981. 77 TABLE 8 Major Diagnostic Categories of Congenital Anomalies (ICD 740-759) by Incidence Rate per 10,000 L i v e b i r t h s and Proportion of a l l Congenital Anomalies Reported i n BC L i v e b i r t h s for the period 1971-1980 Reported Congenital Anomalies Major Categories i n B r i t i s h Columbia of Congenital Anomalies ICD-9 Rate/10,000 % T o t a l Code L i v e b i r t h s Anomalies 1. Nervous System 740-742 26.9 4.5 2. Eye 743 13.7 2.3 3. Ear, Face & Neck 744 19.9 3.3 4. Heart & C i r c u l a t o r y System 745-747 118.3 19.7 5. Respiratory System 748 9.7 1.6 6. C l e f t Palate & C l e f t L ip 749 19.4 3.2 7. Upper Alimentary Tract & Digestive System 750-751 57.8 9.6 8. Genital Organs 752 72.7 12.1 9. Urinary System 753 25.4 4.2 10. Musculo-Skeletal System 754-756 193.3 32.1 11. Integument 757 22.6 3.8 12. Chromosomal Anomalies 758 13.9 2.3 13. Other & Unspecified Congenital Anomalies 759 8.0 1.3 Source: Province of B r i t i s h Columbia, D i v i s i o n of V i t a l S t a t i s t i c s , Health Surveillance Registry Annual Report, 1981. 78 Infant Morbidity (Longterm Growth and Development) While infant m o r t a l i t y w i l l always remain of major concern because of the human s u f f e r i n g involved and the loss of p r o d u c t i v i t y to the community, of equal i f not greater importance i s the impact on the i n d i v i d u a l , family, and community of s u r v i v a l accompanied by d i s a b i l i t y — t h e more severe the d i s a b i l i t y , the higher the cost i n human and economic terms. For the most part, morbidity i s now concerned with disorders of growth and development, such as those associated with malformations, i n t r a u t e r i n e growth retardation, short gestation and neoplasms, rather than the acute i n f e c t i o n s of year's past. Many more infants now survive major i n s u l t s of p e r i n a t a l o r i g i n and require extra support from medical and community services for continued s u r v i v a l and/or to minimize t h e i r handicaps. Continuing attention must be directed toward whether reductions i n p e r i n a t a l mortality are accompanied by s i m i l a r reductions i n morbidity, or since the increased s u r v i v a l rate includes c h i l d r e n with b i r t h defects (e.g., Down's Syndrome) and, thus, increases the prevalence of these disorders i n the community, whether such increases are o f f s e t by the reduction of other preventable disorders. To assess the magnitude of these problems and to monitor any change i n t h e i r incidence or prevalence i n the community requires good morbidity data. Vohr and Hack (1982), i n a review of developmental follow-up of low birthweight i n f a n t s , state that very low birthweight (<1,500 g) continues to be an important factor i n development of subsequent handicap i n the 79 1980s, with r i s k of sequelae i n v e r s e l y r e l a t e d to birthweight. Short term r e s u l t s are often more o p t i m i s t i c than followup of longer duration which report an increased incidence of i n t e l l e c t u a l handicap and learning d i f f i c u l t i e s even i n infants without major sequelae. The paper outlines the p o t e n t i a l r i s k factors to be considered and the type of followup schedules and evaluation procedures that are needed to monitor q u a l i t y of outcome beyond s u r v i v a l . A systematic infant s u r v e i l l a n c e system has not been established i n B r i t i s h Columbia. The Infant Monitoring Program of the Calgary Health Services (1984) provides an excellent working example of such a system. I t i s even more d i f f i c u l t to determine and in t e r p r e t morbidity data than many other reproductive health outcomes. For example, the d i s t i n c t i o n between health and i l l n e s s i s much less objective than between l i f e and death or birthweight categories; diagnostic standards vary much more for non-fatal compared to f a t a l i l l n e s s e s since pathological changes are les s extreme i n the former and les s e a s i l y confirmed; and morbidty i s more d i f f i c u l t to summarize since i t has d i f f e r e n t lengths of duration, can occur more than once to the same i n d i v i d u a l and can take many forms (Leek, 1974). One method of obtaining morbidity data i s through the use of health s e r v i c e contacts such as h o s p i t a l discharge by cause reports. These data measure the frequency of an event in^the population but because the same i n d i v i d u a l may be admitted more than once during the study period, they may require a supplementary survey to determine the frequency and duration of h o s p i t a l i z a t i o n , and type of health problem 80 experienced by a p a r t i c u l a r group of i n d i v i d u a l s . Prevalence or bed occupancy rates are more us e f u l than incidence or discharge rates i n determining the burden of the i l l n e s s e s on the health care services. The i l l n e s s incidence, however, i s the more important i n studying causal associations, or i n assessing the e f f e c t s of intervention. H o s p i t a l data are always biased i n favor of what i s currently a v a i l a b l e i n terms of type and numbers of beds. Given the appropriate reporting system, a s i m i l a r method for determining frequency of physician contact (.other than f o r routine checkup), of need for p r e s c r i p t i o n drugs, and/or the use of s p e c i a l community services ( i . e . , i n fant development program) might be used to measure morbidity by determining the s i z e of the " a f f e c t e d " group. A second method of obtaining morbidity data i s by screening whole populations through routine or s p e c i a l surveys. Routine screening i s c a r r i e d out i n most communities at b i r t h , at the beginning and end of school l i f e and often screening of i n t e l l e c t u a l a b i l i t y i s c a r r i e d out some time during school years—some countries have i n s t i t u t e d routine preschool screening. Data from routine screening are usually l i m i t e d to q u a n t i t a t i v e variables such as weight, height, i n t e l l i g e n c e quotient, and the determination of l a s t i n g d i s a b i l i t i e s such as malformations and impaired hearing or v i s i o n . The main advantage of screening data i s that knowledge i s gained about the e n t i r e population and the proportion and d i s t r i b u t i o n of disorders within i t . The reason for routine screening i s generally to detect unrecognized health problems at an opportune time for preventive treatment. Screening, as w e l l as 81 health service contacts, can be biased by misdiagnoses, by inadequate reporting and by some i n d i v i d u a l s being excluded. Special surveys of p a r t i c u l a r types of morbidity are l i k e l y to be either c r o s s - s e c t i o n a l (retrospective or case-history) or l o n g i t u d i n a l (prospective or cohort) studies. Problems associated with these surveys include inadequate reporting by parents of the study c h i l d r e n and inadequate followup of the survey population i f f a m i l i e s move or are unwilling to p a r t i c i p a t e . With p a r t i c u l a r reference to morbidity of c h i l d r e n of low birthweight or preterm b i r t h , and the measurement of the frequency of rela t e d problems i n l a t e r years, Leek (1974) i d e n t i f i e s three areas of bias: (a) sources of inaccuracy of the measurement recorded ( i . e . , s i n g l e measures, rounding-off); (b) the heterogeneity of the conditions which an abnormal measurement may r e f l e c t ; and (c) the a r t i f i c i a l i t y of c l a s s i f y i n g i n d i v i d u a l s i n t o normal and abnormal on the basis of a fix e d threshold. Murphy, Nichter, and Liden (1982) review a number of d i f f e r e n t methodological issues i n infant development outcome studies. Factors which make t h i s type of research d i f f i c u l t include: l i m i t e d knowledge of "normal" human development; the lack of d i s t i n c t i o n between delayed development, s p e c i f i c developmental disorders and fun c t i o n a l d e f i c i t s ; the problem i n separating transient from longterm outcomes; and the changing nature of the hi g h - r i s k population over the past 20 years. The f i n a l set of problems involve q u a l i t y of assessment i n terms of 82 the s t a b i l i t y of developmental examinations, the comprehensiveness of assessment b a t t e r i e s , and s e r i a l observations of the child r e n . With the many inherent problems of l o n g i t u d i n a l and developmental studies, i t i s not s u r p r i s i n g that very few s p e c i a l surveys are c a r r i e d out and even fewer provide s o l i d evidence of longterm p e r i n a t a l e f f e c t s except at the extremes. An exceptional study chosen f o r review i n t h i s section was reported by Neligan and colleagues (1976) and i s important because the r e s u l t s can be generalized to other populations and because the infants studied were of moderate rather than high r i s k . The r e s u l t s are, therefore, applicable to the large number of infants who are low weight or preterm. Neligan and colleagues o u t l i n e the s p e c i a l features which characterize the study: 1. The population studied was selected s p e c i f i c a l l y for the purpose of comparing a group of ch i l d r e n who had been born too soon and a group who had been born too small, both with each other and with a s u i t a b l e comparison group, i n terms of many aspects of t h e i r l a t e r performance during school years. The groups represented the two abnormalities of in t r a u t e r i n e growth which account for low birthweight and which present c l i n i c i a n s with d i f f e r e n t p r a c t i c a l problems. 2. The two abnormal groups and the comparison group were selected from a l a r g e r , geographically defined population f or whom standardized o b s t e t r i c , p a e d i a t r i c , and s o c i a l data were a v a i l a b l e through the Newcastle Survey of Child Development (Neligan, Prudham, & Steiner, 1974). The s e l e c t i o n was, therefore, free of the biases due to medical, 83 s o c i a l or personal factors which so often prevent gener a l i z a t i o n of study r e s u l t s . 3. The methods of assessment were s u f f i c i e n t l y s o phisticated, comprehensive, and s p e c i f i c to detect very minor differences between i n d i v i d u a l c h i l d r e n over a very wide range of measures of performance, and care was taken so that r e s u l t s could be quantified i n terms of continuous v a r i a b l e s . 4. The methods of s t a t i s t i c a l analysis enabled i d e n t i f i c a t i o n of the s p e c i f i c and independent e f f e c t s of the two abnormalities of i n t r a u t e r i n e growth, a f t e r allowing for a large number of associated and p o t e n t i a l l y confounding f a c t o r s , and to compare the magnitude of the e f f e c t s of these f a c t o r s . The study population consisted of a randomly selected comparison group of 187 f i v e year old c h i l d r e n from amongst one year of b i r t h s of the Newcastle Survey group; 141 c h i l d r e n whose birthweight was below the 10th p e r c e n t i l e for g e s t a t i o n a l age which were separated into two groups—birthweight between the 5th and 10th percentiles and birthweight on or below the 5th p e r c e n t i l e (n = 74); and 59 c h i l d r e n who were born at less than 37 weeks gestation. The groups were studied at 5, 6, and 7 years of age using a wide range of s p e c i f i c psychometric, behavioral and temperamental, neurological and p h y s i c a l growth assessments against a known background of b i o l o g i c a l and s o c i a l v a r i a b l e s . After c a r e f u l analysis of the r e s u l t s , the authors came to the following conclusions: 1. Both the short-gestation and the very l i g h t - f o r - d a t e s groups 84 of c h i l d r e n performed s i g n i f i c a n t l y l e s s w e l l than those i n the random sample over the whole range of measures of performance at ages 5, 6, and 7 years. This conclusion was not altered by the e f f e c t s of 15 associated f a c t o r s . 2. The suggestion that prenatal impairment of growth may be d i r e c t l y responsible for the l a t e r impairment of performance i s strongly supported by the f i n d i n g s — t h e scores from the rather l i g h t - f o r - d a t e s group (mean birthweight of 2,701 g) were almost always intermediate between those of the very l i g h t - f o r - d a t e s (mean birthweight 2,397 g) and the random controls for both o v e r - a l l scores and i n d i v i d u a l subtests. Birthweight made an important contribution to the o v e r - a l l performance scores of the ch i l d r e n w i t h i n a l l groups in c l u d i n g the random sample and short-gestation groups. 3. The o v e r - a l l performance of the c h i l d r e n i n the very l i g h t -for-dates group i s s i g n i f i c a n t l y worse than the short-gestation group. In the study population, impairment of the net rate of i n t r a u t e r i n e growth had a d i r e c t adverse e f f e c t upon the l a t e r performance of the ch i l d r e n which was greater than the e f f e c t of a comparable degree of shortening of i t s duration. 4. The e f f e c t s of some of the associated factors (e.g., s o c i a l c l a s s , mother's care, mother's expectations) which were investigated appear to be of much greater magnitude than those of the v a r i a b l e s i n i n t r a u t e r i n e growth and t h e i r r e l a t i v e importance tends to d i f f e r i n the d i f f e r e n t groups. 85 5. The impaired performance of the c h i l d r e n i n the two extreme abnormal groups p e r s i s t s with very l i t t l e modification a f t e r the e f f e c t s of s i x b i o l o g i c a l , c l i n i c a l , and environmental associated factors have been allowed f o r . 6. The associated factors grouped under the heading of "family f a c t o r s " ( i n p a r t i c u l a r , s o c i a l c l a s s and mother's care of child) were of overriding importance i n the random sample and very l i g h t - f o r - d a t e s group: b i o l o g i c a l and c l i n i c a l factors combine to produce e f f e c t s of almost equal importance i n the short gestation group. Due to the c a r e f u l design and implementation of t h i s study, Neligan and colleagues have provided sound evidence that both c h i l d r e n who are born too soon and those who are born too small show some impairment of performance when they reach school age—those born too small are at a greater disadvantage and seem les s l i k e l y to benefit from recent advances i n p a e d i a t r i c care, whereas those born too soon seem able to b e n e f i t from adequate n u t r i t i o n and care i n the postnatal period. The adverse e f f e c t s a t t r i b u t a b l e to the two disorders of i n t r a u t e r i n e growth (representing 10 percent of the t o t a l study population) cover a wide range of measures of performance and are of a s i m i l a r pattern i n both groups. These groups seem p a r t i c u l a r l y vulnerable to other b i o l o g i c a l , c l i n i c a l , and s o c i a l factors which may also have adverse e f f e c t s . This study re i n f o r c e s the need for an index of reproductive health that w i l l i d e n t i f y the proportion of infants at r i s k for longterm developmental impairment; that w i l l provide a t o o l to measure the 86 e f f e c t of factors on f e t a l growth and ultimately help to determine intervention s t r a t e g i e s which can reduce the preventable portion of f e t a l growth complications. Summary In summary, th i s chapter has reviewed the key outcome measures of reproductive and infant health; the problems of accurate measurement; the outcome data a v a i l a b l e f o r B r i t i s h Columbia; and the type of data required f o r the development of an o v e r a l l index of reproductive health which would r e f l e c t the as s o c i a t i o n between the d i f f e r e n t outcomes i n the community and provide support for s u r v e i l l a n c e and intervention s t r a t e g i e s . This information w i l l be further summarized i n chapters four and f i v e . 87 CHAPTER I I I : THE EFFECT OF LIFESTYLE ON REPRODUCTIVE HEALTH Chapter two began with the p r o p o s i t i o n that the extent to which a population encounters problems producing healthy, l i v i n g infants i s influenced to varying degrees by each of the four elements of the health f i e l d concept—human biology, environment, health care organization, and l i f e s t y l e . The chapter then reviewed the outcome measures which are used to document the reproductive c a s u a l t i e s occurring i n a population, the problems associated with t h i s documentation, and whether the a v a i l a b l e information was comprehensive enough to allow for the assessment of the impact of the l i f e s t y l e element on these outcomes. This chapter examines the extent to which these outcome measures might be affected by l i f e s t y l e . The purpose of the chapter i s to review the evidence that l i f e s t y l e factors do a f f e c t reproductive outcome and to examine what proportion and to what degree might the population be affected. Where data are a v a i l a b l e , the prevalence of these l i f e s t y l e factors i n the BC population of childbearing years w i l l be reported. The focus of the review w i l l be to determine what type of an e f f e c t , i f any, a l i f e s t y l e f a ctor has on reproductive outcome—i.e., whether there i s a growth retarding influence and i f t h i s i s associated with increased mortality and/or morbidity; whether there i s an associated teratogenic, mutagenic, abortifacent, or f e t o t o x i c influence; 88 whether the e f f e c t v aries with early or l a t e pregnancy exposure; and whether the impact can be measured across a l l the reproductive outcome measures or only a s e l e c t number. In the next chapter, the issue of "cause and e f f e c t " w i l l be examined i n d e t a i l . Only the l i t e r a t u r e concerning the e f f e c t of smoking (tobacco and cannabis), n u t r i t i o n and d i e t , and alcohol on reproductive health w i l l be reported i n t h i s study. An equally s u b s t a n t i a l set of l i t e r a t u r e on the r e l a t i o n s h i p between reproductive health outcomes and exposure to drugs, chemicals and other toxins, and to p h y s i c a l and psychological stress could have been reviewed. Given l i m i t a t i o n s of space, i t was decided that a comprehensive assessment of a representative group of l i f e s t y l e factors was preferable to a less comprehensive review of a l l . For the same reason, a representative set of studies for each of the three l i f e s t y l e factors and seven reproductive outcome measures were selected for d i s c u s s i o n — t h e s e h i g h l i g h t human population studies with reference to associated animal, metabolic, and c e l l culture studies. Smoking and Reproductive Health I t i s about 50 years since c l i n i c i a n s f i r s t warned of the detrimental e f f e c t s of smoking on reproductive health. Human and animal research linked maternal exposure to tobacco with increased s t i l l b i r t h and decreased birthweight (Bernard, 1948; Essenberg, Schwind, & Patras, 1940; Schoeneck, 1941; Sontag & Wallace, 1935). Over the past 40 years, an extraordinary amount of research has been conducted i n the area. Comprehensive l i t e r a t u r e reviews have 89 been published (Abel, 1980; Landesman-Dwyer & Emanuel, 1979; S i d l e , 1982; US Publ i c Health Service, 1972, 1980; Wynn & Wynn, 1981) which provide structure to the assessment of the strength of asso c i a t i o n between smoking and reproductive health, mechanisms of action, and int e r a c t i o n s with other l i f e s t y l e and demographic r i s k factor v a r i a b l e s . One of the most recent and comprehensive reports on the health consequences of smoking during.the reproductive period, based on a c r i t i c a l review of the l i t e r a t u r e by basic s c i e n t i s t s and c l i n i c i a n s , was published as a report of the US Surgeon General i n 1979 (US Publ i c Health Service, 1980). This report summarized the evidence of the adverse e f f e c t of smoking on reproductive outcome measures i n the following way: A. The e f f e c t on f e t a l growth. Babies born to women who smoke during pregnancy are, on the average, 200 g l i g h t e r than babies born to comparable nonsmoking women. There i s a dose-response r e l a t i o n s h i p between maternal smoking and reduced birthweight: the more the woman smokes during pregnancy, the greater the reduction i n birthweight. I f a woman gives up smoking early during pregnancy, her r i s k of d e l i v e r i n g a low birthweight baby approaches that of a nonsmoker. The pattern of f e t a l growth re t a r d a t i o n that occurs with maternal smoking i s a decrease i n a l l parameters in c l u d i n g body length, chest circumference, and head circumference. The r e l a t i o n s h i p between maternal smoking and reduced birthweight i s independent of a l l other factors that influence birthweight including 90 race, p a r i t y , maternal s i z e , socioeconomic status, and sex of c h i l d ; i t i s also independent of ge s t a t i o n a l age. Maternal smoking during pregnancy exerts a d i r e c t growth-retarding e f f e c t on the fetus which does not appear to be mediated by reduced maternal appetite, eating, or weight gain. The r a t i o of pl a c e n t a l weight to birthweight increases with increasing l e v e l s of maternal smoking, r e f l e c t i n g a considerable decrease i n mean birthweight and a s l i g h t increase i n mean pl a c e n t a l mass; t h i s may represent an adaptation to r e l a t i v e f e t a l hypoxia. B. The e f f e c t on f e t a l / i n f a n t m o r t a l i t y and morbidity. The r i s k of spontaneous abortion, f e t a l death, and neonatal death increases d i r e c t l y with increasing l e v e l s of maternal smoking during pregnancy; i n t e r a c t i o n of maternal smoking with other factors which increase p e r i n a t a l m o rtality may r e s u l t i n an even greater r i s k . Excess deaths of smokers' infants are found mainly i n the coded cause categories of "unknown" and "anoxia" f o r f e t a l deaths, and the categories of "prematurity alone" (defined by birthweight) and "r e s p i r a t o r y d i f f i c u l t y " for neonatal deaths; suggesting that the excess deaths are due to problems of the pregnancy, rather than to abnormalities of the fetus or neonate. Although there i s l i t t l e e f f e c t of maternal smoking on mean length of gestation, the proportion of f e t a l deaths and l i v e b i r t h s that occur before term increases d i r e c t l y with maternal smoking l e v e l . Up to 14 percent of a l l preterm d e l i v e r i e s i n the US may be a t t r i b u t a b l e to ma t er nal smoking. 91 Increasing l e v e l s of maternal smoking r e s u l t i n a highly s i g n i f i c a n t increase i n the r i s k of abruptio placentae, placenta previa, bleeding early or l a t e i n pregnancy, premature and prolonged rupture of membrances, and preterm d e l i v e r y — a l l of which carry high r i s k s of p e r i n t a l l o s s . Maternal smoking during pregnancy may adversely a f f e c t the c h i l d ' s long-term growth, i n t e l l e c t u a l development, and behavioral c h a r a c t e r i s t i c s . An i n f a n t ' s r i s k of developing the "sudden infant death syndrome" i s increased by maternal smoking during pregnancy. Infants and c h i l d r e n born to smoking mothers may experience more long-term morbidity than those born to nonsmoking mothers; however, studies usually cannot d i s t i n g u i s h between the e f f e c t s of smoking during pregnancy and the e f f e c t s of the infant's or c h i l d ' s passive exposure to c i g a r e t t e smoke af t e r b i r t h . C. E f f e c t on b i r t h defects. There are i n s u f f i c i e n t data to determine whether maternal and/or paternal c i g a r e t t e smoking increases the r i s k of congenital malformations. D. E f f e c t on other reproductive outcome measures. Studies i n women and men suggest that c i g a r e t t e smoking may impair f e r t i l i t y . While i t may be f a i r to assume that t h i s i s an appropriate representation of the evidence to 1979, t h i s report, along with most other reviews of the l i t e r a t u r e i n t h i s f i e l d , does not provide a review of the design, implementation, and s t a t i s t i c a l analysis of the referenced studies and, thereby, does not allow the reader to weigh the evidence provided. Since evidence from large, prospective cohort 92 studies with c a r e f u l l y documented l e v e l s of exposure would provide the most convincing r e s u l t s , a selected review of these studies i s reported. Smoking i n pregnancy has a dual e f f e c t on the mother and f e t u s — i t i s d i f f i c u l t to evaluate the e f f e c t of smoking on the fetus, and equally as d i f f i c u l t to evaluate the e f f e c t of smoking on the mother and her reproductive c a p a b i l i t y . Butler (1975),in an address to the Third World Conference on Smoking and Health, summarized these d i f f i c u l t i e s w e l l . He stated: F i r s t , i n the majority of studies, smoking habits i n pregnancy were inquired about a f t e r d e l i v e r y and were, therefore, subject to f a l a c i e s of memory. Second, smoking i s not a constant phenomenon, as women may begin to smoke or give up smoking during pregnancy, and there may be frequent changes i n the number of c i g a r e t t e s smoked. Third, there are differences i n degree of i n h a l a t i o n and i n type of tobacco smoked. Fourth and perhaps most important, d i f f e r e n t types of women have d i f f e r e n t smoking habits. Smoking habits w i l l d i f f e r with s o c i a l background, p a r i t y and even with the height of the mother. Some o b s t e t r i c conditions may occur more often i n women who smoke than i n women who don't smoke. The changing components of smoking and also the e f f e c t of other s o c i o b i o l o g i c a l influences may completely obscure the pattern and e f f e c t s . Special techniques of analysis and s t a t i s t i c a l treatment can be used to exclude the e f f e c t s of other v a r i a b l e s which could a f f e c t birthweight, p e r i n a t a l m o r t a l i t y or even c h i l d development, where these data are a v a i l a b l e . I t has also been d i f f i c u l t to c o r r e l a t e the c l i n i c a l s i t u a t i o n with experimental fi n d i n g s . A great deal of work has been done with animals . . . but i t must be remembered that the experimental animal i s not the same as a pregnant human female, (pp. 43-44) F e r t i l i t y Many studies suggest that smoking exerts an adverse e f f e c t on f e r t i l i t y for both women and men (US P u b l i c Health Service, 1980). There are many ways that f e r t i l i t y can be affected and, i n the case of smoking, the following associations have been reported: a reduction 93 i n years of f e r t i l i t y for women through early onset of menopause; a reduction i n l e v e l of f e r t i l i t y through impaired spermatogenesis i n men and increased prevalence of menstrual disorder i n women, and a delayed return of f e r t i l i t y a f t e r discontinuation of contraception. Two studies are selected to examine the asso c i a t i o n between smoking and early onset of menopause. W i l l e t t and colleagues (1983) prospectively evaluated the experience of 66,663 female registered US nurses who were pre-menopausal i n 1976 and the 5,004 women who became post-menopausal over the following two years. Data were c o l l e c t e d by mail questionnaire. Of the 121,964 o r i g i n a l study p a r t i c i p a n t s , 78,678 were pre-menopausal and received a questionnaire two years l a t e r . Of th i s group, 89 percent (69,906) responded, 2,887 were post-menopausal f o r s u r g i c a l or therapeutic reasons, 356 did not report menopausal status, leaving a sample s i z e of 66,663 women. Analysis was ca r r i e d out according to four age groups, three smoking c l a s s i f i c a t i o n s and four l e v e l s of smoking, two "weight f o r height" c l a s s i f i c a t i o n s and f i v e q u i n t i l e s of r e l a t i v e weight. Within sample followup suggested accurate reporting of menopausal and r e l a t i v e weight status. Thirty-one percent of the sample were ..current smokers which i s s i m i l a r to the estimate of 32 percent as the prevalence of regular c i g a r e t t e smoking among US women i n 1976 (US, Department of Health Education & Welfare, 1977). Independent e f f e c t s of current smoking, number of cigaret t e s smoked d a i l y , r e l a t i v e weight and n u l l i p a r i t y were confirmed by multiple 94 log i s t i c regression analysis. The rate ratios of menopause for current smokers versus never smokers (95 percent confidence limits) 1.90 (1.10-3.28) for women between 30-39 years, 2.16 (1.73-2.69) for 40-44 year olds, 1.53 (.1.41-1.67) for those aged 45-49 and 1.20 (1.12-1.28) for women aged 50-55 years. Median ages at menopause by level of smoking were 52.4 for never smokers, 51.9 for those smoking 1-14 cigarettes per day, 51.0 for 15-24 cigarettes/day smokers, 50.7 for those smoking 25-34 per day, and 50.4 for women smoking 35 or more cigarettes per day. Any significant relationship between earlier menopause and ex-smokers was confined to those who had quit within the past two years. After adjustment for current cigarette exposure, a weak linear relationship between relative weight and menopause remained among women who smoked but not among nonsmokers. Extremes of underweight were not seen in the study population. The potential effect of a non-response bias was not discussed in this paper but careful consideration was given to potential confounding by other variables. The relation of cigarette smoking to natural menopause was also evaluated by Jick, Porter, and Morrison (1977) in a retrospective cross-sectional survey of two independent groups. Women between the ages of 44-53 years who were hospitalized for medical or surgical reasons were interviewed—one group of 3,076 women being derived from a survey of 25,000 hospital patients from 24 Boston-area hospitals and the other group of 2,479 women from 32,000 inpatients in seven countries participating in the Boston Collaborative Drug Surveillance Program. It is not clear that a l l eligible women aged 44-55 years 95 were interviewed. A f t e r exclusion of those with a h i s t o r y of s u r g i c a l menopause, analysis included data from 2,143 women for group I and from 1,391 women for group I I — n o discussion of the differences between the study women and those excluded for lack of information was presented. Three smoking c l a s s i f i c a t i o n s , two smoking l e v e l s , and f i v e age group c l a s s i f i c a t i o n s were used for analysis. Data were examined for confounding by v a r i a b l e s other than age and smoking (e.g., medical versus s u r g i c a l admission, diagnosis, p a r i t y , alcohol intake). Results showed a progressive r i s e i n the proportion of women who were post-menopausal i n a l l age categories except the youngest (44-45 years) according to amount smoked. The ass o c i a t i o n between smoking and natural menopause was consistent i n each of the seven countries contributing to study I I . For the group of women from Boston, the age-standardized proportions who were post-menopausal by smoking c l a s s i f i c a t i o n were 35 percent, 36 percent, 43 percent, and 49 percent, r e s p e c t i v e l y , for never smokers, ex-smokers (>1 year), and current smokers of one-half pack cigare t t e s per day or one pack or more per day. Corresponding proportions for the group II women were considerably higher at 53 percent, 59 percent, 60 percent, and 65 percent. The consistency of a di f f e r e n c e between the menopause experience of nonsmokers and current smokers seen i n these two studies and others studying d i f f e r e n t population samples (Bailey, Robinson, & Vessey, 1977; Lindquist & Bengtsson, 1979), and the demonstration of a dose-response r e l a t i o n s h i p provides good reason to assume the assoc i a t i o n 96 between smoking and early menopause i s v a l i d . These r e s u l t s are pertinent to a. number of p u b l i c health.issues, but, for t h i s study, they serve as an example of the e f f e c t of smoking on the female reproductive system per se (as compared to an e f f e c t on f e t a l growth) and provide reason to speculate that smoking may be a contributing factor to a range of reproductive problems. The r e t r o s p e c t i v e case-control study of Olsen and colleagues (1983) provides another example. In t h i s study, the a s s o c i a t i o n between tobacco use, alcohol consumption, and subfecundity was investigated i n couples who were examined or treated for an i n f e r t i l i t y problem at Odense Uni v e r s i t y H o s p i t a l and couples who had a healthy c h i l d born at the h o s p i t a l w i t h i n the same time period (1977-1980). Since h o s p i t a l p o l i c y required h o s p i t a l admittance for i n f e r t i l i t y work-up, a l l e l i g i b l e couples ( i n f e r t i l e , f e r t i l e with healthy c h i l d , no i n d i c a t i o n of chronic disease) could be i d e n t i f i e d from the h o s p i t a l r e g i s t r y . Data were c o l l e c t e d by self-administered mailed questionnaires and, f o r a l l i n f e r t i l e cases, enhanced by information from h o s p i t a l records. A response rate of 87 percent for each group was reported providing 927 case and 3,728 con t r o l couples. Analysis involved two c l a s s i f i c a t i o n s of smoking (yes/no),.and four categories of alcohol consumption (by grams of pure alcohol) during the time of h o s p i t a l admission. Three d i f f e r e n t a n a l y t i c a l approaches were used as a way to address methodological problems—between group comparison of i n f e r t i l e and f e r t i l e couples, within-group comparison of i n f e r t i l e couples, and within-group comparison of f e r t i l e couples. Odds r a t i o s 97 for estimating the r e l a t i v e r i s k of smoking and alcohol consumption are adjusted for the e f f e c t s of p a r i t y , age, residence, education, and o r a l contraceptive use. A p o s i t i v e a s s o c i a t i o n between smoking and primary and secondary subfecundity (OR:1.6, C I : l . l - 2 . 2 ; 0R:2.1, CI:1.3-3.6, respectively) i n the between group comparison of cases and controls was supported by a p o s i t i v e a s s o c i a t i o n between smoking and delayed conception of at l e a s t one year wi t h i n the control group—delayed conception i n a past pregnancy (0R:1.6, 1.3-2.0), i n present pregnancy (0R:1.5, 1.2-3.0), and i n both past and present pregnancies (OR:2.0, 1.4-2.8). The odds r a t i o for a l l cases compared to controls for smoking (OR:2.1, CI:1.7-2.5) and alcohol consumption (0R:1.6, CI:1.2-2.1) was s i m i l a r to the odds r a t i o for the subgroup of cases excluding couples with a diagnosed male f e r t i l i t y problem,compared to c o n t r o l s — f o r smoking t h i s was 2.3 (1.8-2.9) and for alcohol consumption, 1.7 (1.3-2.4). Methodological problems were reviewed i n the study: s e l e c t i o n bias which might a f f e c t case-control comparison was u n l i k e l y to a f f e c t w i t h i n - c o n t r o l analysis; non response bias and r e p o r t i n g / s e l e c t i v e memory bias would be less l i k e l y to a f f e c t within group analysis. The reduction i n use of tobacco and alcohol during pregnancy i s a very r e a l p o s s i b i l i t y i n the pregnant c o n t r o l group, but does .not a f f e c t the s i g n i f i c a n t p o s i t i v e a s s o c i a t i o n between smoking and delayed conception wi t h i n the c o n t r o l group. In general, r e s u l t s which are supported by both between and within group analysis provide reasonable evidence of r e a l association. The evidence for an a s s o c i a t i o n between 98 smoking and i n f e r t i l i t y / c o n c e p t i o n delay was consistent and suggests a r e a l a s s o c i a t i o n whereas the a s s o c i a t i o n between alcohol consumption and i n f e r t i l i t y i s not f u l l y supported. Community health surveys ( c r o s s - s e c t i o n a l r e t r o s p e c t i v e studies) i n the US (Sloss & F r e r i c h s , 1983), A u s t r a l i a (Wood, 1978), and Finland (Kaureniemi, 1969) show an a s s o c i a t i o n between smoking and the prevalence of menstrual disorders for which a woman sought medical at t e n t i o n (e.g., dysmenorrhoea and menstrual i r r e g u l a r i t y ) . These studies have not ruled out a spurious a s s o c i a t i o n with low body weight or low percent body f a t and face the common problems of response and memory bias i n r e t r o s p e c t i v e studies. Nevertheless, menstrual disorders are of s u f f i c i e n t s i g n i f i c a n c e i n the population (Kistner, 1979) that a preventable as s o c i a t i o n i s worth examining further. Smoking appears to have an a f f e c t on the reproductive system of men as w e l l as women, an a s s o c i a t i o n supported by animal and b i o l o g i c a l studies (US Public Health Service, 1980). Given the range of studies reporting an adverse ass o c i a t i o n between smoking and reproductive health, a l b e i t of varying weights i n terms of strength of evidence, i t i s d i f f i c u l t not to speculate that the a s s o c i a t i o n i s r e a l and pervasive. The extent of many of these p o t e n t i a l problems cannot yet be estimated. Spontaneous Abortion Problems of ascertainment i n measuring the incidence of spontaneous abortions have been discussed i n chapter two. Retrospective studies are more l i k e l y to allow for complete ascertainment but are subject to 99 errors of r e c a l l , while prospective studies may introduce bias i f groups do not enter the study at comparative times since the early abortions may be missed. While there are a number of studies of both types which report an excess of spontaneous abortions associated with smoking (Downing & Chapman, 1966; Himmelberger, Brown, & Cohen, 1978; O'Lane, 1963; Underwood et a l . , 1965; Zabriskie, 1963), few provide a basis f o r sound estimates of r i s k . A prospective study on the e f f e c t of smoking on pregnancy (Kullander & Kall e n , 1971) obtained data on smoking habits throughout 6,363 pregnancies v i a self-administered questionnaires completed at each subsequent prenatal v i s i t to the physician from diagnosis of pregnancy. Forty-four percent of the women (2,806) smoked during t h e i r pregnancy—almost a l l of these smoked during the enti r e pregnancy (97 percent), only 20 percent smoked more than 10 cigaret t e s per day and 1 percent more than 20 per day. The women whose pregnancies ended i n spontaneous abortions (<28 weeks gestation) were more l i k e l y to be smokers (51.0 percent) than those who delivered a l i v e healthy c h i l d (43.3 percent) (X? = 12.5, p<0.001). The observation of the e f f e c t of smoking on spontaneous abortion showed no differe n c e i n smoking habits of women aborting during the second or t h i r d g e s t a t i o n a l month but af t e r t h i s time, an increasing proportion of women who aborted at each subsequent gestational month were smokers. Graphs (mean percentage smoking, standard error) i l l u s t r a t i n g these r e s u l t s were not accompanied by data on numbers of spontaneous abortions. A l a t e r study of 5,272 .100 women from the same community (Persson et a l . , 1978) did not confirm an a s s o c i a t i o n between smoking and spontaneous abortion. K l i n e and colleagues (1977) sought to provide estimates of r i s k i n a case-control study of c i g a r e t t e smoking during pregnancy among 574 women who aborted spontaneously (cases) and 320 women who delivered l i v e infants a f t e r 29 weeks gestation ( c o n t r o l s ) . Cases were derived from a consecutive ser i e s of women between the ages of 18^40 years h o s p i t a l i z e d for spontaneous abortions at any of three Manhattan hospital s (82 percent of 703 e l i g i b l e cases agreed to p a r t i c i p a t e ) . Controls were selected from women attending public prenatal c l i n i c s of the same hos p i t a l s p r i o r to 22 weeks gestation and matched within two years of age (86 percent of controls selected f or 371 cases consented to interview). Non-respondents were reported to be s i m i l a r to pa r t i c i p a n t s , cases and controls were found to be comparable. Previous reproductive h i s t o r y was not discussed i n the report. Analysis c o n t r o l l e d for age (3 age groups), number of previous spontaneous abortions (3 c l a s s i f i c a t i o n s ) , induced abortions (2 c l a s s i f i c a t i o n s ) , and number of previous l i v e b i r t h s (3 c l a s s i f i c a t i o n s ) and examined the a s s o c i a t i o n of smoking according to three categories (none, 1-19 cigarettes/day, 20+/day). Analysis did not include the p o t e n t i a l e f f e c t of d i f f e r e n t l e v e l s of smoking i n the non-respondents. Study women who aborted spontaneously reported smoking during pregnancy more often than those who delivered a l i v e b i r t h a f t e r 28 weeks gestation: 41 percent of cases and 28 percent of controls 101 smoked. The odds r a t i o for the highly s i g n i f i c a n t a s s o c i a t i o n with smoking (X? = 19.4, p<0.001) was 1.8 (1.3-2.5). The assoc i a t i o n of spontaneous abortion with smoking i s s i m i l a r i n a l l age groups and for o b s t e t r i c h i s t o r y . This study provides the best a v a i l a b l e estimate of r i s k . F e t a l Growth, M o r t a l i t y and Morbidity Of the many epidemiologic studies of smoking and pregnancy over the l a s t three decades, a l l have shown that pregnant women who smoke have babies of lower birthweights than comparable women who do not smoke and many have demonstrated a dose-response r e l a t i o n s h i p — o n average, smokers' infants are 150-250 g l i g h t e r and are twice as l i k e l y to weigh les s than 2,500 g. In populations where 34-54 percent of mothers smoked during pregnancy, from 21-39 percent of the incidence of low birthweight could be at t r i b u t e d to maternal smoking (AR:1.64-2.21) (Andrews & McGarry, 1972; Butler & Alberman, 1969; Fabia, 1973; Meyer, Jonas, & Tonascia, 1976; Niswander & Gordon, 1972). Given that there i s un i v e r s a l agreement about a smoking e f f e c t on birthweight, i t seems appropriate to proceed from t h i s point to examine the implications of the birthweight e f f e c t on aspects of f e t a l s u r v i v a l , growth and development, and to examine the possible i n t e r a c t i o n between smoking and other f e t a l growth retarding f a c t o r s . The study by Andrews and McGarry (1972) of the pregnancies of a l l women i n the c i t y of C a r d i f f between 1965 and 1968 was designed to avoid errors due to sampling and small numbers. Obstetric and s o c i a l 102. data were recorded for the 18,631 pregnancies. Information on smoking habits was unfortunately not documented u n t i l the postpartum period and the authors suggest t h i s may have resulted i n a conservative estimate of cigarettes smoked for some pregnancies. The study r e s u l t s confirm the consistent weight reduction among infants born to smoking mothers (average 170 g, s i n g l e t o n s ) ; an increased proportion of low birthweight infants (10.4 percent among smokers, 6.1 percent among nonsmokers); and a dose-response e f f e c t on birthweight with average weight for infants of ex-smokers (quit before or during f i r s t trimester) between that of nonsmokers and l i g h t smokers—there was a progressive f a l l i n average birthweight of infants and increase i n low birthweight percentage with number of cigarettes (4 categories) smoked by the mother. Smokers had an increased proportion of preterm b i r t h s (9.2 percent versus 6.7 percent; xS = 28.2, p<0.001). Birthweight was adversely affected by other factors such as s o c i a l c l a s s , presence of anaemia, low maternal weight, age and p a r i t y , but i n a l l cases, smoking exhibited an a d d i t i o n a l independent e f f e c t . The study r e s u l t s show a decrease i n p e r i n a t a l mortality among low birthweight infants born to smokers but an o v e r a l l small and s i g n i f i c a n t increase i n m o r t a l i t y of infants of smoking mothers (3.1/1,000 versus 2.5/1,000) associated with antepartum hemorrhage, with s t i l l b i r t h s and pneumonia, r e s p i r a t o r y d i s t r e s s syndrome and immaturity with neonatal deaths. The incidence of congenital anomalies, with the exception of c l e f t l i p and palate, was not s i g n i f i c a n t l y 103 associated with maternal smoking. The incidence of hypertension and pre-eclamptic toxaemia was reduced i n pregnancies of smoking women. The prospective cohort study of Kullander and Ka l l e n (1971), described i n the discussion on smoking and spontaneous abortion, also reported increased m o r t a l i t y among infants of smoking mothers as we l l as reduced birthweight and increased proportion of low birthweight i n f a n t s . A l l f e t a l growth dimensions—body length, head circumference, and shoulder circumference—and placental weights decreased with increases i n amount smoked, while the pla c e n t a l weight/body weight r a t i o increased. Excess mortality re l a t e d to smoking was more l i k e l y to occur i n infants of birthweights greater than (not less than) 2,500 g and to occur a f t e r the f i r s t week of l i f e (RR:2.7). By the end of the f i r s t year, 1.9 percent of infants born to nonsmoking mothers had died, whereas 3 percent of those born to smoking mothers had died (RR:1.6). In 1975, Goujard, Rumeau, and Schwartz reported on a study of maternal smoking and s t i l l b i r t h using data from the 100 s t i l l b i r t h s which occurred to 9,169 women who were p a r t i c i p a t i n g i n the prospective study of pregnancies and b i r t h s i n 13 Paris maternity h o s p i t a l s i n the mid - s i x t i e s . The study showed a su b s t a n t i a l increase i n s t i l l b i r t h s among smokers—the rates per thousand of s t i l l - b o r n for the study population were 23.3 among smokers and 9.2 among nonsmokers (p<0.001; RR:2.5), with the highest incidence occurring to high p a r i t y (4+) smokers. S t i l l b i r t h s according to cause were s i g n i f i c a n t l y r e l a t e d 104 to smoking, during pregnancy i n two categories: abruptio placentae (smoking proportion:46 percent, p<0.0005), which represented 13 percent of s t i l l b i r t h s , and "unknown" cause (smoking proportion:35 percent, p<0.0005) which represented 37 percent of the s t i l l b i r t h s . Study analysis does not r u l e out i n t e r a c t i o n s with other factors ( i n addition to parity) important to outcome. A smoking-related increase i n mortality has not been reported by a l l studies and i t has been suggested that t h i s may be due to problems of sample s e l e c t i o n , s i z e , and methods of analysis (Butler et a l . , 1977; Meyer & Comstock, 1972). Meyer, Tonascia, and Buck (1975) suggest that the f a i l u r e of some studies to f i n d a s i g n i f i c a n t increase i n p e r i n a t a l m o r t a l i t y may be due to s e l e c t i o n of low r i s k study populations where l i g h t smoking i s associated with only a s l i g h t increase i n p e r i n a t a l r i s k , whereas other studies may s e l e c t higher r i s k populations where the influence of smoking on m o r t a l i t y i s stronger. The s i g n i f i c a n c e of the r e s u l t s w i l l depend on the magnitude of the d i f f e r e n c e , the amount smoked, and the s i z e of the study. Data from the Ontario P e r i n a t a l M o r t a l i t y Study (Ontario, Department of Health, 1967) which c o l l e c t e d information on antecedent, prenatal, and p e r i n a t a l factors and events from a l l s i n g l e b i r t h s i n ten teaching h o s p i t a l s i n Ontario during 1960-1961 was selected as an appropriate sample to evaluate further the a s s o c i a t i o n between smoking and f e t a l / i n f a n t m o r tality (Meyer, Tonascia, & Buck, 1975; Meyer, Jonas, & Tonascia, 1976). A t o t a l of 51,49.0 b i r t h s , i n c l u d i n g 701 f e t a l deaths and 655 neonatal deaths are recorded. The o r i g i n a l purpose of 105 the study was to obtain information concerning the causes of p e r i n a t a l l o s s . Interviews with the mother, which included information regarding the maximum number of cigarettes smoked during pregnancy, were conducted i n the early postpartum period. During t h i s same period o b s t e t r i c a l / medical information was confirmed with attending physicians. For a n a l y s i s , three l e v e l s of smoking frequency were established (none, les s than one pack/day, one or more pack/day) for nine population subgroups (37 subgroup categories) and f i v e outcome measures were used. Binary v a r i a b l e multiple regression analysis provided for simultaneous adjustment of outcome rates for multiple factors important to outcome and unevenly d i s t r i b u t e d among smoking-level groups. Smoking information was a v a i l a b l e for 98 percent of study p a r t i c i p a n t s — 5 7 percent were nonsmokers, 30 percent less than one package of cigarettes per day, and 13 percent smoked at l e a s t a pack per day. S i g n i f i c a n t smoking-related increases i n percent low birthweight, and preterm b i r t h s ; placenta previa, abruptio placentae, and p e r i n a t a l m o r t a l i t y were found, independent of mother's height, weight, h o s p i t a l status, age-parity group, b i r t h p l a c e , previous pregnancy h i s t o r y , weight gain, time of r e g i s t r a t i o n , and sex of c h i l d . Maternal smoking had the strongest e f f e c t on birthweight i n the 8 factor regression (weight gain excluded), and percent low birthweight increased d i r e c t l y with smoking l e v e l from 20 percent to 340 percent i n the 37 data subgroups. Preterm b i r t h s of l e s s than 38 weeks increased 20 percent and 50 percent and p e r i n a t a l m o r t a l i t y , 20 percent and 35 percent for infants of le s s than one pack/day and one or more pack/day smokers, 106 a f t e r adjustment. Placental complications increased co n s i s t e n t l y with smoking l e v e l i n a l l but one of 37 data subgroups—adjusted rates f o r placenta previa increased to 25 percent and 92 percent; abruptio placentae, 23 percent and 86 percent for the two l e v e l s of smoking compared to rates for. nonsmokers. These placental complications accounted for one-third to one-half the p e r i n a t a l deaths a t t r i b u t a b l e to smoking. Relative [RR] and a t t r i b u t a b l e r i s k s [AR] for smoking compared to nonsmoking outcomes were reported as follows: birthweight under 2,500 g, RR:2.02, AR:31 percent; preterm b i r t h , RR:1.36, AR:14 percent; p e r i n a t a l m o r t ality, RR:1.27, AR:10.5 percent; p l a c e n t a l complications, RR:1.43, AR:16 percent. Confidence i n t e r v a l s f or r e l a t i v e r i s k were, not reported analysis. Apart from the aspect of r i s k estimates for smoking, these analyses of the Ontario data show that i n each case where a maternal r i s k factor was r e l a t e d to poor reproductive outcome, smoking increased the p r o b a b i l i t y of poor outcome occurring. Data from the Collaborative P e r i n a t a l Project of the National I n s t i t u t e of Neurological and Communicative Disorders and Stroke, a large prospective study of pregnancy c a r r i e d out i n the United States, also confirms an a s s o c i a t i o n between smoking and f e t a l and neonatal deaths that varies with maternal r i s k status (Niswander & Gordon, 1972). This study followed the course of 53,518 pregnancies i n 12 h o s p i t a l s a f f i l i a t e d with US medical schools between 1959-1966 and recorded events of gestation, labor, d e l i v e r y , and the neonatal period. 107 As i n the Ontario study, maternal smoking was associated with p a r t i c u l a r causes of f e t a l and infant death. Naeye, Harkness, and Utts (1977) analyzed factors that might be involved i n abruptio placentae using the data from the Collaborative Project. The p e r i n a t a l m o r tality rate due to abruptio placentae was 3.96/1,000 b i r t h s and was the second most frequent cause of p e r i n a t a l death i n the study population. A f t e r the f i r s t trimester of pregnancy, enough information was a v a i l a b l e for 86 percent of deaths to provide a primary diagnosis regarding the disorder i n i t i a t i n g death. Of the 3,987 deaths reported, 138 s t i l l b i r t h s and 74 postnatal deaths were at t r i b u t e d to abruptio p l a c e n t a e — t h e disorder had peak frequencies between 20-29 weeks and aft e r 38 weeks of gestation. From over 1,000 demographic, hereditary, s o c i a l , medical, and postmortem va r i a b l e s analyzed i n the study, the variables which had a s i g n i f i c a n t influence on the frequency of f a t a l cases of abruptio placentae without having s i g n i f i c a n t i n t e r a c t i o n s with other v a r i a b l e s were found to be: seizures by gravida, low hemoglobin, low pregnancy weight gain, number of cigar e t t e s smoked per day, intrapartum hypertension, sex of i n f a n t , number of previous abortions, number of p r i o r p e r i n a t a l deaths, and number of p r i o r preterm d e l i v e r i e s . Decidual necrosis at the pla c e n t a l margin and large placental i n f a r c t s were the most c h a r a c t e r i s t i c p l a c e n t a l abnormalities i n f a t a l abruptio placentae cases. Decidual necrosis was seen twice as frequently i n women of very low pregnancy weight gain (<10 lb) than with those who had gained more than 19 l b , and was most common i n the heaviest smokers. The association with 108 smoking was greatest at the lower gestational ages, and i n pregnancies where the mother had a low weight gain, was anemic or had a h i s t o r y of p r i o r unsuccessful pregnancies. A r e l a t i o n s h i p between smoking and large p l a c e n t a l i n f a r c t s occurred only with women who had a h i s t o r y of p r i o r spontaneous abortion. Normal pregnancies also had a p o s i t i v e c o r r e l a t i o n between smoking and/or low weight gain and decidual necrosis at the margin of the placenta and smoking and placental i n f a r c t s . The fetuses and neonates who died following abruptio placentae had a pattern of growth re t a r d a t i o n c h a r a c t e r i s t i c of antenatal undernutrition, however, no information i s a v a i l a b l e on growth patterns of infants who did not d i e where the placentae was characterized by decidual necrosis or i n f a r c t s . The r e s u l t s from t h i s study emphasize the importance of recognizing that l i f e s t y l e factors may be i n f l u e n t i a l i n determining cause-specific f e t a l or infant mortality rates. Moreover, they emphasize that these e f f e c t s may go unrecognized i n s i t u a t i o n s where only general m o r t a l i t y rates are investigated, where key maternal r i s k v a r i a b l e s are not reported or reviewed, and where diagnosis of death does not r e f l e c t the disorder that i n i t i a t e d the cause of death. Further to t h i s study, Naeye (1979) found that i n the three p l a c e n t a l conditions most influenced by smoking and r e l a t e d to poor outcome—placenta previa, abruptio placentae, and large p l a c e n t a l i n f a r c t s — t h e frequency of the disorders were influenced by both the numbers of years mothers had smoked (increased frequency with duration of over 6 years) and t h e i r current smoking habits. The same 109 Collaborative data were used to look at past and present smoking habits on birthweight (Wainright, 1983) i n the group of 319 white women who changed smoking patterns between two consecutive pregnancies (cases) and 319 women who did not report a change (c o n t r o l s ) . P a irs were matched by birthweight of t h e i r f i r s t i n f ant (reproductive p o t e n t i a l ) , age, sex of i n f a n t , p a r i t y and time i n t e r v a l between pregnancies. The infants of women who started smoking i n t h e i r second pregnancy, weighed less (not s t a t i s t i c a l l y s i g n i f i c a n t ) on average than the infants of th e i r nonsmoking con t r o l pair or t h e i r previous i n f a n t s . The infants o f women who stopped smoking p r i o r to the second study b i r t h were s i g n i f i c a n t l y heavier (average, 171 g; p<0.01) than infants of controls who continued to smoke. These study'results, and those of a s i m i l a r study by Silverman (1977) suggest the birthweight d e f i c i t between consecutive pregnancies i s greater when smoking was continued over the two pregnancies than when smoking began j u s t p r i o r to the second and that a smoking duration e f f e c t on birthweight should not be discounted. Another example of s p e c i f i c "cause of death" i n v e s t i g a t i o n showing smoking during pregnancy to be a m o r t a l i t y - r e l a t e d factor comes from attempts to understand the etiology of the Sudden Infant Death Syndrome [SIDS], While there i s much s t i l l to be learned about t h i s syndrome, studies that have reported on both maternal smoking habits and SIDS have shown an as s o c i a t i o n between the two (Bergman & Weisner, 1976; Steele & Langworth, 1966). Results from the case-control study of 125 SIDS victims from the Collaborative Project (Naeye, Ladis, & Drage, 1976) are t y p i c a l . A search f o r predisposing factors f or SIDS ( i . e . , 110 events that might have damaged f e t a l or infant brains) compared information c o l l e c t e d during the Collaborative Study for 375 infants matched with SIDS victims for place of b i r t h , date of d e l i v e r y , gestational age, sex, race, and socioeconomic status, and 53,721 liv e b o r n infants who survived the neonatal period (unmatched controls) against data c o l l e c t e d for the 125 SIDS victims. Infants with congenital anomalies were excluded from a l l groups. The demographic p r o f i l e of SIDS f a m i l i e s were the same as those of f a m i l i e s with excessive p e r i n a t a l mortality. Evidence of neonatal b r a i n dysfunction could not be r e l a t e d to events i n labor and d e l i v e r y , and there was l i t t l e evidence of genetic or hereditary influence. A greater proportion of infants who l a t e r died of SIDS were mi l d l y growth retarded at b i r t h or showed postnatal growth retardation p r i o r to death. Their mothers were more l i k e l y to be young, of low socioeconomic status, anemic; to have had v a g i n i t i s or peurperal i n f e c t i o n during pregnancy, to have had p r i o r f e t a l losses, and to have smoked at l e a s t 6 cigarettes/day during pregnancy (46 percent versus 25 percent, p<0.001). A v a r i e t y of neurologic abnormalities were more common i n future SIDS victims than i n the selected.controls and multiple evidence of probable brain dysfunction was found. The mechanism by which smoking might contribute to the r i s k of SIDS i s not known although the frequency with which f e t a l growth re t a r d a t i o n occurs i n r e l a t i o n to both smoking during pregnancy and f e t a l or infant m o r t a l i t y has provided a major focus for such i n v e s t i g a t i o n s . I l l The type of f e t a l growth retardation which occurs may i t s e l f be a clue to the underlying mechanism of action (or perhaps to a common pathway of growth retardant a c t i o n ) — t w o main types of f e t a l growth retardation were discussed i n chapter two. Smoking i s associated with a higher incidence of growth retarded infants with short crown-heel length and proportionately smaller weight, head circumference, and other body dimensions. In the 1974-1975 Swedish prospective study of 5,272 pregnant women described e a r l i e r (Persson et a l . , 1978), ultrasound measurements of f e t a l b i p a r i e t a l diameter [BPD] were made from the 18th-20th week of gestation, growth curves were constructed separately for term b i r t h s of the smoking (51 percent) and nonsmoking group (49 percent). The BPD increased f a s t e r during gestation i n the nonsmoking group, the d i f f e r e n c e being apparent at 22 weeks, s i g n i f i c a n t a f t e r the 28th week and p o s i t i v e l y correlated to number of cig a r e t t e s smoked. The study found a lower birthweight, smaller head circumference, and shorter o v e r - a l l length at b i r t h i n the smokers than i n the nonsmokers — t h e heavy smoking group (>20 cigarettes/day, 3 percent smokers) compared with the nonsmoking group showed a 6 percent reduction i n birthweight and a 2 percent reduction i n length, head circumference, and BPD growth. The study confirmed that f e t a l s i z e at b i r t h i s d i r e c t l y influenced by maternal s i z e and weight gain during pregnancy. The importance of t h i s study i s i t s demonstration of an early onset of o v e r a l l s i z e reduction i n the i n f a n t s of smoking mothers. The disproportionate retardation of s k e l e t a l growth i n premature and low-weight infants of smoking mothers compared to comparable 112 infants of nonsmoking mothers i s reported by others who have c a r r i e d out broad-based population studies of large numbers of pregnancies (Kullander & Kallen, 1971; U l r i c h , 1982) and many who have c a r r i e d out studies on selected pregnant populations (Luke, Hawkins, & P e t r i e , 1981; M i l l e r , Haasanein, & Hensleigh, 1976; M i l l e r & M e r r i t t , 1979). D i f f e r e n t f e t a l growth retardation p r o f i l e s seem to be associated with smoking and low maternal weight gain during pregnancy—higher proportions of short-for-dates infants i n the former and of low ponderal index i n the l a t t e r (although both types occur i n either s i t u a t i o n ) — a n d with d i f f e r e n t patterns of early postnatal growth ( M i l l e r , Haasanein, & Hensleight, 1976). Where studies have s u b c l a s s i f i e d the growth retarded or small-for-ge s t a t i o n a l age i n f a n t for purposes of followup, retarded growth i n the postnatal period has been associated with the short-for-dates as w e l l as the low ponderal index infants (Ounsted et a l . , 1971), with those who did not achieve a "catch-up" growth by 6 months (Fitzhardinge & Steven, 1972), with low birthweight infants of longer gestation (Beck, 1974), with those of s i g n i f i c a n t s k e l e t a l stunting i n height and/or head circumference (Lubchenco et a l . , 1976), and with short rather than lean infants (Holmes, 1977; M i l l e r & M e r r i t t , 1979). I t might be expected then that infants of smoking mothers who are more l i k e l y to be short-for-dates would exhibit signs of growth retardation i n the postnatal period. The r e s u l t s of the l o n g i t u d i n a l study of the 17,000 B r i t i s h c h i l d r e n who were f i r s t studied as part of the P e r i n a t a l M o r t a l i t y 113 Survey (Davie et a l . , 1972) and then investigated at the ages of 7 and 11 years confirms.a l i n k between prenatal smoking and impairment of both mental and p h y s i c a l growth a f t e r adjustment for maternal height, age, s o c i a l c l a s s , p a r i t y and sex of c h i l d . These e f f e c t s were seen at both age 7 and age 11 and the d e f i c i t s increased with the number of cigarettes smoked a f t e r the fourth month of pregnancy. Children of mothers who smoked 10 or more cigare t t e s a day are on average 1.0 cm shorter and between 3-5 months retarded on reading, mathematics, and general a b i l i t y compared with o f f s p r i n g of nonsmoking mothers. While these r e s u l t s are s t a t i s t i c a l l y s i g n i f i c a n t , the smoking-nonsmoking differences are considerably smaller than the differences associated with other s o c i o b i o l o g i c a l factors and would u n l i k e l y be detected at a l e v e l of s i g n i f i c a n c e i n studies of smaller populations (Butler & Goldstein, 1973). Given the r e s u l t s of the Newcastle Survey of Child Development (Neligan et a l . , 1976) showing both short-gestation and l i g h t - f o r - d a t e s groups of c h i l d r e n perform s i g n i f i c a n t l y l e s s w e l l than the random sample comparison over the whole range of measures of performance (psychometric, behavioral and temperamental, neurological and p h y s i c a l growth assessment) studied at ages 5, 6, and 7 years and given that smoking during pregnancy increases the proportion of these two groups, longterm smoking-related impairment i n growth and development i s very p l a u s i b l e . A further example of longterm e f f e c t s related to maternal smoking during pregnancy i s reported by Rantakallio (1978). From a study of 12,068 b i r t h s (96 percent of a l l b i r t h s i n two provinces of Finland), 114 1,819 mothers were smokers (83 percent of these smoked <10 c i g a r e t t e s / day) according to data c o l l e c t e d from the s i x t h month of pregnancy to early postpartum. The infants of these 1,819 mothers and a subgroup of 1,819 infants from the t o t a l group matched for mother's age, p a r i t y , m a r i t a l status, and place of residence were followed for a period of 5 years to examine differences i n mortality and morbidity. Results show postneonatal m o r t a l i t y to be higher among infants of smokers (RO:3.78 for <2,500 g i n f a n t s ; RO:2.22 for >2,500 g i n f a n t s ) . An advantageous p o s i t i o n of the low birthweight infants of smokers during the p e r i n a t a l period when mortality was compared with those of controls, was t o t a l l y l o s t during the postneonatal period when both heavier and low birthweight infants of smokers were more affected than co n t r o l s , and the low birthweight infants more so than the heavier i n f a n t s . Morbidity was s i g n i f i c a n t l y higher (p<0.001) among c h i l d r e n of smokers, as measured by number of admissions to. h o s p i t a l (higher for a l l birthweights) and incidence of disease. Diseases i n which the d i f f e r e n c e between the smokers and controls was s i g n i f i c a n t included: r e s p i r a t o r y , blood and skin diseases, and those of the nervous system and sense organs. Taken together, the r a t i o between the smokers and controls was 3.02 among the low birthweight infants and 1.65 among the infants with birthweights of 2,500 g or more. Clues to the underlying mechanisms of smoking-related reproductive problems are sought i n pathological and p h y s i o l o g i c a l studies of the p l a c e n t a l , membrane, c i r c u l a t o r y , serum, c e l l and t i s s u e changes seen 115 to be associated with smoking (e.g., Asmussen, 1978; Christianson, 1979; Meyer, Jonas, & Tonascia, 1976; Naeye, 1978b; Oschner, 1976; Resnik, Brink, & Wilkes, 1979). The e f f e c t s of three constituents of smoke—carbon monoxide, n i c o t i n e , and cyanide/thiocynate—have been extensively investigated and are reviewed by S i d l e (1982). Carbon monoxide, i n combining with haemoglobin, decreases the • a v a i l a b l e oxygen supply from the blood. Carbon monoxide mediated hypoxia i s one of the most accepted p o t e n t i a l mechanisms by which smoking could influence pregnancy outcome. Consideration i s given to the p o s s i b i l i t y that p l a c e n t a l pathology linked with smoking i s an adaptation to smoking induced hypoxia. There i s evidence that n i c o t i n e has both independent and s y n e r g i s t i c e f f e c t s when combined with carbon monoxide (Boyle et a l . , 1957; Krisna, 1978; Krous et a l . , 1981; S i d l e , 1982). Nicotine i s seen to produce v a s o c o n s t r i c t i o n v i a sympathetic stimulation, leading to a reduction i n uterine blood flow. The presence of both n i c o t i n e and cyanide has implications f or the a v a i l a b i l i t y and requirements of e s s e n t i a l n utrients, for the l e v e l of metabolic a c t i v i t y , and for the subsequent growth and development p o t e n t i a l of the fetus (Andrews & McGarry, 1972; Crosby et a l . , 1977; Olubadewa et a l . , 1978; Rowell & Sastry, 1978; Sontag & Wallace, 1935). Smoking causes a s i g n i f i c a n t increase i n maternal cyanide and thiocyanate and i n f e t a l thiocyanate l e v e l s (Pettigrew et a l . , 1977). 116 A negative c o r r e l a t i o n between maternal thiocyanate l e v e l s and birthweight i s seen (Meberg et a l . , 1979). I t i s not known i f the fetus i s exposed d i r e c t l y to cyanide and i f the exposure i s at toxic l e v e l s , or i f only to the d e t o x i f i e d product thiocyanate, which i s i t s e l f an a n t i - t h y r o i d agent that could e f f e c t thyroid dependent organ development. In a review of the e f f e c t of smoking on placental and maternal immune competence, Wynn and Wynn (1981) state " f a i l u r e to r e t a i n the fetus appears i n many cases to be a f a i l u r e i n immunological mechanisms invol v i n g both the placenta and the mother" (p. 30). The studies of Naeye and colleagues (Naeye & Blanc, 1970; Naeye et a l . , 1971) associate placenta and membrane i n f e c t i o n to f e t a l growth retardation, bleeding, preterm b i r t h , s t i l l b i r t h , and p e r i n a t a l death. This i n f e c t i o n can i n d i c a t e either increased exposure or reduced resistance to i n f e c t i o n . Many constitutents of smoke have independent immunosuppressive e f f e c t s (Esber et a l . , 1973; Gulsvik & Fagerhol, 1979; Kraal, 1978; S i l v e t t e , Larson, & Haag, 1957; Thomas, Holt, & Keast, 1975). Smokers are seen to have a higher prevalence of i n f e c t i o n s than nonsmokers (Holt, Thomas, & Keast, 1973; Martin-Boyce, David, & Schwartz, 1977) and i n the Collaborative Study (Naeye, 1978b) mothers who smoked had an increased incidence of v a g i n a l , c e r v i c a l , and amniotic f l u i d i n f e c t i o n s . Immune status, as has been discussed, i s modified by n u t r i t i o n a l status. Since an intimate r e l a t i o n s h i p e x i s t s between the endocrine and immune systems (Castro, 1978) i t could be speculated that depression 117 of maternal endocrine function r e l a t i n g to smoking, n u t r i t i o n , or other factors may lead to reduced c e l l r e p l i c a t i o n of embryo, fetus, and placenta, and, i n turn, to reduced immune competence i n the i n f a n t (Wynn & Wynn, 1981). Consideration of the i n t e r a c t i o n between immune system function, and the changes i n nutrient a v a i l a b i l i t y and requirements with smoking and/or i n f e c t i o n , as w e l l as the i n t e r a c t i v e e f f e c t s of other known r i s k f a c t ors provides a t h e o r e t i c a l explanation of the many types and v a r i a t i o n s i n smoking-related pregnancy outcome (Meyer & Tonascia, 1977; Naeye et a l . , 1973; Naeye, Harkness, & Utts, 1977). Examples of v a r i a t i o n i n outcome include the increased r i s k (doubled) to c h i l d r e n of the lower socioeconomic groups a t t r i b u t e d to maternal smoking (Meyer, Tonascia, & Buck, 1975) and the 6-fold increase i n abruptio placentae with heavy maternal smoking or alcohol intake reported by Goujard (.1978) that becomes a 30-fold increase when both occur together. Smoking Prevalence The most recent broad-based data concerning the prevalence of smoking amongst women of childbearing age i n B r i t i s h Columbia i s reported as part of the r e s u l t s of the 1978-1979 Canada Health Survey (Ottawa, Canada, 1981). This c r o s s - s e c t i o n a l survey was intended to provide health s t a t i s t i c s compatible with the o u t l i n e of the health f i e l d c o n c e p t — i n p a r t i c u l a r , to provide missing data about the d i s t r i b u t i o n of r i s k factors of l i f e s t y l e o r i g i n i n the Canadian population. 118 The sample design consisted of approximately 12,000 households from 100 geographical c l u s t e r s ( s t r a t i f i e d representation of the population by p r o v i n c i a l region and population, urban centers and r u r a l areas), representing 40,000 persons to be interviewed by s p e c i a l l y trained personnel, and a subset of 4,200 households to provide phy s i c a l measures information. Budget constraints required that the i n i t i a l survey design be altered somewhat p r i o r to i t s i n i t i a t i o n by a reduction i n subset p a r t i c i p a n t s . The response rate was 86 percent (10,571/12,218 households) for the p h y s i c a l measures component. Non-response adjustments at the household l e v e l i n e f f e c t replaced the non-respondent with an "average" household from the same c l u s t e r and month of survey. Non-responses at the person l e v e l were excluded (determined as a n e g l i g i b l e e f f e c t on r e s u l t s ) and non-response for i n d i v i d u a l items was weighted i n a manner that assumed respondents and non-respondents from the same province-stratum-age-sex group were the same, although a study of non-response indicated that non-respondents tended to be s l i g h t l y l e s s healthy. The data are considered to be from a representative, non-volunteer sample of the n o n - i n s t i t u t i o n a l i z e d population of Canada that do not l i v e i n the T e r r i t o r i e s , i n remote areas, or on Indian Reserves (3 percent of t o t a l population). In general, the survey found differences between the sexes i n terms of both the rate and l e v e l of exposure to l i f e s t y l e r i s k s , as w e l l as a change with age, and v a r i a t i o n with s o c i a l status (defined by occupation and income). Data on tobacco use were c o l l e c t e d from persons 15 years of age and over by self-administered questionnaire. The survey reports about 40 percent of adult Canadians smoke cigarettes d a i l y , and one-third of these smoke 23 or more cigarettes per day, and nearly one-quarter of adult Canadians are former smokers. Smoking was most prevalent among those with low education, among the unemployed, and among those i n blue c o l l a r jobs. Income was not strongly r e l a t e d to d a i l y c i g a r e t t e smoking. Heavy smokers were the l e a s t l i k e l y to have recently t r i e d to cut down on t h e i r smoking. Current drinkers were the most l i k e l y to be current smokers; heavy drinkers were most l i k e l y to be heavy smokers. The percentage of women reporting d a i l y c i g a r e t t e smoking according to the data from the Canada Health Survey i s compared to data from the Smoking Habits of Canadians Surveys conducted i n 1977 and again i n 1979 (Ottawa, Canada, 1979, 1980), as follows. 120 TABLE 9 Proportion of Female Population 15 and over Reporting Daily Cigarette Smoking, by Age, i n the Smoking Habits of Canadians Survey, December 1977 and December 1979, and Canada Health Survey, 1978-1979 Percentage of population who are d a i l y smokers Smoking Habits of Canadians Canada Health Survey 1977 1979 1978-1979 Women: A l l ages 31.1 30.1 35.8 15-19 years 26.7 26.0 35.6 20-24 years 40.7 39.8 46.5 25-44 years 36.6 36.0 38.9 45-64 years 30.5 28.9 34.8 Source: Adapted from Canada Health Survey, Table IV, p. 48. A previous study which examined the r e s u l t s of the Smoking Habits of Canadians Surveys from 1965-1975 (Thompson, 1978) found t o t a l c i g a r e t t e consumption to be underreported by 14-20 percent when compared to tobacco sales data. This suggests that the higher estimates of the Canada Health Survey may provide a closer approximation of the prevalence of smoking i n Canada. The proportion of men and women smokers under 25 are v i r t u a l l y the same, a f t e r t h i s , r e l a t i v e l y more men than women smoke. Over h a l f the women under 25 are l i k e l y to have started smoking before age 16, which contrasts with t h e i r mothers who were les s l i k e l y to begin before age 21 and, i n f a c t , were les s l i k e l y to smoke. 121 Table 10 compares the prevalence of smoking among women of d i f f e r e n t age groups i n Canada and i n B r i t i s h Columbia. TABLE 10 Tota l Female Population 15 Years and Over and Percent D i s t r i b u t i o n by Type of Cigarette Smoker and by Number of Cigarettes Smoked Dai l y , by Age Group, for...Canada-.and.for.-British Columbia, Canada Health Survey, 1978-1979 Type of Smoker Numbers of Women i n Population ( i n thousands) ..BC Age Groups: Canada T o t a l .15-19 20-24 25-44 45-64 65+ Never smoked % 38.8 % 38.4 % 27.9 % 33.2 % 40.2 % 58.7 Former smoker 21.6 16.6 19.8 21.3 17.3 11.2 Current smoker Occasional 1.8 6.2 4.3 4.0 2.1 1.3 Daily 30.8 33.9 45.2 37.3 31.9 13.8 # Cigarettes/day 1-12 10.8 16.8 16.9 1.1.0 9.8 7.4 13-22 14.0 11.6 18.9 14.5 12.9 3.3 23-32 4.0 4.4 7.2 9.9 7.0 1.2 >33 1.6 0.0 0.2 2.7 1.7 0.0 Type of smoker unknown 6.9 4.9 2.8 4.1 8.3 14.9 Tot a l Population % 99.9 100.0 100.0 99.9 99.8 99.9 N 974 1146 1108 3242 2279 . 1132 Source: Ottawa, Health & Welfare Canada, S t a t t i s t i c s Canada (1981) I t can be seen from these data that from 20-45 percent of the female population of childbearing years are regular smokers. At l e a s t as high 122 a proportion of the male population are regular smokers as w e l l . The percentage of women smoking throughout pregnancy i s unknown, although i t can be speculated that some w i l l q u i t when pregnancy i s confirmed. The prevalence of smoking around the period of conception and during pregnancy i s at a l e v e l where even a low r e l a t i v e r i s k of adverse reproductive outcome from exposure can be of public health s i g n i f i c a n c e . The implications of t h i s w i l l be examined i n the following chapter. Cannabis and Reproductive Health Cannabis i s i d e n t i f i e d as an independent health hazard i n t h i s study for two reasons: (a) m i l l i o n s of young people of reproductive age currently use the drug and, thus, adverse consequences, even i f infrequent, could be of p u b l i c health s i g n i f i c a n c e ; and (b) recent documentation of the health hazards associated with cannabis use (Fehr & Kalant, 1983) provides an exemplary model for i n v e s t i g a t i n g and reporting the e f f e c t s of l i f e s t y l e factors on health and reproductive outcome. However, since large prospective cohort studies to examine the e f f e c t of cannabis on reproductive outcome i n humans have not been conducted, i t i s not yet possible to determine a causal r e l a t i o n s h i p . During the early 1980s three expert committees published reports on the health implications of cannabis u s e — t h e Addiction Research Foundation of Ontario i n c o l l a b o r a t i o n with the World Health Organization (Addiction Research Foundation, 1981); the I n s t i t u t e of Medicine, National Academy of Sciences, USA (National Academy of Sciences, 1982); and the Advisory Council on the Misuse of Drugs, UK (Home O f f i c e , 1982). 123 These reports came to s i m i l a r conclusions, namely: That cannabis i s c l e a r l y capable of causing adverse e f f e c t s on health, that some of these have already been recognized c l i n i c a l l y i n heavy users, that others have been produced i n experimental animals but do not necessarily occur i n humans at the dose l e v e l s that have been employed by voluntary users u n t i l now, and that there i s a serious lack of systematic epidemiological information about the incidence and prevalence of such e f f e c t s i n d i f f e r e n t parts of the world, as a function of the extent of cannabis use i n those regions. (Fehr & Kalant, 1983, p. 10) The 1983 proceedings of the Addiction Research Foundation/World Health Organization S c i e n t i f i c Meetings provides a comprehensive review of the adverse e f f e c t s of cannabis use (the text of 800 pages includes over 2,000 references) in c l u d i n g general and c e l l u l a r t o x i c i t y ; the immune system; e f f e c t s on endocrine function, reproduction and development; epidemiology and consideration of the c l i n i c a l relevance of e f f e c t s seen i n animal experiments. This p u b l i c a t i o n i s the primary reference for the following overview. Other key references, i n add i t i o n to the previously mentioned expert committee reports, include Nahas 0.976) and Nahas and Paton (1979). A number of extensive reviews have focused on the i n t e r a c t i o n of marihuana and cannabinoids with endocrine and reproductive function i n animals and humans (Bloch, 1983; Nahas & Paton, 1979; Rosenkrantz & Esber, 1980; Rosenkrantz & Hayden, 1979; Smith et a l . , 1979b). There are approximately 60 compounds of cannabinoid structure w i t h i n the group of 421 compounds i d e n t i f i e d i n the cannabis plant (Turner, Elsohly, & Boeren, 1980). I t i s the euphoric e f f e c t s of ^9-tetrahydrocannabinol [THC] that a t t r a c t s i t s users (Mechoulam et a l . , 124 1980) and although other compounds are also implicated, t o x i c i t y appears r e l a t e d to the THC content of cannabis products (Fehr & Kalant, 1983). In general, marihuana (dried leaves) preparations contain 1-8 percent THC, hashish ( r e s i n and flowers) contains up to 15 percent, and hashish o i l (solvent extracts of l e a f , flower, or resin) can contain up to 60 percent THC. When cannabis products are smoked, the health consequences are considered s i m i l a r to those associated with tobacco smoke—the important d i f f e r e n c e i s that the concentration of toxic compounds i s greater i n cannabis smoke and the smoke contains the a d d i t i o n a l t o x i c i t y of the cannabinoid compounds (Leuchtenberger, 1983). The common constituents i n tobacco and marihuana c i g a r e t t e smoke considered to be health hazards by v i r t u e of t h e i r known carcinogenic, cocarcinpgenic and toxic properties (US,Department of Health, Education, & Welfare, 1972, 1979) are: tar, phenols, c r e s o l s , and polynuclear aromatic hydrocarbons i n the p a r t i c u l a t e phase (Hoffman et a l . , 1975; Lee, Novotny, & B a r t l e , 1976; Magus & H a r r i s , 1971); and nitrous oxide, carbon monoxide, hydrogen cyanide and nitrosamines i n the gas vapour phase (Leuchtenberger, 1983). Those smoking tobacco and cannabis concurrently are seen to be at greatest r i s k f o r pulmonary disease and lung cancer (Tennant, 1983). Lung biopsies from young males using both tobacco and hashish had precancerous lesions generally found only a f t e r years of heavy tobacco smoking (Tennant, 1980). In animals, THC decreases cardiac output and cerebral blood flow; i n humans, tachycardia and postural hypotension are co n s i s t e n t l y 125 reported (Hardman & Hosko, 1976; Tennant, 1983). Since s i m i l a r amounts of carbon monoxide are found i n cannabis and tobacco smoke, impaired oxygenation of the myocardium due to the formation of carboxyhemoglobin i s l i k e l y . Other general toxic e f f e c t s are associated with reduced growth and body weight (animals) (Rosenkrantz, 1983), g a s t r o i n t e s t i n a l d i s t r e s s (Halikas, Goodwin, & Guze, 1971; Rosenkrantz et a l . , 1975; Tennant, 1974; Thompson et a l . , 1973), enhancement of alcohol-induced l i v e r t o x i c i t y (human) (Tennant, 1983), decreased resistance to i n f e c t i o n (Munson & Fehr, 1983; N a l i n et a l . , 1978), and altered endocrine function (Bloch, 1983). In vivo and i n v i t r o studies of c e l l u l a r t o x i c i t y associated with cannabinoids have focused on the p o t e n t i a l f or mutagenicity, c a r c i n o g e n i c i t y and impairment of biosynthesis of nu c l e i c acids and proteins (Leuchtenberger, 1983). In both animal and human tissue culture studies p u r i f i e d cannabinoids were linked with a higher proportion of hypoploid c e l l s during the period of exposure with no measurable increase i n chromosone breaks or gaps (Bloch,_1983; Leuchtenberger, 1983; Matsuyama et a l . , 1977; Morishama et a l . , 1979; Nichols et a l . , 1974). Marihuana smoke, on the other hand, i s associated with both chromosome aberrations and hypoploidy (Leuchtenberger, 1983), mutagenicity, and impaired development i n the second generation of treated animals (Dalterio,. 1980; F r i e d & Charlebois, 1979). Cannabis smoke i s considered to have s i g n i f i c a n t carcinogenic p o t e n t i a l , not because of evidence that i t has produced cancer i n 126 humans, but because cannabis produces the same sequence of c e l l u l a r abnormalities produced by tobacco smoke. Abnormalities i n mit o s i s , DNA complement, chromosome number and c e l l d i v i s i o n were more severe a f t e r exposure to cannabis smoke than to tobacco smoke (Leuchtenberger, 19.83; Munson & Fehr, 1983). In animals, a decrease i n DNA content of immature sperm as w e l l as i n h i b i t i o n of RNA synthesis and of protein synthesis (Munson & Fehr, 1983) i s at t r i b u t e d to the cannabinoids i n marihuana smoke. Cannabinoid-induced i n h i b i t i o n of both c e l l growth and d i v i s i o n i s con s i s t e n t l y reported for a l l c e l l types as i s i n h i b i t i o n of the i n t r a c e l l u l a r synthesis of macromolecules, and i n h i b i t i o n of the incorporation of precursors of n u c l e i c acids and protein across systems. The relevance of these i n v i t r o findings to the general population i s not yet known, although abnormalities i n morphology and/or nucleoproteins of sperm from marihuana smokers has been reported (Hembree et a l . , 1979). I s s i d o r i d e s (1983) proposes a hypothesis regarding the possible biochemical mechanism of cannabis which provides: a s i n g l e explanation both for the deficiency i n nuclear histone synthesis and f o r the u l t r a s t r u c t u r a l a l t e r a t i o n s encountered i n the leukocytes and spermatozoa of chronic human users of hashish. These c e l l s are characterized by a depletion of the amino acid arginine and by abnormal chromatin condensation. According to t h i s hypothesis, the b i o l o g i c a l depletion which, i n i t s e l f , can cause chromosomal aberrations, decreased sperm maturity and m o t i l i t y , defective ovulation, growth re t a r d a t i o n , immunosuppression and the r e a c t i v a t i o n of l a t e n t v i r a l i n f e c t i o n s , and CNS e f f e c t s such as anorexia, motor incoordination, and lethargy. Furthermore, the enzymes reported to be affected by cannabis possess e s s e n t i a l arginine residues at t h e i r a c t i v e s i t e s , which would permit a THC/arginine i n t e r a c t i o n . (p. 13) 127 There i s consistent evidence that THC and marihuana induce immunological defects i n mice and rats (Morahan et a l . , 1979; Munson & Fehr, 1983). Both humoral and cell-mediated immune suppression are reported. The e f f e c t s are greater when exposure occurs during the early phase of antibody formation, i n young animals (Luthra et a l . , 1980; Pruess & Lefkowitz, 1978). There i s suggestive evidence of s i m i l a r immune dysfunction i n humans based on i n v i t r o studies ( I s s i d o r i d e s , 1979; Juel-Jensen, 1972; Petersen et a l . , 1975; Stefanis & I s s i d o r i d e s , 1976). Munson and Fehr (1983) state: I t i s l i k e l y that the degree of immunosuppression produced by d i f f e r e n t doses of cannabis, l i k e that of other immunosuppressants, w i l l vary along a continuum, ranging from s l i g h t e f f e c t s on resistance to i n f e c t i o n s , to a marked decrease i n resistance r e s u l t i n g i n death of the host. I t i s l i k e l y that we would now be aware of profound changes i n the resistance of human cannabis smokers.lif -these occurred frequently. However, we lack the epidemiological observations on large numbers of users that would be necessary to e s t a b l i s h the occurrence of small degrees of impairment that are suggested by the animal experiments. (p. 339) The Addiction Research Foundation/World Health Organization report concludes: A minor degree of immunosuppression i n a s u b s t a n t i a l number.of cannabis users might r e s u l t not i n any sudden and dramatic increase i n incidence of unusual i n f e c t i o n s , but rather i n a s l i g h t increase i n incidence, s e v e r i t y , and duration of common ones. Cumulatively, t h i s could have considerable s i g n i f i c a n c e for p u b l i c health and health care d e l i v e r y systems. (Fehr & Kalant, 1983, p. 16) E f f e c t s of cannabis on endocrine function are seen c o n s i s t e n t l y i n a n i m a l s — p a r t i c u l a r l y with respect to gonadel and adrenal hormone production and hypothalamic-pituitary r e g u l a t i o n — b u t i n c o n s i s t e n t l y 128 i n humans (Bloch, 1983; Bloch et a l . , 1978). The Addiction Research Foundation/World Health Organization expert committee (1981) suggest that since the endocrine responses of animals to most drugs are at l e a s t q u a l i t a t i v e l y s i m i l a r to human response, animal studies showing a consistent r e a c t i o n across species and classes are l i k e l y to be relevant to humans. There s t i l l remain c r i t i c a l gaps i n animal work examining endocrine response and reproductive outcome associated with cannabis exposure and there are few human studies. Bloch (1983), i n h i s review of the 200 key references i n t h i s f i e l d , summarizes the a v a i l a b l e evidence that cannabis adversely e f f e c t s (or has p o t e n t i a l to) reproduction. The most s u b s t a n t i a l work has been with rodents. The conclusions that can be drawn from these studies are as follows (Bloch, 1983): 1. In male animals, cannabinoids disrupt normal reproductive physiology and are associated with a r e v e r s i b l e decrease i n t e s t i c u l a r metabolic a c t i v i t y and i n v i t r o testosterone synthesis; and lowered plasma testosterone and LH l e v e l s (Bloch et a l . , 1978; Fujimoto et a l . , 1978; Huang, Nahas, & Hembree, 1979; Rosenkrantz & Hayden, 1979). With chronic exposure the androgenic target tissues show fu n c t i o n a l and morphological i n v o l u t i o n and spermogenesis i s diminished. The mechanism of action may be v i a reduced c h o l e s t e r o l esterase a c t i v i t y or prostaglandin synthesis, and through diminished metabolic a c t i v i t y and macromolecule synthesis. 2. In non-pregnant female animals, A.9-THC i n h i b i t s functioning 129 of the hypothalamic-pituitary-gonadal axis r e s u l t i n g i n decreased plasma LH and p r o l a c t i n ; suppressed pre-ovulatory LH surge and delayed estrus (Nir et a l . , 1973). With chronic exposure, the u t e r i and vagina show fu n c t i o n a l and morphological i n v o l u t i o n and ovulation may be interrupted or blocked (Cordova et a l . , 1980). As with males, these e f f e c t s are seen to be r e v e r s i b l e . During pregnancy, exposure to high doses i s associated with reduced maternal weight gain, a reduction i n p r o l a c t i n l e v e l s and with continued exposure, inadequate l a c t a t i o n i n the postpartum period (Bloch, 1983). 3. Cannabinoids stimulate adrenal c o r t i c a l function, r e s u l t i n g i n an increase i n adrenal weight and a decrease i n thymus weight, and diminish thyroid function (Bloch e t a l . , 1978). In animals, exposure during the f i r s t two trimesters i s associated with increased incidence of f e t a l resorption and growth re t a r d a t i o n i n the survivors (Bloch et a l . , 1978; S o f i a , Strasbaugh, & Banerjee, 1979). The growth re t a r d a t i o n was linked to exposure during the embryogenic and organogenic phases of development, not the f e t a l phase. When THC i s administered during pregnancy, the uptake by the placenta i s greater than that of the fetus and release i s slower. This suggests the placenta acts both as a b a r r i e r to and a re s e r v o i r for cannabinoid transfer (Bloch, 1983). THC may disrupt p l a c e n t a l development and function (Sassenrath, Chapman, & Goo, 1979) and i n t h i s way influence the p o t e n t i a l f o r f e t a l abnormalities. Large doses of A9-THC can increase the incidence of malformations i n mice and hamsters (Bloch, M o r r i l l , & Fujimoto, 1979; Geber & Schramm, 130 1969; Harbison, Mantilla-Plata, & Lubin, 1977; Joneja, 1976; Mantilla-Plata, Clewe, & Harbison, 1975; Persaud & Ellington, 1968). But since the incidence of malformationsassociated with THC varies across species and between studies, the question of teratoginicity of cannabinoids per se and cannabinoids versus other toxins in cannabis smoke remains unresolved. Studies on monkeys and nonhuman primates have not been extensive, but tentatively confirm the suppressive effect of cannabinoids on pituitary gonadotrophin release and on ovulation (Asch et a l . , 1979; Besch et a l . , 1977; Smith & Munson, 1976; Smith et a l . , 1979a). A number of mechanisms for cannabinoid action have been proposed, but conclusions cannot be drawn at present. The fact that the effects of cannabis are not exclusive to the cannabinoid compounds suggest a variety of mechanisms must be c l a r i f i e d . It is known that ^ 9-THC concentrates in tissues rich in l i p i d components (Bloch, 1983) and that in some respects cannabinoids mimic estrogen action and in others, act as estrogen antagonists (Chakravarty & Sengupta, 1980; Chakravarty et a l . , 1976; Harmon & Aliapoulios, 1972). Bloch (1983) , in answer to the question, "How well have animal studies served as models for the human situation?", states: Where a particular endpoint or system reacts uniformly to cannabinoid exposure in several species of several classes, including monkeys, a qualitatively similar response in humans may be inferred. This uniformity seems to be the case for acute and short-term responses of the reproductive system, including lactation. (p. 416) 131 Rosenkrantz and colleagues (1974, 1975, 1976, 1979) have developed a methodology for relating animal study dosages to human marihuana intake. According to these calculations, most studies approximate the daily smoking of .5-1 marihuana cigarettes given the current (and rising) A9-THC content of marihuana (US Secretary of Health, Education, & Welfare, 1980). It is suggested that approximately 10 percent of the marihuana-smoking population equal or exceed this level of daily exposure (Smart, 1983). In humans, marihuana acutely depresses the activity of the male pituitary-testicular axis (Hembree et a l . , 1979; Kalodny et a l . , 1974) and, according to Bloch (1983), a similar inhibitory affect may be expected to exist in females based on animal studies. One of the few human studies in this area (Bauman et a l . , 1979) provides preliminary evidence that marihuana-smoking women exhibit an abnormal proportion of menstrual cycles that are anovulatory or have an inadequate luteal phase. Further to this, there is no evidence as yet that cannabis exposure i s associated with i n f e r t i l i t y i n humans or animals (Grilly, Ferraro, & Braude, 1974; True et a l . , 1980; Wright et a l . , 1976). Human studies have not examined in any comprehensive manner the relationship of cannabis exposure to reproductive outcomes such as spontaneous abortion, s t i l l b i r t h , birthweight, or perinatal death (Adamec, 1976; Bloch, 1983). Teratogenicity in humans exposed to cannabis i s presently under study (Zukerman et a l . , 1981) as is the question of postnatal morbidity (Fried, 1980). Studies suggesting 132 marihuana exposure concurrent with alcohol, tobacco, other drugs or poor n u t r i t i o n enhances the adverse e f f e c t s of these health hazards have been reported (e.g., Benowitz & Jones, 1977; Siemens, 1980). U n t i l further research i s a v a i l a b l e , an assumption that the p r o f i l e of adverse e f f e c t s of marihuana smoking on reproductive health i s best represented by the p r o f i l e associated with tobacco smoking seems reasonable. Estimates of the prevalence of cannabis use i n Canada are derived from several large surveys of grade 7-13 students and household surveys of adults 18 years of age and older, between 1975 and 1979. The student surveys found that 20-27 percent of female and 25-36 percent of male students had used cannabis within the year of the survey, compared to 2-7 percent of female adults and 8-12 percent of male adults (Rootman, 1979; Smart & Fejer, 1975; Smart & Goodstadt, 1976; Smart et a l . , 1979). The percentage of adults (usually between 18-25 years of age) using cannabis d a i l y i s estimated at 2 percent for females and 5 percent for males. In the United States, higher prevalence of cannabis use i s seen among students (45-50 percent) and adults (10 percent) with smaller dif f e r e n c e s between male and female users (Abelson & Atkinson, 1975; Blackford, 1977; Johnston, Bachman, & O'Malley, 1979). Cannabis use i s associated with alcohol, tobacco, and other drug use. The high prevalence of cannabis use i n teenagers and young adults provide ample j u s t i f i c a t i o n f o r further study of the e f f e c t of cannabis exposure on reproductive health i n humans. 133 Diet and N u t r i t i o n The measurement of the impact of n u t r i t i o n and d i e t on reproductive performance presents problems not encountered i n studying other l i f e s t y l e f a c t o r s . Food intake i s e s s e n t i a l to s u r v i v a l ; d i e t i s an .integral component of human metabolism and the functioning of a l l c e l l s , t i s s u e s , organs and systems; and dietary requirements vary widely according to each i n d i v i d u a l ' s p h y s i c a l , p h y s i o l o g i c a l , biochemical, genetic makeup, age and his/her health status. At present, there i s no simple standardized method of determining whether a s p e c i f i c d i e t i s optimal, adequate or inadequate f o r an i n d i v i d u a l ( l e t alone a study population) and, thus, no straightforward method of determining the impact of an optimal versus adequate versus inadequate d i e t on reproductive performance. This has led to a heavy r e l i a n c e on animal and c e l l model studies which allow precise measures and manipulation of nutrients but can only provide grounds for speculation of a n u t r i t i o n a l impact i n humans. More recent c l i n i c a l studies of n u t r i t i o n a l status and fu n c t i o n a l e f f e c t s represent p o t e n t i a l f o r the future. For a l l of the reproductive outcome measures i d e n t i f i e d by th i s paper, animal and c e l l research describe a s i g n i f i c a n t r o l e f o r n u t r i t i o n — b u t i n most cases, the appropriate human studies have yet to be conducted. For thi s reason, apart from an independent section on n u t r i t i o n and i n f e r t i l i t y , a review of the a s s o c i a t i o n between n u t r i t i o n and reproductive outcome with a focus on birthweight and p e r i n a t a l mortality i s presented. Despite the considerable l i t e r a t u r e on n u t r i t i o n a l status and p l a c e n t a l function; on maternal-fetal exchange; on postnatal consequences of 134 maternal malnutrition; on s p e c i f i c nutrient d e f i c i e n c i e s ; and on the p o t e n t i a l mutagenic, teratogenic, and growth retarding e f f e c t s of malnutrition during the preconception period, these areas w i l l not be reviewed. The material which follows i s considered an ample demonstration of the d i f f i c u l t y of estimating n u t r i t i o n - r e l a t e d reproductive c a s u a l t i e s without s u f f i c i e n t knowledge of the n u t r i t i o n a l and anthropometric status of the population who do and do not experience reproductive problems and of the importance of baseline assessment. Background Information To b r i e f l y i l l u s t r a t e the complexity involved i n assessing the f u n c t i o n a l consequences of malnutrition, the following section reviews the i n t e r r e l a t i o n s h i p between n u t r i t i o n and the endocrine system Csince the importance of the r e l a t i o n s h i p between endocrine function and reproductive performance i s w e l l established) and examines some . s p e c i f i c problems and new concepts i n measuring n u t r i t i o n a l status. N u t r i t i o n and the endocrine system. The endocrine system, comprised of a. number of d i f f e r e n t glands each having a s p e c i f i c function (e.g., thyroid, adrenal medulla, adrenal cortex, gonads, parathyroids, and pancreas), adjusts and correlates the a c t i v i t i e s of the various body systems to meet the changing demands of the external and i n t e r n a l environment (in c l u d i n g reproduction, growth and development) and does t h i s by the s e c r e t i o n of hormones which regulate the metabolic processes of various c e l l s . In pregnancy, maternal, p l a c e n t a l and f e t a l endocrine function i s d i s t i n c t yet i n t e r r e l a t e d . 135 The involvement of any hormone i n the c o n t r o l of c e l l u l a r function depends on a s e r i e s of reactions, beginning with the synthesis of hormones from t h e i r precursors and ending with e x t r i n s i c or i n t r i n s i c feedback control mechanisms—each stage can be extremely complex because of the many i n t e r a c t i o n s involved, and, thus, there are many p o s s i b i l i t i e s f o r d i s r u p t i o n at various l e v e l s i n the system (Laycock & Wise, 1983). Hormones influence the synthesis and catabolism of carbohydrates, p r o t e i n and f a t i n a v a r i e t y of ways (Goodhart & S h i l s , 1980). In turn, hormonal balance can be profoundly changed by malnutrition—protein-energy malnutrition [PEM] i s a w e l l documented example. Since vitamins are required as precursors of hormones, as a c t i v a t o r s of steps i n hormonal synthesis, and as potentiators of the i n t e r a c t i o n of hormones with c e l l membranes, i t follows that an excess or d e f i c i e n c y of s p e c i f i c vitamins must induce some modification i n the q u a l i t y of hormones which they help synthesize, or some change i n the e f f e c t of those hormones on t h e i r target c e l l s . In addition, defective vitamin status may modify the capacity of the target c e l l to react to normal hormonal stimulus. (Jennings, 1970, p. 113) Most, i f not a l l , metabolic reactions also require the assistance of enzymes to f a c i l i t a t e and accelerate the reaction. The enzymes with the assistance of coenzymes (usually vitamin derivatives) or cofactors ( e l e c t r o l y t e s or trace mineral cations) s i g n i f i c a n t l y a f f e c t c e l l u l a r function l i k e the synthesis of new protein, through t h e i r c o n t r o l of the metabolic processes which regulate these functions. Many of the enzymes are highly s p e c i f i c and catalyze only one step i n a complicated metabolic pathway. In p a t h o l o g i c a l s t a t e s , such as vitamin d e f i c i e n c i e s 136 i n which e s s e n t i a l coenzymes are missing, there may be a buildup of r e a c t i o n products, or even a r e v e r s a l of the reaction. The synthesis of new p r o t e i n which i s v i t a l to successful reproduction and which requires the a c t i v a t i o n of p a r t i c u l a r gene sequences ( s t r u c t u r a l gene DNA to ribosomal RNA, messenger RNA and transfer RNA) i n the c e l l nucleus, i s dependent on both hormone and enzyme regulation, which are, i n turn, dependent on an adequate supply of n u t r i e n t s . In an excellent review of the actions and i n t e r a c t i o n s of hormones, vitamins, and minerals i n the human system, Kutsky (1981) states: The chief d i f f e r e n c e between a vitamin and a hormone seems to be the s i t e of biosynthesis, the types of organic compounds present i n vitamins as opposed to the hormones, and some of the modes of act i o n . These differences between vitamins and hormones i n e s s e n t i a l properties are small compared to t h e i r s i m i l a r i t i e s . (.P- v i ) The s i m i l a r i t y i s made obvious by the following d e f i n i t i o n s : a hormone i s defined as: a b i o l o g i c a l l y a c t i v e , organic compound,.a c o n t r o l l i n g agent e s s e n t i a l f or normal health and growth ( i t s absence causing a d e f i c i e n c y disease or d i s o r d e r ) , synthesized within the human  organism i n ductless glands which release the agent i n very small concentrations into the c i r c u l a t o r y system to act on target organs or t i s s u e s . (Kutsky, 1981) and a vitamin i s defined as: a b i o l o g i c a l l y a c t i v e , organic compound, a c o n t r o l l i n g agent e s s e n t i a l f o r normal health and growth i n humans ( i t s absence causing a d e f i c i e n c y disease or d i s o r d e r ) , not synthesized  w i t h i n the organism, a v a i l a b l e i n the d i e t i n small amounts, and c a r r i e d i n the c i r c u l a t o r y system i n small concentrations ..to. act on target organs or t i s s u e s . (Kutsky, 1981) 137 The d i f f e r e n c e s between c e r t a i n hormones and vitamins are becoming more d i f f i c u l t to d i f f e r e n t i a t e given the recent evidence that small amounts of Vitamin D and n i a c i n can be synthesized, that some steroid s are a c t i v e i n a dietary form, and that the same molecule (e.g., Vitamin C) functions as eit h e r a hormone or vitamin depending on the species. Both vitamins and hormones can be divided into a f a t - s o l u b l e and water-soluble s e r i e s . The f a t - s o l u b l e vitamins [ADEK] and hormones (steroids) are s i m i l a r i n f u n c t i o n — b o t h groups a f f e c t the permeability of c e l l membranes, the redox p o t e n t i a l , the a c t i v a t i o n of enzymes—but i n a d dition, the hormones a f f e c t RNA t r a n s c r i p t i o n i n the c e l l nucleus. The water-soluble vitamins (B-complex, C) and hormones ( p i t u i t a r y , thyroid-pancreas-ovary, adrenal, hypothalamic r e l e a s i n g factors) share the common t r a i t s of enzymes a c t i v a t i o n or action (vitamins d i r e c t l y , hormones i n d i r e c t l y v i a c y c l i c AMP), and a f f e c t on the c e l l n u c l e u s — but i n addition, some of these vitamins have redox p o t e n t i a l . Trace elements and mineral cofactors are required for the function of some hormones and of most coenzyme systems. The i n t e r a c t i o n , antagonism and synergism among these vitamins, hormones, and minerals i s w e l l recognized. Further evidence regarding the r o l e of s p e c i f i c amino acids, e s s e n t i a l f a t t y acids, and the l o c a l t i s s u e regulators (vascular, neurotransmitter, mitogenic agents) i s l i k e l y to c l a r i f y and enlarge t h i s p i c t u r e of interdependence. Given these complex.interactions and the pattern of intimate and co n t r o l l e d balance, i t i s small wonder that e f f o r t s to measure the i s o l a t e d influence of one component are les s successful i n human studies 138 where components cannot be knowingly manipulated. The i n s i g h t gained from animal studies and tissue cultures confirms that s i g n i f i c a n t change occurs at the c e l l u l a r l e v e l i n the absence of, or p r i o r to, overt c l i n i c a l measures and that the system i s most vulnerable to i n s u l t at times of p h y s i o l o g i c a l s t r e s s — p a r t i c u l a r l y during the maturation period of the sperm and ovum, and during embryogenesis. To date, abnormalities i n metabolic performance i n humans can only be measured at the extremes of the continuum. Thus, although the pattern of adaptive i n t e r a c t i o n within and between systems has been recognized f o r a long time, the d i s t i n c t i o n between adaptative versus optimal performance and the relevance of adaptive change at the c e l l u l a r l e v e l has yet to be c l a r i f i e d . Further to t h i s point, the following section d e t a i l s some of the basic methodology involved and problems encountered i n determining the n u t r i t i o n a l . s t a t u s of i n d i v i d u a l s and groups. N u t r i t i o n a l assessment. Although the f i r s t p r i n c i p l e s f o r the studies on the n u t r i t i o n of populations were documented i n the mid-19303 and for standardization of survey methodology i n the early 1960s, the focus of concern was the n u t r i t i o n a l problems of the populations of underdeveloped nations. I t was not u n t i l the l a t e 1960s and early 1970s that re c o g n i t i o n was given to the f a c t that malnutrition was also a health problem i n developed countries. Population surveys (e.g., Ten-State N u t r i t i o n Surveys, HANES survey, N u t r i t i o n Canada) i n the US and Canada revealed vulnerable groups and i n d i v i d u a l s with evidence of c l i n i c a l and s u b c l i n i c a l m alnutrition and linked undernutrition with 139 growth cessation and developmental handicaps, poor outcomes of pregnancy, s u s c e p t i b i l i t y to i n f e c t i o u s diseases, delayed recovery from i l l n e s s and shortened l i f e expectancy (Simko, Cowell, & G i l b r i d e , 1984). Since t h i s discovery, there have been increased e f f o r t s to apply and improve methods of n u t r i t i o n assessment for s u r v e i l l a n c e and monitoring (mandated by US Congress, 1977) and for n u t r i t i o n a l i n t ervention and preventive health care (Simopoulos, 1982). Common procedures for i n d i v i d u a l n u t r i t i o n a l assessment i n a c l i n i c a l s e t t i n g include anthropometric measurements for s e r i a l readings, q u a l i t a t i v e and q u a n t i t a t i v e dietary evaluation, observations of p h y s i c a l signs and symptoms, review of medical and socioeconomic f a c t o r s , and s e l e c t i v e biochemical t e s t s . Although n u t r i t i o n screening and diagnosis i s the exception rather than the r u l e i n antenatal care, t h i s i s not for lack of established methodology (NAS [The National Research C o u n c i l ] , 1978; Simopoulos, 1982). Techniques for community surveys or population studies, on the other hand, are usually streamlined ( i . e . , 24-hour d i e t r e c a l l as compared to m u l t i p l e 7-day prospective d i e t h i s t o r y ) i n accord with cost, convenience, and concern for minimizing invasiveness. However, the s e l e c t i o n of screening measures that are convenient, yet adequately s e n s i t i v e and s p e c i f i c , i s f a r from straightforward. Regardless of choice, these findings must be correlated with c l i n i c a l , anthropometric, d i e t a r y , biochemical, and e c o l o g i c a l factors to ensure accuracy. The report of a recent conference (Simopoulos, 1982) on assessment of n u t r i t i o n a l status r e i t e r a t e s the d i f f i c u l t i e s of assessment i n 140 epidemiologic studies and surveys of populations. P r i m a r i l y the d i f f i c u l t i e s a r i s e from the complexity of i n t e r a c t i o n s among dieta r y , and other environmental factors with genetic, personal, and other s o c i a l v a r i a b l e s that influence the occurrence of the various conditions of concern. R e a l i s t i c a l l y , these d i f f i c u l t i e s are u n l i k e l y to be s a t i s f a c t o r i l y overcome. P r a c t i c a l l y , the associations of importance for epidemiological study should be c l e a r l y i d e n t i f i e d by p r i o r c l i n i c a l and applied p u b l i c health research. More o p t i m i s t i c evaluation of recent advances i n food consumption methodology and systems (Hegsted, 1982; Nesheim, 1982; Schucker, 1982; Schultz, 1982; Schwerin et a l . , 1982) suggest these techniques could be used to i n v e s t i g a t e r e l a t i o n s h i p s between food consumption and n u t r i t i o n a l status of a representative or target population integrated with measures of health status i n d i c a t o r s , morbidity, mortality, regional economic estimate of food a v a i l a b i l i t y , and other c h a r a c t e r i s t i c s of the population (e.g., health habits and nutrient composition of the food supply). Such systems have been developed i n the context of applied p u b l i c health research. M a l n u t r i t i o n . J e l l i f f e (1966) i d e n t i f i e d four types of malnutrition: (a) undernutrition r e s u l t i n g from lack of s u f f i c i e n t food over a period of time; (b) o v e r n u t r i t i o n caused by an excess of food over time; (c) s p e c i f i c d e f i c i e n c y states r e s u l t i n g from a lack of i n d i v i d u a l nutrients; and (d) imbalance caused by a disproportionate amount of required n u t r i e n t s , either through d i e t or supplementation. Primary ma l n u t r i t i o n 141 r e f e r s to inadequacies and imbalances i n the d i e t , while secondary malnutrition i s the r e s u l t of disease and d i s a b i l i t y . F a i l u r e i n n u t r i t i o n a l health has been categorized by inadequate intake, inadequate absorption, defective u t i l i z a t i o n , increased losses (excretion), and increased requirements (Wellman, 1978). The e f f e c t s of these conditions present c l i n i c a l l y as weight l o s s , delayed wound healing, e l e c t r o l y t e and f l u i d imbalance, depressed c e l l u l a r immunity, progressive weakness, s k i n breakdown, and endocrine abnormalities. Increased understanding of the c o r r e l a t i o n between n u t r i t i o n a l assessment, intervention, and outcome parameters has resulted from increased i n t e r e s t within the c l i n i c a l and h o s p i t a l s e t t i n g — a n i n t e r e s t stemming i n part from the "discovery" by the medical profession of h o s p i t a l m a l n u t r i t i o n (Butterworth, 1974; Butterworth & Blackburn, 1975) , the advent of hyperalimentation, the introduction of computer-assis t e d n u t r i t i o n a l assessment support, and advanced research on the ef f e c t s of n u t r i t i o n a l status on f u n c t i o n a l states. One of the r e s u l t s relevant to community-based studies i s the recognition of a p a r t i c u l a r malnutrition syndrome, described by Blackburn and B i s t r i a n (1977) as "a v i s c e r a l a t t r i t i o n state or kwashiorkor-like syndrome," where a patient who appears well-nourished and has maintained normal anthropometric measurements, has depressed serum l e v e l s of albumin, t r a n s f e r r i n and other c i r c u l a t i n g proteins, as well as compromised c e l l u l a r immunity. Recommended therapy includes protein and p r o t e i n -sparing regimes such as "more than adequate" c a l o r i e s , f l u i d , e l e c t r o l y t e s , 142 and vitamins and minerals. I t i s l i k e l y that the methods of assessment used at the community l e v e l are not s e n s i t i v e enough to i d e n t i f y t h i s type of syndrome although studies have not been ca r r i e d out to c l a r i f y t h i s . A second r e s u l t of c l i n i c a l studies which has relevance to the community i s the recognition that a consistent r e l a t i o n s h i p between the parameters of n u t r i t i o n a l assessment and prognosis i s not always seen. In a h o s p i t a l s e t t i n g t h i s can be p a r t l y explained by the non-n u t r i t i o n a l aspects of the i l l n e s s and treatment. But Russell and Jeejeebhoy (1983) suggest that another reason for the discrepancy may be that the f u n c t i o n a l aspects of malnutrition are often not manifested by s i g n i f i c a n t changes i n t r a d i t i o n a l measurements of body composition. Morbidity and mor t a l i t y , they suggest, may c o r r e l a t e better with the adverse e f f e c t s of malnutrition on organ function than, with the changes i n body composition which occur much l a t e r . The authors review the e f f e c t s of maln u t r i t i o n on three v i t a l organ f u n c t i o n s — h e p a t i c secretory protein function, immunocompetence,:. and s k e l e t a l muscle f u n c t i o n — b y way of explanation (Russell & Jeejeebhoy, 1983). This review would suggest that screening f o r a " v i s c e r a l a t t r i t i o n " state of malnutrition might appropriately incorporate the measurement of serum l e v e l s of pre-albumin and r e t i n o l - b i n d i n g protein (which respond to short term changes i n pr o t e i n and energy intake) i n combination with some form of di e t a r y intake (to help d i s t i n g u i s h between dietary inadequacy and disease processes). 143 Since malnutrition i s the commonest cause of secondary immunodeficiency and i n f e c t i o n i s one of the most frequent complications of undernutrition, an a d d i t i o n a l support to screening could come from the measurement of immune function. Cell-mediated immune response i s affected e a r l i e r and more severely by undernutrition (Chandra, 1979 ), but severe undernutrition also a l t e r s serum immunoglobulins (Chandra, 1979b, 1981; McFarlane & Hamid, 1973). In f a c t , almost a l l facets of host resistance are known to be affected (e.g., the complement system, opsonic function of plasma, polymorphonuclear c e l l function) by malnutrition. Selected nutrient d e f i c i e n c i e s can also a l t e r immune f u n c t i o n — t h e s e include zinc (Driezen, 1978), i r o n and magnesium (Chandra & Dayton, 1982), pyridoxine, f o l i c a c i d , Vitamin A and Vitamin E ( B i e s e l , 1981). In a t r e a t i s e on the immunology of n u t r i t i o n a l disorders, Chandra (1980) states that n u t r i t i o n a l modulation of immunity has wide-ranging c l i n i c a l , b i o l o g i c a l and therapeutic implications. The f a c t that immunological changes occur i n advance of the " v i s i b l e " chain-reaction outcomes of poor reproductive health signals a p o t e n t i a l use for them as a screening measure (Chandra, 1980; Metcoff, 1977), although accurate i n t e r p r e t a t i o n of these measures s t i l l present d i f f i c u l t y (Harper & Simopoulos, 1982). S i m i l a r l y , recent studies of the e f f e c t s of n u t r i t i o n a l status on autonomic and sympathetic nervous system function, which may precede some of the changes i n other organ function that accompany a l t e r a t i o n s i n n u t r i t i o n a l status and which can, i n part, be assessed by norepinephrine turnover rates, speak to the future a v a i l a b i l i t y of more s e n s i t i v e 144 screening measures of malnutrition than those of current and past use. A t h i r d aspect of c l i n i c a l studies relevant to the community and reproductive outcomes i s the r e l a t i o n s h i p between n u t r i t i o n a l status and behavior. I t i s suggested that behavioral consequences may be among the most important consequences of inappropriate dietary p r a c t i c e s (Harper & Simopoulos, 1982), but that these require measurement of s p e c i f i c functions (e.g., memory, problem-solving) rather than the global measures (e.g., cognitive function) used previously i n followup studies of infants and c h i l d r e n . One study i n p a r t i c u l a r has provided us with a benchmark for the e f f e c t of severe, acute malnutrition on reproductive performance i n a human population. Through a d e t a i l e d i n v e s t i g a t i o n of the consequences of undernutrition for 40,000 c h i l d r e n conceived and born during the Dutch "hunger winter" of 1944/45, Stein and colleagues (1975) provide clear evidence of r e l a t i o n s h i p s between severe food shortage, birthweight, and other reproductive casualties when famine i s abruptly imposed on an adequately nourished population. Under these conditions, m a l n u t r i t i o n was associated with a f e t a l growth d e f i c i t of 300-400 g, excess prematurity (defined as low birthweight), excess p e r i n a t a l and neonatal m o r t a l i t y , and an excess of congenital abnormalities. Similar but l e s s d e t a i l e d accounts of epidemics of malformations, low birthweight, and i n f a n t death have been associated with acute war-r e l a t e d famine conditions since the siege of Paris i n 1870 (Wynn & Wynn, 1979). These same adverse outcomes have been c o n s i s t e n t l y 145 demonstrated across animal species with a broad spectrum of energy d e f i c i e n t , s i n g l e and multiple nutrient d e f i c i e n c y conditions (Hurley & Eckhert, 1981; Hurley, Keen, & Lonnerdal, 1983). The unique strength of the Dutch study conducted by Stein and colleagues l i e s i n the q u a l i t y of data (events surrounding the Dutch famine were c a r e f u l l y and systematically documented as part of populations s t a t i s t i c s , food r a t i o n i n g records, b i r t h and death r e g i s t r a t i o n , maternity h o s p i t a l records, e t c . ) ; the f a c t that t h i s cohort data represents the t o t a l rather than a sample population; and the f a c t that famine conditions existed i n a defined area of the Netherlands for a s p e c i f i c and measurable time period. These data allowed the researchers to compare reproductive outcomes i n famine areas against those i n non-famine area of the Netherlands during the same time period, to compare famine e f f e c t s during early versus l a t e pregnancy, and to compare outcomes i n the famine area before, during and a f t e r the famine occurred. Followup of the male cohort at the time of t h e i r m i l i t a r y medical examination was possible for a l l but 3 percent of the s u r v i v o r s . The study c l e a r l y shows that maternal n u t r i t i o n can adversely a f f e c t a l l measures of reproductive performance, from the a b i l i t y to conceive and maintain a pregnancy to the number of malformations, s t i l l b i r t h s , and i n f a n t deaths. I t also determined that maternal n u t r i t i o n a f f e c t s maternal weight, p l a c e n t a l weight, and f e t a l growth dimensions seemingly i n a time-ordered chain of events. Thus, with reinstatement of food supplies, maternal weight gain and placental weight gain preceded increases i n infant birthweight. 146 The data suggest a threshold phenomenon which r e f l e c t s a s h i f t from adaptive to maladaptive state below a c e r t a i n l e v e l of c a l o r i c intake. For example, below a threshold value of food r a t i o n s , f e r t i l i t y (and infecundity by inference) i n the population declined p a r a l l e l with the a v a i l a b i l i t y of food and by v i r t u e of a s o c i a l class gradient e f f e c t , affected the lowest classes to the greatest degree. The data did not, however, support the presence of famine e f f e c t s i n those who survived to adulthood. Some observations of the Dutch famine help to d i f f e r e n t i a t e between the e f f e c t of malnutrition around conception and i n early pregnancy from the e f f e c t of malnutrition l a t e r i n pregnancy. For example, food shortage i n the t h i r d trimester had the greatest impact on i n t r a u t e r i n e growth and early postnatal m o r t a l i t y (7-90 days). P a r t i a l regression c o e f f i c i e n t s i n d i c a t e a change i n d a i l y average rations of 100 c a l o r i e s predicts a change of 1.2 deaths/1,000 l i v e b i r t h s at age 7-89 days. Birthweight was more s e n s i t i v e to n u t r i t i o n a l e f f e c t s than length and head circumference. Slowed f e t a l growth was the mediating factor between mal n u t r i t i o n and excess deaths i n the f i r s t three months of l i f e and affected a s u b s t a n t i a l number of i n f a n t s . Food shortage i n the f i r s t trimester had the greatest e f f e c t on abnormal development of the c e n t r a l nervous system, preterm b i r t h associated with very low birthweight, s t i l l b i r t h , and f i r s t week death. Smaller numbers of infants were affected by f i r s t trimester famine. The data suggest a d i f f e r e n t low birthweight syndrome i s associated with t h i r d versus f i r s t trimester n u t r i t i o n deprivation. 147 Susser (1981) has suggested the l a t t e r a s s o c i a t i o n may have to do with a famine-refeeding combination. Wynn and Wynn (1981) reviewed the same data i n an i n v e s t i g a t i o n of the e f f e c t of food shortage around the time of conception and present a d i f f e r e n t perspective. They report that the peak incidence of p e r i n a t a l m ortality ( s t i l l b i r t h s and deaths from "prematurity") and infant m o rtality from malformations of the c e n t r a l nervous system and from "other" malformations, occurred when famine conditions coincided with the period around conception. Further to t h i s , they found that excess deaths ( p a r t i c u l a r l y those associated with malformations) pe r s i s t e d at famine l e v e l s among babies conceived during the f i r s t four months following the r e s t o r a t i o n of food supplies and remained above pre-famine l e v e l s for up to 12 months. Although f e r t i l i t y rates returned to pre-famine l e v e l s and above immediately following food r e s t o r a t i o n , there seemed to be a time lag of at l e a s t four months before recovery from the e f f e c t s of malnutrition on other reproductive outcomes was complete. I t i s speculated that t h i s f i n d i n g may r e f l e c t the s e n s i t i v i t y of the maturing ovum and sperm to i n s u l t , as i d e n t i f i e d by current mutagenic research (Wynn & Wynn, 1984). I n f e r t i l i t y and Impaired Fecundity In the Dutch famine c i t i e s , the r e l a t i o n s h i p between food rations and f e r t i l i t y (number of b i r t h s ) was convincing below a threshold of 1,500 c a l o r i e s (r = 0.92) (Stein & Susser, 1978). Amenorrhea was commonly experienced but not q u a n t i f i e d . Non-manual classes were les s e f f e c t e d than manual—the s o c i a l c l a s s v a r i a t i o n i s suggested by the 148 author to represent a greater access to the l i m i t e d food supplies i n the higher socioeconomic population. The work of F r i s c h (1977) and others (Crisp, 1979; F r i e s , 1974; Holmberger & Nylander, 1971; N i l l i u s , 1978) with small population groups, which indicated that onset and maintenance of regular menstrual function i n the human female i s dependent on a minimum weight for height ( r e f l e c t i n g a p a r t i c u l a r body composition of r e l a t i v e fatness) has led to the suggestion that 47 kg i s a " c r i t i c a l weight" below which menarche does not occur i n the North American population. Similar c r i t i c a l weight ranges have been suggested for other countries (Wynn & Wynn, 1981). However, Garn and LaVelle (1983) investigated the " c r i t i c a l weight" theory i n a pooled sample of 79,000 North American females (from four d i f f e r e n t surveys) which included 3,549 women at or below 47 kg. The authors concluded that low body weight may represent a delaying f a c t o r to menarche and e f f e c t reproductive outcome i n terms of birthweight, but because menarche, conception and m u l t i p a r i t y occurred within the low weight group, i t could not be considered " c r i t i c a l " . They observed that n o n - n u t r i t i o n a l genetic components play a r o l e , along with n u t r i t i o n a l status, i n the timing of menarche. Others (Wynn & Wynn, 1983) have suggested that ponderal index (as a p e r c e n t i l e grid) may provide a more appropriate screening t o o l i n i d e n t i f y i n g those approaching the " i n f e r t i l i t y threshold" but t h i s has yet to be studied i n a representative population. The findings of Garn and LaVelle do not contradict studies of 149 malabsorption, c a l o r i c deprivation or chronic malnutrition r e l a t e d to delayed menarche or prolonged postpartum amenorrhea, and acute malnutrition r e l a t e d to the i n t e r r u p t i o n of menstrual cycles ( E l l i s o n , 1981; Garn, 1979). The e f f e c t s of severe malnutrition on fecundity are unequivocal and include disturbed endocrine function i n both adult males and females; delayed puberty i n the young; cessation of menstruation and ovulation; underdeveloped g e n i t a l i a and atrophy of the seminiferous tubules i n young boys; s i g n i f i c a n t reduction i n semen volume, sperm count and mo b i l i t y , and l i b i d o ; and morphologic changes i n gonadal ti s s u e (Calloway, 1983; Rechcigal, 1981). The recovery period a f t e r refeeding, which i s c o n s i s t e n t l y reported as four to s i x months f o r adult males (Jacobs, 1948; K l a t s k i n , S a l t i e r , & Humm, 1947), i s l i k e l y to be between two to four months for adult females. The evidence that moderate, chronic malnutrition a f f e c t s fecundity i s more equivocal, but s i m i l a r changes have been reported i n c o n t r o l l e d experiments on adult males (Crisp et a l . , 1982; Keys et a l . , 1950; Smith et a l . , 1975) and studies of anorexia nervosa patients (Crisp, 1979; F r i e s , 1974) with weight los s of 10-15 percent below i n i t i a l normal weight. Hypothalamic, p i t u i t a r y and Leydig c e l l function a l l are involved i n the response to chronic undernutrition. Hormonal c h a r a c t e r i s t i c s return toward normal with r e s t o r a t i o n of body weight but f u l l recovery may take an extended time (Crisp et a l . , 1982; Woolam, 1981). 150 Hormonal l e v e l s are known to be very responsive to acute food deprivation. In men of normal weight, and even i n obese men, f a s t i n g for only a few days leads to reduced serum l e v e l s of FSH and testosterone, decreased excretion of 17-ketosteroids, and decreased responsiveness to LHRH (Klibanski et a l . , 1981; M i l l e r , Mickelsen, & Keys, 1948). Much has been reported about the prevalence of amenorrhea and i n f e r t i l i t y i n women following or concurrent with self-imposed weight loss (Bergh, N i l l i u s , & Wide, 1978; Crisp, 1979; F r i e s , 1974; F r i s c h , 1977; Hirvonen, 1979; N i l l i u s , 1978). Bates, Bates, and Whitworth (1982) found, i n a small c l i n i c a l study of 47 women with unexplained i n f e r t i l i t y or menstrual dysfunction who practised weight c o n t r o l , that even 5-10 percent reductions i n weight were associated with subtle a l t e r a t i o n s i n the menstrual cycle and reproductive f a i l u r e . Of the 36 study women who agreed to follow a d i e t designed to increase t h e i r weight to within the " i d e a l " range (Metropolitan L i f e Standards, 1968), 73 percent of those who were i n f e r t i l e conceived spontaneously and 90 percent with secondary amenorrhea resumed menstruation when weight was restored to with i n 5 percent of i d e a l . Differences i n the serum gonadotrophin l u t e i n i z i n g hormone:follicle stimulating hormone r a t i o [LH:FSH] were found to be s i g n i f i c a n t l y r e l a t e d to differences i n the percentage of i d e a l body weight. There i s also an ass o c i a t i o n between impaired fecundity and q u a l i t y of d i e t or s p e c i f i c nutrient d e f i c i e n c i e s (as compared to general undernutrition or energy deprivation r e f l e c t e d i n weight l o s s ) . Animal models in d i c a t e important and s p e c i f i c roles for Vitamin A and E, each 151 of the B-complex vitamins and for e s s e n t i a l f a t t y acids. In a review of n u t r i t i o n and reproductive function i n males, Calloway (1983) concludes: Of the e s s e n t i a l n u t r i e n t s , only zinc has been linked unequivocally to gonadal development and function i n men . . . [but] almost a l l nutrients involved i n metabolism can a f f e c t gonadal function d i r e c t l y , or i n d i r e c t l y v i a the pituitary-hypothalamic axis. Several trace elements are toxic to the gonads, as are excess Vitamin A and ethanol. (p. 377) Single nutrient deficiency studies i n human populations are v i r t u a l l y impossible (as i n the case of z i n c , the opportunity for natural experiments occasionally occur) since inadequate d i e t s w i l l be d e f i c i e n t i n more than one e s s e n t i a l nutrient and standards of ethics no longer allow the imposition of a n u t r i e n t d e f i c i e n c y . I t i s of i n t e r e s t that when Nelson and colleagues (1951, 1954), i n one of a s e r i e s of studies on n u t r i t i o n and reproduction i n r a t s , used hormone i n j e c t i o n s to treat d i e t - r e l a t e d endocrine dysfunction, the hormones s u c c e s s f u l l y supported conception and o f f s e t most of the f e t a l e f f e c t s r e l a t e d to the dysfunction (e.g., embryo resorption) but did so at the expense of the mother's health and n u t r i t i o n a l status. The relevance of t h i s study to humans i s unknown. Given that n u t r i t i o n a l assessment i s not a common component of current i n f e r t i l i t y i n v e s t i g a t i o n and that there i s a s u b s t a n t i a l body of knowledge l i n k i n g n u t r i t i o n to the i n t e g r i t y of the endocrine system which i s believed pertinent to mammals and humans (Campbell, 1981), i n v e s t i g a t i o n of the n u t r i t i o n a l status of some couples r e c e i v i n g medical attention for i n f e r t i l i t y problems seems warranted. U n t i l t h i s occurs i t i s not possible to determine the incidence of n u t r i t i o n - r e l a t e d i n f e r t i l i t y . 152 At best one could only speculate from population estimates of the number of men and women who are anorexic, bulemic, chronic d i e t e r s , s u b s t a n t i a l l y underweight, or have n u t r i e n t - r e l a t e d diseases or d e f i c i e n c i e s . Birthweight, P e r i n a t a l M o r t a l i t y and F e t a l Growth Studies of groups of pregnant women have con s i s t e n t l y shown that an increase i n weight gain during pregnancy i s associated with a p a r a l l e l increase i n birthweight and a progressive decrease i n the number of infants under 2,500 g. Increased prepregnancy weight i s also associated with increased birthweight and reduced incidence of low birthweight infants (Gormican, Valentine, & Satter, 1980; Naeye, 1981; Niswander & Jackson, 1975; Ounsted & Scott, 1981; Peckham & Christianson, 1971; Simpson, Lanlow, & Mitchel, 1975). One of the largest s e r i e s to confirm these c o r r e l a t i o n s was c a r r i e d out by Eastman and Jackson (1968) for f u l l - t e r m pregnancies of over 11,000 pregnant women (6,675 white, 5,236 black women) d e l i v e r i n g at the Johns Hopkins H o s p i t a l , Baltimore. Cases associated with f e t a l deaths, multiple pregnancies, toxemias, and maternal disease states were not included i n the study. Weight gain and prepregnancy weight are found to act independently of each other. When they vary i n the same d i r e c t i o n , t h e i r e f f e c t s are add i t i v e ; when they vary i n opposite d i r e c t i o n s , t h e i r separate e f f e c t s tend to o f f s e t each other. Thus, the lar g e s t infants are born to women whose pregravid weight and pregnancy weight gain i s high, the smallest to women whose pregravid weight and weight gain were both low, and infants of intermediate weight are born to women with either a low pregravid weight or pregnancy weight gain. In the Baltimore study, for example, 153 no low birthweight term infants were born to women (black or white) of high maternal weight with adequate pregnancy gain whereas the percentage of low birthweight term infants associated with low maternal weight and pregnancy gain was 16 percent i n the group of black women and 5.8 percent i i n the group of white women. Other studies also report an increased number of preterm b i r t h s (X4) associated with low weight gain ( M i l l e r & M e r r i t t , 1979). From a n u t r i t i o n a l perspective, i t has been i n f e r r e d that women with both an adequate pregravid weight for height and an adequate pregnancy weight gain (11-12 kg), enter t h e i r pregnancy with adequate maternal nutr i e n t stores and maintain a l e v e l . o f n u t r i e n t intake to support pregnancy needs. Tompkins, Wiehl, and M i t c h e l l (1955) were some of the e a r l i e s t i n v e s t i g a t o r s to report that underweight women who gained more than an average amount from mid-pregnancy had fewer low birthweight i n f a n t s . This has been further substantiated (Brown et a l . , 1981; Higgins, 1976; Simpson, Lanlow, & Mitchel, 1975). Postpartum weight as a percentage of i d e a l body weight/height i s considered to be a r e f l e c t i o n of the n u t r i t i o n a l status of a woman at the conclusion of the gestat i o n a l period. Luke and Rosso (1978) studied the r e l a t i o n s h i p between postpartum weight and optimal i n f a n t birthweight i n 254 Black and Hispanic singleton, term d e l i v e r i e s and found a l i n e a r c o r r e l a t i o n up to 110 percent i d e a l weight/height, at which point the curve begins to l e v e l o f f , reaching a plateau at about 125 percent. To achieve a postpartum weight of at l e a s t 110 percent meant a minimum weight gain of 20 kg for underweight gravida i n the study group, but was reached by the obese group i n s p i t e of some weight l o s s . 154 In a group of 467 term, singleton d e l i v e r i e s whose low-income mothers had been re f e r r e d f or dietary counselling, Rosso and Cramoy (1977) examined the influence of maternal weight gain on the incidence of f e t a l growth retardation. Based on pregravid weight/height, the mothers were c l a s s i f i e d as underweight, normal, or obese. Low weight gain was c l a s s i f i e d as underweight, normal, or obese. Low weight gain was defined as 13 kg or less f or the underweight group and 7 kg for the other two groups. The incidence of i n t r a u t e r i n e growth retardation among women (with no pregnancy complications) was 2 percent i n those with adequate gain, compared to 33 percent among women with low gain. The o v e r a l l incidence of IUGR was 9.8 percent (46 cases) of which 41.3 percent (19 cases) were delivered of women with low gain alone and an ad d i t i o n a l 30.4 percent (14 cases) were from women with low gain plus one other i n f l u e n c i n g factor (e.g., preeclampsia, smoking, hypertension); 13 percent (6 cases) of IUGR occurred i n gestations of normal weight gain and no complications, and 15 percent (7 cases) a t t r i b u t e d to a path o l o g i c a l condition other than weight gain. Thus, inadequate gain, e i t h e r alone or i n combination with other f a c t o r s , was present i n 72 percent of the cases of i n t r a u t e r i n e growth retardation. In addition, studies have demonstrated that the usual excess of low birthweight infants i s not found i n teenagers or underweight women i f they a t t a i n an above-average pregnancy gain. This apparent "weight gain p r o t e c t i o n " may apply i n cases where women smoke or are under s t r e s s as well, (Garn, Hoff, & McCabe, 1979; Luke, Hawkins, & P e t r i e , 1981; Picone et a l . , 1982; Rush, 1974). From a n u t r i t i o n a l 155 perspective, these women and those who are underweight are seen to be at r i s k by v i r t u e of increased n u t r i t i o n a l requirements or of decreased u t i l i z a t i o n of nutrient intake. Body weight i s affected by both height and maternal c a l o r i c intake, height being more important except at the extremes of v a r i a t i o n i n weight-for-height. Numerous studies have demonstrated a r e l a t i o n s h i p between maternal height and the incidence of s t i l l b i r t h s and labor d i f f i c u l t i e s , birthweight and p e r i n a t a l m o r t a l i t y (Baird, 1952; Butler & Bonham, 1963; Thomson, 1959a; Thomson & B i l l e w i c z , 1963). From a n u t r i t i o n a l perspective, i t i s i n f e r r e d that women who were favored by good n u t r i t i o n and health i n childhood are l i k e l y to have reached t h e i r f u l l p o t e n t i a l for height, whereas those less favored may be stunted and more vulnerable to reproduct ive probXeins (NAS [The Na.tiona.1 Research. Council], 1970). The a s s o c i a t i o n between height-adjusted maternal weight and i n f a n t birthweight i d e n t i f i e s the r o l e of p r i o r c a l o r i c intake. An example of t h i s type of study i s that of Peckham and Christianson (1971) of 3,939 pregnant white women who were members of the Kaiser Foundation Health Plan i n C a l i f o r n i a . From t h i s group, a l l women within the height range of 64+'. 3 inches were selected and those with weights i n the lowest (n=394), highest (n = 395), and mid (n=393) 10 percent of the group were studied. The r e s u l t s are confounded by "opposite e f f e c t " d ifferences i n maternal age and p a r i t y between the l i g h t and heavy group (more under 20 years and primigravida i n the l i g h t e s t group) and by differences i n weight gain (higher i n the l i g h t e s t group). The mean birthweight of infants 156 increased with maternal weight for height (3,191.6 g, 3,388.7 g, and 3,531.8 g for the l i g h t , medium, and heavy groups, respectively) while the proportion of low birthweight decreased with increasing maternal weight for height (7.8, 4.2, and 2.3 percent). Thus, i n stepwise multiple l i n e a r regression analysis of the determinants of birthweight (c o n t r o l l e d for gestational age and f e t a l sex), prepregnant weight for height and maternal weight gain are most important (e.g., Blidner, Anderson, & S i n c l a i r , 1982; Niswander & Gordon, 1972) with other anthropometric measurements explained by prepregnant weight. Consistent c o r r e l a t i o n s between p a r t i c u l a r dietary components and weight gain, pregravid weight, or birthweight have not been shown across studies (Leader, 1983; NAS, 1970; Rosso & Cramoy, 1977). In a prospective l o n g i t u d i n a l study of 95 pregnancies of 54 women, Beal (1971) investigated the food intake of the study women monthly. Throughout pregnancy, there was a p o s i t i v e c o r r e l a t i o n of c a l o r i c intake to weight gain, but the c o e f f i c i e n t was s t a t i s t i c a l l y s i g n i f i c a n t only for c a l o r i c intake i n the second trimester. No s i g n i f i c a n t r e l a t i o n s h i p between dietary components and birthweight or b i r t h length were found. However, the author described the women as being generally w e l l nourished before and during pregnancy. Both b i r t h length and birthweight were s i g n i f i c a n t l y correlated with maternal weight and pregnancy weight gain ( t h i r d trimester only for length). A s i m i l a r c o r r e l a t i o n c o e f f i c i e n t (+.30) for c a l o r i c intakes during the l a s t h a l f of pregnancy was reported by Thomson (1959b) i n a study of the influence of maternal body s i z e and c a l o r i c intake on birthweight 157 (489 primigravidas). In t h i s study, c a l o r i c intake was d i r e c t l y r e l a t e d to birthweight, but also to s o c i a l class and to height. The authors interpreted t h e i r r e s u l t s as i n d i c a t i n g that maternal body s i z e was the antecedent factor. In a mul t i v a r i a t e analysis of birthweight i n a low-socioeconomic black population, Rush, Davis, and Susser (1972) found maternal weight gain during pregnancy correlated with c a l o r i c intake and birthweight independent of "maternal s i z e " and interpreted these r e s u l t s as i n d i c a t i n g that c a l o r i c intake determines f e t a l s i z e and birthweight. Studies with women from low-socioeconomic groups( whose n u t r i t i o n a l status i s more l i k e l y to be inadequate) over time have shown s i g n i f i c a n t c o r r e l a t i o n between food intake and birthweight. Many do not meet the standards of research design that would allow the r e s u l t s to be regarded as "evidence" but the studies are h i s t o r i c a l l y important and put perspective on the l a t e r , more rigorous i n v e s t i g a t i o n s of the e f f e c t of n u t r i t i o n on reproductive performance. An often reported n u t r i t i o n i n tervention study of the early 1940s was c a r r i e d out i n Toronto by Ebbs, Tisdall,..and Scott (1941). A 7-day prospective dietary h i s t o r y was completed by 380 women attending the prenatal c l i n i c of the Toronto General Hospital and analyzed by the s t a f f d i e t i t i a n . Those with a poor d i e t record and poor income were a l t e r n a t e l y l e f t on a poor d i e t (120 cases) or given d a i l y food supplements beginning the 5th-6th month of pregnancy to provide a good d i e t (90 cases); those with s u f f i c i e n t income were provided with 158 n u t r i t i o n counselling to improve t h e i r d i e t (170 cases). A second 7-day food record was analyzed at eight months which showed improvement i n a l l three groups but s u b s t a n t i a l l y better diets i n the supplemented and counselled group. I n i t i a l records f or the poor d i e t group showed a mean d a i l y intake of 56 g p r o t e i n and 1,627 c a l o r i e s with an increase to 62 g and 1,837 c a l o r i e s by the second review; the supplemented group began at a l e v e l of 56 g protein and 1,690 c a l o r i e s and increased to 94 g protein and 2,424 c a l o r i e s ; and the good d i e t group began with 81 g p r o t e i n and 2,206 c a l o r i e s which..increased to 92 g p r o t e i n and 2,521 c a l o r i e s . The poor d i e t and supplemented groups were s i m i l a r i n terms of socioeconomic, demographic and o b s t e t r i c h i s t o r y ; the good d i e t group contained more primips, had higher incomes, and better o b s t e t r i c a l h i s t o r i e s . Pregravid weight, height and pregnancy gain were not reported. Obstetric and nursing s t a f f were not informed of the women's d i e t group. Results showed the incidence of abortions, low birthweight, s t i l l b i r t h s , and neonatal deaths was higher i n the group on a poor d i e t . The women i n the groups with supplemented and good diets had fewer o b s t e t r i c a l complications in c l u d i n g toxemia, and had fewer d i f f i c u l t i e s during labor, d e l i v e r y , and the postpartum period (reduced i n f e c t i o n , increased a b i l i t y to nurse). The infants born to mothers on a poor prenatal d i e t were more susceptible to i n f e c t i o n s and n u t r i t i o n a l diseases f or the f i r s t s i x months and experienced a below normal growth rate even though t h e i r mean birthweight was not s i g n i f i c a n t l y d i f f e r e n t . The study has been c r i t i q u e d for 159 methodological problems—e.g., noncompliance not adequately c o n t r o l l e d , sample s i z e too small to show s t a t i s t i c a l s i g n i f i c a n c e . By today's standards some of the outcome measures lack s p e c i f i c i t y and the timing of the intervention i s l a t e (preconception or f i r s t trimester i n t e r v e n t i o n would be seen to be more appropriate) but the approach i s sound. A study of the s t i l l b i r t h rates i n England and Wales from 1928-1944 showed a sharp decline i n a l l counties (greatest i n the, poor counties) a f t e r the i n s t i t u t i o n of war-time prenatal food rations by government. Neonatal death rates and low birthweight also declined but to a les s e r degree. Since these changes occurred at a time when a l l conditions other than n u t r i t i o n had deteriorated, the r e s u l t was considered to be due to the improved n u t r i t i o n a l status of the poorer women (Baird, 1947; Sutherland, 1946). Descriptive dietary studies i n England seemed to support t h i s explanation (Cameron & Graham, 1944). Following t h i s , a serie s of studies were conducted at the Boston Lying-In H o s p i t a l (Burke, 1948; Burke et a l . , 1943; Burke, Harding, & Stuart, 1943) to determine the influence of d i e t during pregnancy on growth and development of the fetus or the course of pregnancy, labor and d e l i v e r y , or the postpartum period. The die t s of a group of 216 pregnant women attending the h o s p i t a l prenatal c l i n i c were c a r e f u l l y assessed over the l a s t two trimesters and a formula used to rate them as excellent or good (14 percent), f a i r (46 percent), poor (23 percent), or very poor (17 percent). An independent p e d i a t r i c appraisal of the infants was ca r r i e d out and then examined against the dietary r a t i n g 160 of the mother. The length and weight of the i n f a n t s , t h e i r o v e r a l l p h y s i c a l condition and osseous development were c l o s e l y associated with the protein content of the mother's d i e t , and the occurrence of toxemia was r e l a t e d to the q u a l i t y of the d i e t . A s i b l i n g study (Burke et a l . , 1949) reported that the d i e t s of the mothers entering subsequent pregnancies remained at the same l e v e l of adequacy or inadequacy. This allowed for speculation of a n u t r i t i o n a l component i n the "repeater" syndrome df reproductive outcome i n which a mother with a h i s t o r y of a previous low birthweight i n f a n t , for example, i s at greater r i s k of giving b i r t h to a second low birthweight i n f a n t . The need for more rigorous research design and l a r g e r , more representative study populations led to a s e r i e s of supplementation studies which have f a i l e d to c l a r i f y the e f f e c t of n u t r i t i o n on reproductive outcome. Some have demonstrated a clear b e n e f i c i a l e f f e c t of food or n u t r i e n t supplementation, others have been less conclusive, and some have shown no e f f e c t . In the early 1970s a number of major studies of n u t r i t i o n and pregnancy were conducted on the premise that i f inadequate d i e t . contributed to low birthweight and poor reproductive performance, n u t r i t i o n supplement/non-supplement comparison would help quantify t h i s r e l a t i o n s h i p . Three representative prospective s t u d i e s — t h e New York study of 769 singleton b i r t h s (Rush, St e i n , & Susser, 1980), the Bogota study of 413 b i r t h s (Mora, Clement, & Christiansen, 1977; Mora et a l . , 1978, 1979), the Guatemala study of over 1,500 pregnancies 161 (Habicht et a l . , 1974; Lechtig & K l e i n , 1980), and the Taiwan study of 225 women over two consecutive pregnancies (Blackwell et a l . , 1973; Wohlleb et a l . , 1983) have been c r i t i c a l l y reviewed by Lechtig and K l e i n (1981), Rush, Stein, and Susser (1980) , Stein, Susser, and Rush (1978), and Susser (1981). The New York study c l o s e l y conforms to the experimental model with random double-blind assignment of an adequate sample of high r i s k /N women, to a cont r o l or one of two beverage supplemented groups. The Bogota study used randomized assignment of malnourished f a m i l i e s to s i x prenatal and postnatal c o n t r o l and treatment groups using regular food as a prenatal supplement from the s i x t h month of pregnancy; the Guatemala study used two d i f f e r e n t l i q u i d d i e t supplements with treatment assigned by residence i n one of four small v i l l a g e s ; and the Taiwan study randomly a l l o c a t e d pregnant women to a beverage supplement or placebo group from the second or t h i r d trimester of the i n i t i a l study pregnancy through to the weaning of a subsequent b i r t h . The r e s u l t s from these major studies were mixed and emphasize the complexity of the nutrition-birthweight r e l a t i o n s h i p . The New York study showed no s i g n i f i c a n t supplement e f f e c t on birthweight except i n infants of women who smoked. Susser (1981) concludes from these studies, that among women at r i s k of producing low birthweight i n f a n t s , prenatal dietary supplementation can lead to a modest r i s e of 40-60 g i n birthweight. The degree of e f f e c t on birthweight seems to be con d i t i o n a l and depends on the n u t r i t i o n a l status of the women. In adequately nourished or heavy women, supplements are l i k e l y to show no 162 e f f e c t , whereas i n t h i n or undernourished women, the birthweight e f f e c t i s l i k e l y to be i n the upper range. Supplement e f f e c t s on maternal weight gain did not cons i s t e n t l y r e f l e c t improved f e t a l growth nor were perceived n u t r i t i o n - r e l a t e d benefits i n cognitive function or psychological test r e s u l t s n e c e s s a r i l y associated with improved f e t a l growth. A best time for supplement i n i t i a t i o n and a best n u t r i t i o n a l content f o r supplements was not f u l l y c l a r i f i e d . The l a t t e r r e f l e c t s the inherent problem of supplement studies where t o t a l d i etary intake and n u t r i t i o n a l needs are not taken into account. In the Guatemalan study, the average non-supplemented d i e t of women during pregnancy was about 1,500 kcal and 40 g protein (mostly from vegetable sources) and the average pregnancy gain was only 7 kg. The two types of supplements compared p r o t e i n - c a l o r i c versus c a l o r i c e f f e c t s . In both supplemented groups, birthweight showed a consistent a s s o c i a t i o n with the number of c a l o r i e s received during pregnancy but no r e l a t i o n s h i p to protein. The mean birthweight of infants of women rec e i v i n g more than 20,000 kc a l of supplement over t h e i r pregnancy was 3,105 g compared to 2,994 g when supplementation was less than 20,000 k c a l . The proportion of low birthweight infants i n the former groups was 9 percent compared to 19 percent i n the l a t e r . The re l a t i o n s h i p between c a l o r i c supplement and birthweight remained a f t e r c o n t r o l l i n g for maternal height, p a r i t y , or gesta t i o n a l age. Supplements were taken under supervision at the v i l l a g e c l i n i c s , but since attendance at the center was voluntary, t h i s resulted i n a wide range of supplemented intake among and with i n the groups of women. The d i f f i c u l t y i n 163 s u c c e s s f u l l y implementing the study as designed created obvious problems concerning the strength of evidence of r e s u l t s . On the other hand, the New York City study was able to implement a research design of high standard but has been widely c r i t i c i z e d i n terms of i t s n u t r i t i o n a l design. Jacobson (.1980) has suggested that the intervention was too l a t e to allow meaningful r e h a b i l i t a t i o n to occur; others have c r i t i c i z e d the pro t e i n content (which, at the higher l e v e l s , may have had detrimental effects) i n the absence of supportive c a l o r i e s (Barnes, 1980) and the vitamin-mineral balance (Hegsted, 1980).. Wynn and Wynn (1980) noted that the ponderal index of the women (20.32 +/- 2.18) was at a l e v e l c i t e d by F r i s c h (1978) as the i n f e r t i l i t y threshold, and that poor n u t r i t i o n at conception could be a component of the poor health of the study infants (e.g., o v e r a l l p e r i n a t a l m ortality rate of 65.5/1,000). Leader (1983) comments that since protein d e f i c i e n c y may not have been the n u t r i t i o n a l problem (Bowering, Lowenberg, & Morrison, 1980), "the New York Study was analogous to a f i e l d t r i a l of a new a n t i b i o t i c i n a population not s u f f e r i n g from i n f e c t i o n s s e n s i t i v e to the new drug." He continues: "With the ben e f i t of hindsight, the study r a i s e s more questions than i t answers. Whatever i t s suggested shortcomings, the study was valuable i n pointing out that there i s a need f or a balanced and i n d i v i d u a l i z e d approach to n u t r i t i o n i n pregnancy" (p. 38). There are some i n t e r e s t i n g s i m i l a r i t i e s between the findings i n some of the supplement studies and observations by St e i n and colleagues 164 (1975). For example, when food was restored post-famine, maternal weights increased before birthweights increased—some supplement programs were associated with s i g n i f i c a n t increase i n maternal gain but not i n birthweights. Male birthweights at the height of the famine were affected more than female birthweights—some supplement programs were seen to a f f e c t male f e t a l growth to a greater degree than female f e t a l growth. These observations contribute to the questions raised as to the timing and type of supplementation selected for the studies. Why the d i s p a r i t i e s i n the r e s u l t s of the major supplement studies of the 1970s? I t i s now recognized that a d d i t i o n a l nourishment for the already adequately nourished pregnant woman bestows no measurable b e n e f i t on in f a n t health and growth, nor does inappropriate dietary intervention of the malnourished. When the " n u t r i t i o n a l l y at r i s k " woman i s c o r r e c t l y i d e n t i f i e d , her n u t r i t i o n a l status, i f unaltered, becomes p r e d i c t i v e of adverse f e t a l growth outcome but i f improved, shows s i g n i f i c a n t f e t a l growth and health b e n e f i t s . The work of Wharton and h i s colleagues (Bissenden et a l . , 1981a, 1981b; Viegas et a l . , 1982a, 1982b; Wharton, 1983) i n England and Higgins (1984) i n Canada are reviewed as examples. In a serie s of studies s t a r t i n g i n 1974, Wharton, Bissenden, and colleagues (Bissenden et a l . , 1981) found evidence that undernutrition was a factor i n the aetiology of poor i n t r a u t e r i n e growth seen i n infants of many Asian mothers d e l i v e r i n g i n th e i r Birmingham h o s p i t a l . Their i n i t i a l question was a f a m i l i a r o n e — d i d low weight Asian babies 165 r e f l e c t e t h n i c i t y ( i . e . , were they simply " l i g h t " normals) or growth f a i l u r e ? A comparison of weight, length, head circumference, skin f o l d s , and arm circumference dimensions i n 28 normal European babies, 8 growth retarded European babies from medically compromised pregnancies, and 12 low weight Asian babies from normal pregnancies showed the Asian babies deviated from normal i n exactly the same way the growth retarded European babies did (Bissenden et a l . , 1981a). In addition, both l i g h t Asian babies (n = 21) and growth retarded European babies (n = 30) showed s i g n i f i c a n t l y r a i s e d cord blood plasma t r i g l y c e r i d e s — a n a l y g o u s to the r a i s e d plasma t r i g l y c e r i d e s of c h i l d r e n with marasmus—when compared to normally grown babies of e i t h e r race (n = 35) (Bissenden et a l . , 1981b). The s i m i l a r i t y i n anthropometric and biochemical variables suggested growth f a i l u r e . Following a systematic i n v e s t i g a t i o n of factors associated with growth f a i l u r e i n the Asian population, a n u t r i t i o n a l hypothesis was established based on observed food customs and weighed di e t a r y intake studies. A prospective study was designed to compare the n u t r i t i o n a l status of mothers having normally grown and l i g h t for gestational age babies. Mothers who produced l i g h t babies put on less weight and less f a t during the second trimester (t r i c e p s s k i n f o l d was the best predictor of a growth retarded infant) and had biochemical measures suggestive of poor n u t r i t i o n a l status (Viegas et a l . , 1982a). The f i n a l study i n the s e r i e s by Wharton attempted to confirm the n u t r i t i o n a l cause of growth re t a r d a t i o n through dietary i n t e r v e n t i o n . One hundred and ninety-seven women were a l t e r n a t i v e l y assigned to one 166 of three supplement programs (con t r o l = vitamins; energy plus vitamin supplement; p r o t e i n and energy plus vitamin supplement) p r i o r to t h e i r 20th week of pregnancy. When a l l mothers were considered together, the prenatal d i e t supplements provided no f e t a l growth b e n e f i t s . Within t h i s group, infants of mothers who had l a i d down f a t adequately during pregnancy (determined not to be n u t r i t i o n a l l y at ri s k ) showed no f e t a l growth benefits from the protein-energy supplement but infants of mothers who had not put on f a t adequately ( t r i c e p s s k i n f o l d increment <0.02 mm per week) showed s i g n i f i c a n t l y enhanced i n t r a u t e r i n e growth with protein-energy supplementation. The authors comment that the r e s u l t s i l l u s t r a t e a general p r i n c i p l e of therapeutics—treatment should be given only to those who have the disorder; unselective or " b l i n d " treatment of a l l patients i s not c o s t - e f f e c t i v e and may be detrimental (Viegas et a l . , 1982b; Wharton, 1983). Wharton suggests the d i f f e r e n t i a l e f f e c t of supplementation i n the- major studies of the 1970s i s l i k e l y to r e f l e c t the i n i t i a l n u t r i t i o n a l status of the study women. This point i s also emphasized by the r e s u l t s of the Montreal Diet Dispensary-Royal V i c t o r i a H o s p i t a l [MDD-RVH] c o l l a b o r a t i v e s i b l i n g study (1963-79) (Higgins et a l . , 1984). This prospective, l o n g i t u d i n a l study was designed to examine the e f f e c t , i f any, of a n u t r i t i o n assessment and int e r v e n t i o n program on pregnancy outcome among a low socioeconomic unselected population derived from the RVH public maternity c l i n i c s . P r i o r to the study i n 1963, the MDD had developed a standardized outcome-related n u t r i t i o n r e h a b i l i t a t i o n method (known as The Higgin's 167 N u t r i t i o n Intervention Method) based on 15 years of experience providing n u t r i t i o n counselling for disadvantaged women with poor o b s t e t r i c h i s t o r i e s and poor n u t r i t i o n a l status. Baseline data showed a small mean d i f f e r e n c e i n birthweight (120 g) between the public and priv a t e patients at the RVH h o s p i t a l but a large d i f f e r e n c e i n the incidence of low birthweight and p e r i n a t a l m o r t a l i t y (60% and 50% higher i n the public patient group, r e s p e c t i v e l y ) . The purpose of the Montreal study was to demonstrate a means of reducing the birthweight d i s p a r i t y between the infants of advantaged and disadvantaged women ( i . e . , break the poverty cycle) and to test the effectiveness of d i f f e r e n t components of the n u t r i t i o n i n tervention method. C r i t e r i a f o r enrollment i n the study included attendance at public c l i n i c s f o r antenatal care and d e l i v e r y at the RVH, i n i t i a t i o n of the MDD program for referred women ( a l l women from two pub l i c c l i n i c s ) during the f i r s t two trimesters of pregnancy, and singleton l i v e b i r t h at or a f t e r 28 weeks of gestation. Since multiple regression analysis had shown that only 45 percent of the variance i n infant birthweight could be explained by parameters known to influence birthweight, the study examined the e f f e c t s of the MDD program using "within mother" rather than "between mother" comparison using s i b l i n g c o n t rols. Analysis of program e f f e c t used l e a s t squares methods, with appropriate c o r r e c t i o n for infant sex and b i r t h order. The MDD treatment groups (cases) were defined as singleton, l i v e , RVH-born i n f a n t s , gestation 28 weeks or more, with the mother being enrolled i n the MDD 168 program before the 27th week of pregnancy and having been counselled by the d i e t i t i a n at le a s t 5 times. The control groups of s i b l i n g s born to the 1,139 mothers with treated pregnancies were divided into those born p r i o r to the mother r e c e i v i n g MDD treatment (Pre-MDD; n = 872) and those born a f t e r a MDD-treated pregnancy (Post-MDD; n = 326). Cases had a mean birthweight of 3,352 g, which was 129+/-30 g greater than controls Cp<0.001), had a lower p e r i n a t a l m ortality rate (7.8 compared with 14.3/1,000; p<0.01) and lower birthweight proportion (4.1 compared with 7.9 percent; p<0.01). The MDD n u t r i t i o n i n tervention program i s b u i l t around an i n i t i a l comprehensive n u t r i t i o n assessment (diagnosis) with subsequent dietary treatment t a i l o r e d to the needs of each i n d i v i d u a l mother. The f a c t that mean birthweight (3,469 g) and weight gain (13.4 kg) was highest, and low birthweight l e s s frequent (2.9 percent) f o r those assessed as having a good i n i t i a l n u t r i t i o n a l intake demonstrates the assessment method i s accurate f o r s e l e c t i n g mothers who do not need a c o r r e c t i v e dietary allowance. S i g n i f i c a n t improvements i n birthweight and i n f a n t outcomes were s p e c i f i c to those women (79 percent of to t a l ) determined to be n u t r i t i o n a l l y at r i s k . The r e s u l t s support the assumption that infants of women who are underweight pregravid, with a low rate of pregnancy weight gain, with a p r i o r poor o b s t e t r i c h i s t o r y , etc., ben e f i t i f maternal- n u t r i t i o n i s improved. However, 46 percent of the study group who were determined to be undernourished by v i r t u e of the pro t e i n content of th e i r d i e t at i n i t i a l intake (a "leader" nu t r i e n t 169 c r i t e r i o n ) a n d who were the most malnourished of the study r i s k groups (by 540-650 kc a l and 17-30 g protein at i n i t i a l assessment) were not underweight (the majority were overweight) and would most l i k e l y be missed as a high r i s k pregnancy i n studies where dietary intake i s not assessed. This group showed the most s i g n i f i c a n t birthweight increase compared to t h e i r s i b l i n g controls (161+/-39 g). That the study group was economically disadvantaged i s revealed by the f a c t that 80 percent of the cases required an income supplement (given i n the form of basic foods) to meet t h e i r n u t r i t i o n a l requirements. I t i s reasonable to suggest that the women were c h r o n i c a l l y malnourished p r i o r to intervention. I n i t i a l mean intakes ranged from 51-79 g protein and 1,800-2,400 c a l o r i e s depending on the r i s k category and were increased to 99-105 g protein and 2,600-2,950 c a l o r i e s through the counselling program. Relative to the Dutch famine study t h i s i s not an extreme l e v e l of malnutrition and yet major benefits i n terms of birthweight and infant m o r t a l i t y were r e a l i z e d through appropriate assessment and improved d i e t p r i o r to the t h i r d trimester. The need f o r a comprehensive n u t r i t i o n a l assessment (which requires time and s p e c i a l s k i l l s ) i n measuring the impact of n u t r i t i o n a l status on reproductive outcome; the recognition that adverse e f f e c t s of malnutrition can be associated with s i n g l e or multiple nutrient inadequacies, excesses or imbalance; the knowledge that n u t r i t i o n a l needs are modified by a wide range of factors (e.g., i n f e c t i o n , drug use, a c t i v i t y patterns, e t c . ) ; that nutrition-outcome r e l a t i o n s h i p s may vary at the extremes (e.g., high pregnancy gain i n the overweight 170 may r e f l e c t poor rather than good d i e t ) ; and that n u t r i t i o n can have a mediating influence on the a f f e c t of other reproductive r i s k factors (e.g., rate of d e t o x i f i c a t i o n , endocrine function, immune status) i d e n t i f y l i m i t s to the a b i l i t y to quantify the t o t a l impact of n u t r i t i o n on reproductive performance. The new generation of post-70s n u t r i t i o n studies provide a more d e t a i l e d multimeasure examination of maternal-reproductive outcome-related v a r i a b l e s i n smaller populations compared to the i n d i r e c t or l e s s s p e c i f i c measures of the previous large scale studies. There are, of course, inherent research design problems i n studying smaller populations across a larger set of v a r i a b l e s . However, they can provide increased d e t a i l which seems necessary to bridge the gap between animal studies and ti s s u e or c e l l c ulture studies and human n u t r i t i o n studies. Metcoff and colleagues (1980, 1981) i n Oklahoma, working through a s e r i e s of stages i n a complex, c a r e f u l l y executed study of over 400 "apparently" health pregnant women attending the University Obstetric G l i n i c show that birthweight can be predicted from a set of maternal c h a r a c t e r i s t i c s and n u t r i t i o n - r e l a t e d measurements