UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The relationship between learning, health beliefs, weight gain, alcohol consumption, and tobacco use.. Strychar, Irene 1988

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
UBC_1988_A2 S77.pdf [ 22.02MB ]
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0055820.json
JSON-LD: 1.0055820+ld.json
RDF/XML (Pretty): 1.0055820.xml
RDF/JSON: 1.0055820+rdf.json
Turtle: 1.0055820+rdf-turtle.txt
N-Triples: 1.0055820+rdf-ntriples.txt
Original Record: 1.0055820 +original-record.json
Full Text
1.0055820.txt
Citation
1.0055820.ris

Full Text

THE RELATIONSHIP BETWEEN LEARNING, HEALTH BELIEFS, WEIGHT GAIN, ALCOHOL CONSUMPTION, AND TOBACCO USE OF PREGNANT WOMEN by IRENE STRYCHAR B.Sc. Universite de Montreal, M.Ed. The University of Ottawa A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF EDUCATION in THE FACULTY OF GRADUATE STUDIES ADMINISTRATIVE, ADULT AND HIGHER EDUCATION We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March 1988 © IRENE STRYCHAR, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at The University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. ADMINISTRATIVE, ADULT AND HIGHER EDUCATION The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: March 1988 ABSTRACT Understanding how women learn during pregnancy is the foundation for planning prenatal education programs. To date, adult educators have not investigated, in any depth, the learning process during pregnancy. The purpose of this study was to examine learning during pregnancy and relate this learning to learning outcomes. The principal research questions were: "What are the learning patterns of pregnant women?" and "What is the relationship between learning and health behavior of pregnant women?" It is unknown whether learning during pregnancy is directly associated with behavior or mediated through health beliefs. The objectives of this research were to identify pregnant women's health behaviors, learning patterns, and health beliefs. The three health behaviors examined in this study were eating, drinking, and smoking. These behaviors were operationalized in terms of their outcomes: weight gain, alcohol consumption, and tobacco use. These factors are amenable to an education intervention and are behavioral risk factors associated with low birth weight. The process of investigating learning patterns consisted of identifying: what was learned during the pregnancy, which resources were utilized, what advice was given, what amount of time was spent in learning, who initiated the learning episodes, and what learning transaction types emerged. Determining learning transaction types was based upon an adaptation of Tough's (1979) concept of planners and Knowles's concept of self-directed learners. The process of investigating health beliefs consisted of identifying pregnant women's concerns, perceived risk, perceived use of the information, and perceived barriers, defined according to an adaptation of the Health Belief Model. ii The principal hypotheses of the study were: (1) self-initiated learning will be positively correlated with knowledge scores, (2) self-initiated learning will be positively correlated with ideal health behaviors, and (3) health beliefs will be positively correlated with ideal health behaviors: ideal weight gain during pregnancy, reduced alcohol consumption, and reduced cigarette smoking. The research, an ex post facto design, involved a one hour structured interview with women within the week following delivery of their infants in hospital. A proportional sample of 120 primigravidas was selected from seven hospitals with average number of monthly births greater than 100. Reporting of results was based upon 120 interviews conducted as part of the main sample and eight interviews conducted during the pilot study. Pilot responses were included because these responses were similar to responses provided by the main sample, with the exception of health belief data. One case was excluded from the sample, making for N = 127. Data analyses were based upon the entire sample N = 127, with the exception of health belief measures. Since alcohol and smoking health belief questions were administered to drinkers and smokers and since health belief measures related to weight gain, alcohol, and smoking were missing data, health belief analyses were based upon N=123 for weight gain, N = 88 for alcohol, and N = 43 for smoking. Women had spent an average of forty-one hours learning about weight gain, alcohol consumption, and tobacco use during pregnancy. The principal resources used were: reading materials, physicians, family members, and prenatal classes. The majority of pregnant women had engaged in other-initiated learning episodes in the one to one setting, that is with a health professional, family member, or friend. iii Self-initiated learning about weight gain was associated with higher knowledge scores and ideal prenatal weight gain (p<0.05); and, weight gain health beliefs were negatively correlated with ideal prenatal weight gain (p<0.05). Finding a negative correlation, in contrast to the predicted positive correlation, may have been due to the fact that in a retrospective study the behavior precipitated reporting of health beliefs. Other-initiated learning about alcohol was associated with higher knowledge scores and reduced alcohol intake (p<0.05); however, alcohol health beliefs were not associated with reduced alcohol intake. For smoking, neither self-initiated nor other-initiated learning was associated with knowledge scores or reduced cigarette smoking; however, a low degree of perceived risk was predictive of reduced cigarette smoking (p<0.05). Knowledge about tobacco use was positively correlated with health beliefs, suggesting that learning may be indirectly related to smoking behaviors. This study contributes to the knowledge about learning during pregnancy by providing a descriptive profile of learning patterns during pregnancy, and by examining the relationship between learning, health beliefs, and behavior. Fostering a learning environment which stimulates self-initiated learning may assist women reach ideal weight gain during pregnancy. For alcohol, encouraging health professionals, family members, and friends to initiate learning about the hazards of consuming alcohol during pregnancy seems warranted. Self-initiated learning may not be superior to other-initiated learning but may be topic specific, due to the nature of the health behaviors examined. Identification of women's smoking health beliefs seems warranted during prenatal education. Further research is required to better understand the role of learning with respect to changing smoking behaviors during pregnancy. TABLE OF CONTENTS Abstract ; ii Table of Contents v List of Tables viiList of Figures xii Acknowledgements xiiCHAPTER I THE RESEARCH PROBLEM 1 Background of the Problem 2 Learning 3 Health Behavior 6 Gravity of the Problem 7 Definition of Terms 13 Summary and Organization of the Dissertation 14 CHAPTER II REVIEW OF THE LITERATURE 5 Learning 1Learning Opportunities During Pregnancy 19 Description of Health Behaviors and Learning about Health Behaviors ... 28 Weight Gain: Behaviors and Learning 2Behaviors: Weight Gain and Dietary Practices 29 Learning about Weight Gain: Information Sources and Intervention Strategies 34 Alcohol Consumption: Behaviors and Learning 39 Behaviors: Alcohol Consumption 3Learning about Alcohol: Media Campaigns, Intervention Strategies, and Information Sources 47 Tobacco Use: Behaviors and Learning 51 Behaviors: Tobacco Use 5Learning about Tobacco: Intervention Strategies and Information Sources 9 Health Behaviors and Perinatal Outcomes 64 Weight Gain and Perinatal Outcome 5 Alcohol Consumption and Perinatal Outcome 69 Tobacco Use and Perinatal Outcome 71 Health Beliefs of Pregnant Women and Behaviors 74 Summary of Chapter II 77 CHAPTER III THE CONCEPTUAL FRAMEWORK 79 Framework of the StudyThe Learning Component 80 The Health Beliefs Component 8 The Research Questions 92 Hypothesis of the Study 3 Statement of Hypothesis 5 v Summary of Chapter III 97 CHAPTER IV METHODOLOGY 9 Research Design 100 Population and Sample 102 Data Collection Procedures 5 Interview Schedule 6 Development of the Research Instrument 109 Prepilot Study 10Pilot Study 113 Content of the Research Instrument 115 Health Behaviors of Pregnant Women 11Learning Patterns of Pregnant Women 120 Health Beliefs of Pregnant Women 126 Demographic Information 128 Newborn Information 9 Reliability of the Research Instrument 130 Data Analyses 135 Summary of Chapter IV 151 CHAPTER V SAMPLE CHARACTERISTICS 153 Data Collection 15Sample Description 6 Participation Rate 157 Demographic Characteristics 159 Health Behaviors of Pregnant Women 166 Weight Gain 16Alcohol Consumption 173 Tobacco Use 18Summary of Chapter V 191 CHAPTER VI LEARNING PATTERNS AND HEALTH BELIEFS 193 Learning Patterns 19What Had Been Learned 194 Weight Gain Knowledge Results 19Alcohol Knowledge Results ; 197 Tobacco Knowledge Results 200 Utilization of Learning Resources 2 Prenatal Class AttendanceWeight Gain Resources 205 Alcohol Resources 208 Tobacco Resources 212 Advice and Recommendations from Resources 215 Advice about Weight Gain 216 Advice about Alcohol Consumption 219 Advice about Tobacco Use 221 Time Spent in Learning 223 Time in Learning: Weight Gain 22Time in Learning: Alcohol 226 Time in Learning: Tobacco 9 vi Initiators of the Learning 232 Initiators of Weight Gain Discussions 23Initiators of Alcohol Discussions 233 Initiators of Tobacco Discussions 4 Initiators of Attendance at Prenatal Classes 23Initiators of Reading Episodes 235 Initiators of Viewing Audiovisual Productions 236 Learning Transaction Types 237 Weight Gain Learning Transaction Types 23Alcohol Learning Transaction Types 239 Tobacco Learning Transaction Types 241 Health Beliefs 243 Weight Gain Health Beliefs 244 Alcohol Health Beliefs 6 Tobacco Health Beliefs 8 Summary of Chapter VI 249 CHAPTER VII RESULTS OF HYPOTHESIS TESTING 252 Outcome of Hypotheses Tested 25Other Findings 26Weight Gain Goal 269 Alcohol Goal 271 Smoking Goal 2 Summary of Chapter VII 273 CHAPTER VIII DISCUSSION AND CONCLUSIONS 276 Summary of the Study 27Significance of the Study 8 Discussion of Hypotheses Tested and Other Findings 280 Limitations of the Study 289 Discussion of the Findings 291 Health BehaviorsWeight Gain 2 Alcohol Consumption 295 Tobacco Use 303 Learning Patterns 9 What Pregnant Women Learn 30How Pregnant Women Learn 311 Health Beliefs 327 Conclusions 32Recommendations for Theory, Practice, and Future Research 332 Recommendations for Theory 33Recommendations for PracticeRecommendations for Future Research 335 Epilogue 337 BIBLIOGRAPHY 8 APPENDICES 355 vii List of Tables 1. Tough's Typology of Planners in Learning 83 2. Annual and Monthly Delivery Numbers of Hospitals Participating in the Study with Corresponding Number of Research Subjects 103 3. Learning Transaction Types 125 4. Hoyt's Estimate of Reliability for the Knowledge and Health Belief Measures 132 5. Means and Standard Deviations of the Weight Gain Learning Episode Measure 140 6. Means and Standard Deviations of the Alcohol Learning Episode Measure . 144 7. Means and Standard Deviations of the Tobacco Learning Episode Measure 145 8. Participation Rate 157 9. Age Distribution of the Sample and the B.C. Population 160 10. Births by Weight Categories for the Sample and the B.C. Population 163 11. Gestational Age Distribution of the Sample and the B.C. Population 165 12. Weight Gain Patterns of Under, Normal, and Over Weight Women 168 13. Changes in Eating Habits 169 14. Sample Distribution of Weight Gain Ratio Values 171 15. Alcohol Consumption Patterns of Drinkers Before and During Pregnancy .. 174 16. Average Weekly Alcohol Consumption of Drinkers in the Sample Before Pregnancy and all Canadian Female Drinkers Participating in the Health Canada Survey 175 17. Average Weekly Number of Drinks Consumed by the Sample Before and During Pregnancy 1718. Percentage of Drinkers Consuming Various Types of Alcoholic Beverages Before and During Pregnancy 176 19. Number of Individuals Consuming Beer, Cider, Wine, and Liqueur Before and During Pregnancy and their Corresponding Number of Alcoholic Beverages Consumed on a Single Drinking Occasion 178 viii 20. Alcohol Consumption Behavior Changes During Pregnancy and Mean Numbers of Drinks Consumed per Month Prior to Pregnancy 179 21. Average Number of Cigarette Smoked Daily by Smokers in the Sample Before Pregnancy and by all Canadian Female Smokers Participating in the Health Canada Survey 184 22. Cigarette Smoking Patterns of Smokers Before and During Pregnancy 185 23. Average Number of Cigarettes Smoked Daily by the Sample Before and During Pregnancy 1824. Cigarette Smoking Behavior Changes During Pregnancy and Mean Numbers of Cigarettes Smoked per Day Prior to Pregnancy 187 25. Age and Socioeconomic Differences Among Smokers and Non-Smokers 189 26. Percentage of the Sample Providing Optimal Responses to the Weight Gain Knowledge Test Items 195 27. Information Sources Used to Answer the First Weight Gain Knowledge Test Question 196 28. Percentage of Drinkers and Non-Drinkers Providing Optimal Responses to the Alcohol Knowledge Test Items 197 29. Information Sources Used to Answer the First Alcohol Knowledge Test Question 199 30. Percentage of Smokers and Non-Smokers Providing Optimal Responses to the Tobacco Knowledge Test Items 200 31. Information Sources Used to Answer the First Tobacco Knowledge Test Question 201 32. Location Sites of Prenatal Class Attended 203 33. Resources Utilized for the Topic of Weight Gain 205 34. Types of Reading Materials and Audiovisual Productions Used for the Topic of Weight Gain 207 35. Resources Utilized for the Topic of Alcohol Consumption 209 36. Types of Reading Materials and Audiovisual Productions Used by Drinkers for the Topic of Alcohol Consumption 211 37. Resources Utilized for the Topic of Tobacco Use 3 ix 38. Types of Reading Materials and Audiovisual Productions Used by Smokers for the Topic of Tobacco Use 214 39. Summary of Weight Gain Advice from Various Resources 217 40. Summary of Alcohol Advice Given to Drinkers by Various Resources 220 41. Summary of Smoking Advice Given to Smokers by Various Resources 222 42. Percentage of Women Having Differing Number of Discussions about Weight Gain with Health Professionals, Family Members, and Friends . 224 43. Estimated Time in Learning about Weight Gain Issues by the Sample .... 225 44. Percentage of Drinkers Having Differing Number of Discussions about Alcohol with Health Professionals, Family Members, and Friends 227 45. Estimated Time in Learning about Alcohol Issues by Drinkers 228 46. Percentage of Smokers Having Differing Number of Discussions about Smoking with Health Professionals, Family Members, and Friends 229 47. Estimated Time in Learning about Tobacco Issues by Smokers 230 48. Profile of the Sample's Engagement in Self-Initiated and Other-Initiated Learning Episodes Regarding Weight Gain Issues 233 49. Profile of Drinkers' Engagement in Self-Initiated and Other-Initiated Learning Episodes Regarding Alcohol Issues 234 50. Profile of Smokers' Engagement in Self-Initiated and Other-Initiated Learning Episodes Regarding Tobacco Issues 235 51. Dominant Weight Gain Learning Transaction Types for the Sample 238 52. Dominant Alcohol Learning Transaction Types for Drinkers 240 53. Dominant Tobacco Learning Transaction Types for Smokers 242 54. Mean Scores of the Weight Gain Health Belief Components 244 55. Mean Scores of the Alcohol Health Belief Components for Drinkers 247 56. Mean Scores of the Smoking Health Belief Components for Smokers 248 57. Intercorrelations Among the Four Components of Weight Gain Health Beliefs and Ideal Weight Gain 265 58. Intercorrelations Among the Four Components of Alcohol Health Beliefs and Reduced Alcohol Intake During Pregnancy 267 x 59. Intercorrelations Among the Four Components of Smoking Health Beliefs and Reduced Cigarette Smoking During Pregnancy 268 60. Summary of Weight Gain, Alcohol, and Tobacco Hypotheses Tested 270 61. Summary of Demographic Characteristics of the Sample 279 62. Average Weekly Alcohol Consumption Levels Prior to and During Pregnancy of Women Participating in Six Research Studies 298 63. Number of Cigarettes Smoked Daily Prior to and During Pregnancy of Women Participating in Six Research Studies 305 xi List of Figures 1. The Link Between the Goals of Educators and Health Professionals 12 2. Concept of Adjusted Weight Gain Ratio 147 xii ACKNOWLEDGEMENTS I extend sincere and heart-felt thanks to Dr. William S. Griffith, my research supervisor and mentor, for his guidance, support, and encouragement. The process of completing a dissertation is facilitated by a strong research supervisor and I am fortunate to have had Dr. Griffith take me through the task. A very special thank you must be given to Dr. Bob Conry for his assistance in the design and analyses of my study. The many hours spent discussing data output and analyses issues are much appreciated. I also extend my appreciation to Dr. Nancy Schwartz for her nutrition education focus and to Dr. Tom Sork for his valuable input in developing learning transaction types. I thank Dr. James King for his constructive criticism during the development of the hypotheses of my study. To my parents, with whom I have shared the sorrow and the joys involved in writing my dissertation, I thank you for your ever encouraging words and confidence in my abilities. To my grandmother, sister and her family, I thank you all so very much for your support. There are many colleagues and friends who have helped me through my studies at UBC. I want to acknowledge, with thanks, Dorothy Fisher for her assistance in developing the ^ weight gain ratio measure and Nand Kishor for his help in setting up my data files for analyses. I also especially value the friendship of Lawrence Hilton, Elizabeth Mumba, the Kohns, Shauna Butterwick, and Jane Munro. Most of all I would like to thank the women who participated in my research and the hospital administrators and directors of nursing who made this study possible. It is the end of an era and the beginning of new challenges to come. xiii 1 CHAPTER I THE RESEARCH PROBLEM Pregnancy is a time when numerous changes affect a woman's physical, psychological, and social disposition. These changes coupled with a desire to have a healthy pregnancy outcome are powerful motivating forces for pregnant women to engage in learning. Understanding the nature and scope of learning during pregnancy is the basis upon which educators can assist pregnant women during their learning as well as have a positive effect on pregnancy outcome. There are numerous factors that influence a pregnancy outcome. Some of these include the pregnant woman's age, parity, obstetrical history, race, socioeconomic status, health status, use of the health care system, nutrition, weight gain, alcohol, smoking, and drug use (Worthington-Roberts, Vermeersch, and Williams, 1985). The complex interrelationships among these variables have not been clearly delineated in the literature. Nevertheless, several factors are amenable to intervention, and more specifically an education intervention. These variables include nutritional status, alcohol consumption, and smoking practices. Furthermore, these variables are three primary behavioral risk factors associated with low birth weight, a principal predictor of perinatal mortality (Institute of Medicine, 1985). The purpose of this study is to examine learning behaviors during pregnancy with regard to those variables that are amenable to an education intervention and are associated with a healthy pregnancy outcome. The principal research questions are "What are the learning patterns of pregnant women?" and "What is the relationship between learning and health behavior of pregnant women?" 2 In order to develop the principal research questions posed, the first section of Chapter I contains the background of the problem; the second section consists of identifying the gravity of the problem; the third section contains a definition of the terms used in this study; and, the final section contains a summary of Chapter I and an outline of this dissertation. Background of the Problem Health educators are concerned about the learning patterns and behaviors of pregnant women regarding nutritional practices, alcohol consumption, and tobacco use. What individuals choose to eat, what amounts they choose to drink, and what amounts they choose to smoke are all voluntary decisions and thus may be amenable to an education intervention. In some cases poverty or other environmental factors may restrict options open to pregnant women but these might also be regarded as conditions calling for increased educational efforts. The purpose of health education is to change behavior and therefore professional health educators dealing with pregnant women strive for optimal behavior associated with healthy pregnancy outcome. The medical literature (Butler and Alberman, 1969; Metcoff et al., 1981; Naeye, 1979; Little, 1977; and Sexton and Hebel, 1984) provides evidence that poor maternal weight gain, alcohol consumption, and tobacco use are all associated with an increased incidence of low birth weight, a principal predictor of perinatal mortality (Piekkala et al., 1985). Low birth weight is defined by the World Health Organization (1984) as a birth weight of less than 2,500 grams. Low birth weight is an important determinant of the infant's chances of survival and healthy development. Because of the association of low maternal weight gain, alcohol consumption, and tobacco use with increased incidence of low birth weight, the learning that occurs 3 about weight gain, alcohol consumption, and tobacco use and its relationship with health behaviors of pregnant women is of particular interest to educators. Therefore two important components need to be examined for the purpose of clarifying the background of the research problem. Component one is learning and component two is health behavior. The problems and focus of the research on learning during pregnancy and behavior are described as follows. Learning People engage in learning for a multitude of reasons and the learning transactions undertaken by adults are complex as well as multifaceted. Every individual enters a learning activity with previous experience and prior learning which mesh with the activity undertaken. Therefore educators can never predict with total certainty how each adult will respond to new information and experiences (Brookfield, 1986). The multifaceted nature of the learning process should challenge researchers to obtain a better understanding of learning, and in this case, learning experiences during pregnancy. In an attempt to better understand the learning that occurs among adults, Tough (1979), in his work "The Adult's Learning Projects", addressed the question of how adults learn. Specifically he examined the decision and planning aspects of learning and focused on what adults learn, why they learn it, and the types of planners involved in learning. The results of research in several studies reported by Tough (1978) indicate that about 73 percent of highly deliberate attempts to learn are planned by learners themselves (self-planned learning projects), and approximately 20 percent are planned by professionals on a one-to-one basis or in a group setting. To date, much of the research reported in the medical, nursing, and nutrition literature on learning during pregnancy has been diverse and has focused on: 4 providing a descriptive profile of women attending prenatal educational opportunities; examining pregnant women's sources of information; and, describing media campaigns encouraging healthy behaviors. By utilizing Tough's framework for examining planners of the learning, attention is shifted from evaluating learning in one particular educational setting to obtaining a more global view of learning during pregnancy. At the same time, it shifts attention away from examining only professionally guided learning to examining learning that occurs in all settings. Often health professionals assume that adult learning is restricted to educational programs provided by government agencies, physicians, and other health professionals. The extent of individual learning and of learning with nonprofessionals has yet to be documented. Adult learning is a phenomenon and a process that can take place in any setting. Tough (1982) found that learning outside the institutional setting was not perceived by learners to be as useful or valuable as learning within the institutional setting. Although a great deal of value is placed on professionally guided learning, the purpose of this study is not restricted to examining learning in the institutional setting. Instead, the purpose is to examine the learning that occurs among pregnant women in any setting. The extent of group guided learning during pregnancy can be identified by examining attendance at prenatal classes. British Columbian data from Moricky (1985) indicate that approximately 70 percent of women experiencing their first pregnancy and approximately 40 percent of women experiencing their second or subsequent pregnancy attend B.C. Ministry of Health prenatal classes, an organized learning activity. The extent of individually guided learning by pregnant women has not been documented. A potential indirect measure of this type of learning could be the amount of prenatal care a woman obtains. Although no evidence exists to establish 5 the association between amount of prenatal care obtained and the extent of learning, one could surmise that learning might occur during a visit with the physician, nurse, or dietitian/nutritionist. The extent and quality of learning would vary greatly. In the United States, 1984 data from the Advance Report of Final Natality Statistics (National Centre for Health Statistics, 1986) indicate that approximately 75 percent of women begin their prenatal care during the first three months of pregnancy; approximately 17 percent of women begin their prenatal care during the fourth, fifth, and sixth month of pregnancy; approximately 4 percent of women begin their prenatal care during their seventh, eighth, and ninth month of pregnancy; approximately 2 percent of women had no prenatal care; and, approximately 2 percent of women had an unknown amount of prenatal care. Furthermore, the median number of visits made by women to obtain prenatal care was twelve. The extent of learning with nonprofessionals and the extent of learning that occurs through reading and viewing audiovisual productions during pregnancy has also not been examined or addressed. Educators and health professionals have focused their attention primarily on professionally guided learning. Although professionally guided learning during pregnancy is an important phenomenon, the relative importance of learning with nonprofessionals is not known. This study is an investigation of the learning of pregnant women from a new perspective, one that provides educators with an understanding of learning in all settings, and of how pregnant women learn. To date this type of study identifying how women learn has not been conducted with the pregnant population. In this research, Tough's view of learning and his concept of planners serves as the basis for the identification of how pregnant women learn. 6 Health Behavior The second component of the research problem focuses on health behaviors of pregnant women. Much of the research in this area has focused on describing health behaviors of pregnant women and has focused on examining the effect of intervention strategies by health professionals on behaviors of pregnant women. Intervention strategies have included smoking cessation programs for pregnant women, nutrition counselling programs for pregnant women, and alcohol programs for pregnant problem drinkers. What factors influence behavior is a question that has perplexed educators over the centuries. Personality variables, situational characteristics, and norms all play a role in determining an individual's behavior. Numerous forces interact to determine behavior and this study focuses on examining the relationship between learning and behavior of pregnant women. Experience in the health care field tells us that pregnant women not only deliberately seek information about their pregnancy but also are given information deliberately or incidentally by health professionals. Knowles (1975) assumes that individuals who self-direct or self-initiate their learning efforts will make better use of the information than individuals whose learning is initiated by others. Therefore a profile of the learning episodes of pregnant women might reveal pertinent information about learning and its relationship with health behavior. It is unknown whether learning during pregnancy is directly associated with health behaviors or whether learning is mediated through health beliefs. The Health Belief Model was developed by Hochbaum, Regies, Rosenstock, and Leventhal (Rosenstock, 1974b) to explain preventive health behaviors. Women monitoring their weight gain, and changing their alcohol and smoking practices during pregnancy would 7 be taking preventive action to increase their chances of having a healthy pregnancy outcome. The model indicates that mass media campaigns, advice from others, and newspaper or magazine articles can potentially influence beliefs (Janz and Becker, 1984). Learning during pregnancy may be mediated through health beliefs. The measurement of learning as well as health beliefs may provide useful insights towards understanding the outcomes of learning during pregnancy. Little work has been done to understand the phenomenon of learning during pregnancy, and the relationship between learning and behavior as well as between health beliefs and behavior. This phenomenon is explored in this research. Gravity of the Problem There are two aspects which underlie the gravity of the research problem. First, adult educators have not examined, in any depth, the learning process during pregnancy. Second, those studies which have examined how adults learn have not related this learning to learning outcomes. The gravity of not having adult educators examine learning during pregnancy is embedded in the fact that certain health behaviors, amenable to an education intervention, are associated with pregnancy outcome. Each of these aspects is expanded upon in order to describe the underpinnings of studying the process of how pregnant women learn and relating this learning to outcome. Understanding how women learn, specifically during pregnancy, has not been a high priority among adult educators. It has been included as an incidental activity when learning had been studied. For example, Cross and Valley (1974) examining learning of non-traditional education programs found that 31 percent of 1,893 respondents in a national probability sample in the United States were engaged in some form of adult learning. More specifically, 56 percent indicated that they would 8 be interested in learning about home and family living and 54 percent in learning about personal development. The topic of pregnancy was not mentioned as a separate category. Similarly, Waniewicz (1976) in his study of part-time learners in Ontario had also not included pregnancy as a subject area. More recently, Devereaux (1985) reported in "The Survey of Adult Education in Canada" that 23 percent of adult learners had taken a course for "personal development/general interest." Included as one sub category of "personal development/general interest" was a catch-all division of "other" comprising marriage preparation, prenatal instruction, driver training, and first aid. Although it is recognized that a small percentage of the population are pregnant at any given time, it is an important period of learning, one that can potentially influence the next generation. Brookfield (1984) has been critical of the research studies which have examined how adults learn, specifically self-planned learning. He has cited lack of attention to the quality of learning as a serious omission. Tough (1982) in his more recent research efforts examined intentional changes among adults, and focused on identifying what changes had occurred and how that process was implemented primarily from a learning perspective. Tough attempted to evaluate whether individuals had attained their goal by asking "What percentage of your desired change did you actually achieve?" This type of questioning is a beginning step in evaluating behavior change and its relationship to learning. More objective criteria are required. Adult educators have primarily studied the process of how adults learn and have not related learning to objective outcomes. In contrast, health professionals have primarily focused on learning outcomes and have not examined the process of learning in any depth. For example, health studies (Robitaille and Kramer, 1985; and Thordarson and Costanzo, 1976) have identified birth weight outcome of prenatal class 9 participants and nonparticipants, however these studies have not addressed the dynamics of that learning process. Health professionals would be interested in understanding how women learn during pregnancy and the association between learning and health behaviors, but this interest would primarily be grounded in final outcomes, reduction of low birth weight incidence. The literature on the health aspects of low birth weight is extensive. Although low birth weight is not the primary focus of this study, the problems associated with low birth weight are addressed. This presentation is intended to strengthen the rationale for examining the learning that occurs about three primary behavioral risk factors associated with low birth weight: inadequate weight gain, alcohol consumption, and tobacco use. The problems of mortality associated with low birth weight are clearly identified in the classical study, the 1958 British Perinatal Mortality Survey. Butler and Alberman (1969) demonstrated that there were 227.0 deaths per 1,000 infants born weighing under 2,500 grams; compared with 28.2 deaths per 1,000 infants born weighing between 2,500 and 3,000 grams; and, compared with 12.4 deaths per 1,000 infants born weighing above 3,000 grams. More recently, Lee et al. (1980) as well as Saugstad (1981) found that low birth weight was an excellent predictor of perinatal mortality. The problems of increased morbidity associated with low birth weight incidence are documented in the Surgeon General's Report on Health Promotion and Disease Prevention (U.S. Department of Health, Education, and Welfare, 1979) and in the Institute of Medicine's (1985) report on Preventing Low Birthweight. Problems identified included increased occurrence of mental retardation, birth defects, growth and development problems, respiratory tract conditions, blindness, autism, cerebral palsy, and epilepsy. The incidence of each of these problems, as a result of low birth 10 weight, has not been clearly identified in the literature. The proportion of infants weighing 2,500 grams or less at birth in 1984 was 5.7 percent for all of Canada and 5.1 percent in British Columbia. These percentages translate into 20,997 infants in Canada as a whole and 2,206 infants in British Columbia (B.C. Ministry of Health, 1985). Variations exist across the province of British Columbia in the incidence of low birth weight. The range was from a 3.4 percent incidence in the Upper Fraser Valley Health District to a 6.0 percent incidence in the East Kootenay Health District. The Upper Fraser Valley Health District has a large Mormon community, a group of individuals who do not smoke or consume alcoholic beverages. The low birth weight incidence may reflect the health behaviors of the population of this health district. In the East Kootenay Health District, as a result of the high low birth weight incidence, the Kamloops low birth weight study was conducted by the B.C. Ministry of Health (Carlson and Phillon, 1985). A media campaign was organized along with production of two publications "Birth Weight & Why It's Important" and "Low Birth Weight Infants, Recognition and Preventive Management of High Risk Mothers." The low birth weight incidence in the East Kootenay Health District dropped to a 4.5 percent level in 1985. Although conclusive evidence is not available, one could speculate about the benefits of the education campaign in the East Kootenay District. Reduction of low birth weight incidence is a priority at both the federal and provincial governmental levels. The B.C. Ministry of Health has as one of its objectives to reduce the incidence of low birth weight to a 3 percent level (Fisher, 1984). This objective is close to the Swedish incidence of low birth weight which in 1981 was 3.4 percent (World Health Organization, 1984), the lowest reported incidence in the world. In 1983, The Canadian Department of Health & Welfare developed a "Five Year Federal-Provincial Plan on Nutrition in Health Promotion for 11 Pregnant Women" (Report of the Federal-Provincial Advisory Committee on Health Promotion, 1983). The purpose of the plan was to improve and maintain maternal and infant health through nutrition intervention in an attempt to decrease the incidence of low birth weight, thereby reducing infant morbidity and mortality rates. Health care costs associated with the problems of low birth weight have been one of the underlying incentives for the federal and provincial governments to focus on reducing low birth weight incidence. Budiansky (1986) reports that the Institute of Medicine, National Academy of Science, estimates the annual cost of neonatal intensive care to be approximately $2 billion in the United States. One stay in an intensive care unit can range from $15,000 to over $100,000. Other health care costs associated with low birth weight range from treating developmental problems to treating mental retardation. The costs of supporting a severely handicapped child throughout his lifetime is estimated by the Manitoba Community Task Force on Maternal and Child Health to be $750,000 (Fisher, 1984). Behrman (1985) estimates the unit cost of initial hospitalization of a low birth weight baby to be approximately $13,616 per low birth weight infant; the unit cost of rehospitalization to be approximately $372 per day per rehospitalized low birth weight infant; and, the unit cost for one year of ambulatory medical care for noninstitutionalized infants with morbidity to be approximately $1,805 per child. Documentation exists on the costs of neonatal intensive care units and other health care expenditures related to low birth weight, however, the psychological and social costs of long term disability and anxiety of parents are an intangible consideration (Papernik et al., 1985). Problems of increased morbidity and mortality associated with low birth weight, coupled with health care and social costs associated with low birth weight provide strong support for educators to pursue the study of learning related to weight gain, alcohol consumption, and tobacco use of pregnant women. 12 The goal of educators is to better understand the learning process during pregnancy and its relationship with health behavior, whereas the goal of health professionals is to ultimately reduce the low birth weight incidence. In this study, the goals of educators and health professionals are linked as schematically conceptualized in Figure 1. LEARNING AND HEALTH BELIEFS OF PREGNANT > WOMEN RELATIONSHIP WITH WOMEN'S HEALTH BEHAVIORS -weight gain -alcohol consumption -tobacco use RELATIONSHIP WITH LOW BIRTH WEIGHT INCIDENCE Fig. 1. The Link Between the Goals of Educators and Health Professionals In order to mesh the goals of educators and health professionals in this research study, the following questions are asked: (1) What are the learning patterns of pregnant women, that is: "What are pregnant women learning about weight gain, alcohol consumption, and tobacco use?" and "How do pregnant women learn about weight gain, alcohol consumption, and tobacco use?", (2) "What is the relationship between learning and behavior of pregnant women?", and (3) "What is the relationship between health beliefs and behavior of pregnant women?" Knowing what women learn and how they learn, along with the association of learning, health beliefs, and behaviors, provides educators with the tools to assist pregnant women 13 during their learning. As well, it can assist educators plan prenatal education programs for women. Definition of Terms A definition of the education terms used in this study is listed below. A description of the medical terms used is contained in Appendix 1. Education: A process by which human beings seek to improve themselves by increasing their skill, knowledge, or sensitiveness (Houle, 1972). Education is a condition established to facilitate learning. Knowledge: A cognitive or intellectual component acquired and retained through education or experience (Houle, 1972). Learning: A change in human disposition or capacity not ascribable to growth (Gagne, 1970). (Learning) Incident: An occurrence which may include a discussion, reading, or observation and whose end result is learning. (Learning) Episode: A series of learning incidents. (Learning) Transaction: The event of carrying through a learning incident with another individual, a group, or a nonhuman resource. Self-Directed Learning: A process whereby individuals take the initiative, with or without the help of others, to learn .(Adapted from Knowles, 1975). Self-Planned Learning: A process whereby individuals themselves are responsible for more than half of the detailed planning and deciding during learning (Tough, 1979). Resource: Any object or person which can be used for support in or help during the learning process. 14 Summary and Organization of the Dissertation In summary, Chapter I contains the rationale as to why educators are concerned about the relationship between learning during pregnancy and weight gain, alcohol consumption, and tobacco use. A brief overview of the two major components of the research problem, learning and health behaviors, was also provided. The research questions were identified, and definitions of the education and medical terms used in this text were provided. Chapter II consists of a review of the literature on learning, learning during pregnancy, behaviors related to perinatal outcomes, and behaviors related to health beliefs during pregnancy. Chapter III is a presentation of the conceptual framework of the study drawing primarily upon Tough's (1979) report "The Adult's Learning Projects," Knowles's concept of self-directed learning, and the Health Belief Model developed by Hochbaum, Kegeles, Leventhal, and Rosenstock (Rosenstock, 1974b). These works provide a framework for answering the research questions asked. This conceptual framework leads to a statement of the hypotheses of the study. Chapter IV outlines the methodology of the study and includes a description of the research design, population and sample, data collection procedures, interview schedule, and data analyses. Chapter V contains a description of the data collection and the sample characteristics. Participation rate, demographic characteristics, and health behaviors constitute the framework for describing the sample of this research study. Chapter VI is a presentation of the learning patterns and health beliefs of pregnant women as they relate to the research questions identified. Chapter VII is a report of the findings of the hypotheses tested. Chapter VIII is a discussion of the results. In addition, the significance and limitations of the study along with the conclusions and recommendations for theory, practice, and research are provided. 15 CHAPTER II REVIEW OF THE LITERATURE The review of the literature consists of five principal sections. Section one begins with an overview of adult learning and section two consists of an overview of learning opportunities during pregnancy. Section three specifically addresses the literature describing health behaviors of pregnant women and studies of learning about these health behaviors. Section four addresses the literature related to health behaviors and their association with perinatal outcome. Section five addresses the literature describing health beliefs and their association with behavior of pregnant women. The three health behaviors addressed in this dissertation are eating, drinking, and smoking. The final section is a summary of Chapter II. Learning Adults learn continuously throughout their lives. Havighurst (1969) pointed out that learning is necessary throughout life because of new developmental tasks that occur during the life cycle. The greatest changes in life prompt learning, and preparation for childbirth would be no exception. Learning, as defined by Gagne (1970), is a change in human disposition or capacity, which is not ascribable to growth, and that definition has been adopted for this study. Learning can occur in numerous settings and Jensen (Verner, 1964) has classified these settings into two major categories: (1) the natural societal setting, and (2) the formal instructional setting. In the natural societal setting learning may result from the day to day experiences of participating in the regular activities of life. Learning in this setting can be casual and incidental or it may be intentional. 16 Learning in the formal instructional setting occurs under the auspicies of an agent who has the task of inducing change in behavior (Verner, 1964). Little (1980) has further extended these two categories identified by Jensen and states that learning in the natural societal setting can be either (1) fortuitous, that is occurring by chance, or (2) intentional. Learning in the formal instructional setting can be either (1) education directed by self, or (2) education directed by others. In this classification Little distinguishes between learning and education. The educational process is a set of selected actions which systematically establish and maintain conditions that contribute to the achievement of learning objectives. Learning occurs entirely within the individual, whereas education occurs outside the individual and is a condition established to facilitate learning. Participation of adults in various educational situations has been extensively studied in the adult education literature, and the first major study of participation was the landmark work by Johnstone and Rivera in 1962-63. Comprehensive information about learning habits and practices of adults was obtained on a national probability sample of 11,957 households in the United States. Johnstone and Rivera (1964) estimated that one in five persons in the United States had been active in at least one form of learning during the twelve month period prior to the year of the study. Results showed that 15.0 percent of adults had been enrolled in courses on a part-time basis; 7.9 percent had been engaged in independent study; and, 2.3 percent had been full-time students. People were also asked if they had ever undertaken an educational course since leaving high school, and as many as 47 percent said they had. Approximately 38 percent recalled at least one occasion where they had tried to teach themselves something on their own. Research on participation rates of adults in learning activities has been reported to be between 12 and 98 percent of adults over the age of seventeen 17 (Cross, 1982). Cross attributes the wide discrepancy in participation rates to the different definitions of learning activities used in these studies. The U.S. Bureau of Census reported that 12 percent of adults were engaged in some form of organized adult education, defined to consist of courses and activities organized by a teacher or sponsoring agency (Cross, 1982). Taking a different approach, Tough examined learning activities from a new perspective, one that began with the individual learner. Tough examined deliberate learning efforts which he defined as learning projects. In a 1970 study of sixty-six individuals, Tough (1979) found that the typical person engaged in about eight learning projects per year with 98 percent of his sample having engaged in at least one learning project in the year prior to the actual study. These figures are higher than participation rates of other surveys of adult learning and Tough attributed these differences to the intensive probing questions asked by the interviewers to assist adults in recalling their learning activities. Tough had devised a methodology for the investigation of learning which was subsequently replicated by other researchers in different settings. By 1980 nearly fifty surveys had been conducted in Australia, the United Kingdom, the United States, Israel, Jamacia, New Zealand, and Zaire. Tough (The International Encyclopedia of Education, 1985) summarized the results of these studies and indicated that approximately 90 percent of all individuals had conducted at least one learning project a year with the average or typical adult learner conducting five learning projects in one year. He estimated that the typical learner spends an average of approximately 100 hours per learning project for a total of 500 hours per year or ten hours a week. Of those learning projects undertaken, about 73 percent were planned by the learner, 7 percent planned by a nonprofessional, and 20 percent planned by a professional educator or a guided set of materials. The largest national survey using Tough's framework to examine the extent of learning in the population was that conducted by Penland (1977) on a national 18 probability sample in the United States. The sample was not totally representative of the adult population. Five thousand four hundred and ninety-three households were initially contacted. No more than two contacts per household were made. Of the contacts made, 1,501 individuals completed the interview and 1,193 refused to participate. Some household members could not be reached, and some households eligible for a second call did not actually receive one since interviewers were instructed to stop interviewing in a location after completing a specific number of interviews. Penland categorized the 1,501 participants in the study into four groups and found that: 60 percent were self-initiating learners planning their own learning projects; 16 percent were combination learners involved in self-planned learning and course-like activities; 3 percent were learning for credit; and, 21 percent were nonlearners. Although it cannot be stated that the total adult population is engaged in learning, there is evidence that a great deal of the population is engaged in learning outside the organizational setting. A question that remains to be answered is "What is the extent of learning that is occurring during pregnancy, and what is the relationship between this learning and a pregnant woman's health behavior?" To the best knowledge of this researcher, no in-depth study on learning during pregnancy exists in the literature. One related study was conducted by Cobb (1978) to determine whether prospective parents were engaged in self-directed learning about parenting. The findings revealed that 97 percent of the sample were engaged in self-directed learning, however the author acknowledges that the results cannot be generalized to all prospective parents since the sample consisted of eighty-six parents enrolled in Lamaze Childbirth Education Classes, a formal education program appealing to middle and upper class adults. 19 In summary, learning can occur in numerous settings and consists of a change in human disposition or capacity not ascribable to growth. Participation of adults in various learning activities has been reported to range from 12 to 98 percent, depending on the definition of learning activity used in the study. Tough reported that a large percentage of the population is engaged in learning activities outside the organizational setting. An unanswered question is: "What is the extent of learning that occurs among the pregnant population?" Learning Opportunities During Pregnancy Numerous opportunities are available to pregnant women to engage in learning. Learning with family members, learning with friends, learning with a health professional, learning at prenatal classes, and individual learning are some examples. Houle's typology of educational design situations is the framework used to classify these learning opportunities. Houle (1972) in his Design of Education states as one of his assumptions that "Any episode of learning occurs in a specific situation and is profoundly influenced by that fact". ^ Houle defines episode as a related succession of acts, making an educational whole. He identifies eleven categories of educational design situations during which learning can occur. These eleven categories are described on the following page. In classifying various learning opportunities during pregnancy according to Houle's eleven educational design situations, four categories seem to dominate prenatal learning activities. These include: Category 1, an individual designs an activity for herself. This could include individual learning opportunities such as reading or viewing * Cyril O. Houle, The Design of Education, (San Francisco: Jossey Bass, 1972), p. 32. 20 MAJOR CATEGORIES OF EDUCATIONAL DESIGN SITUATIONS INDIVIDUAL C-l An individual designs an activity for himself C-2 An individual or group designs an activity for another individual GROUP C-3 A group (with or without a continuing leader) designs an activity for itself C-4 A teacher or group of teachers designs an activity for, and often with, a group of students C-5 A committee designs an activity for a larger group C-6 Two or more groups design an activity which will enhance their combined programs of service INSTITUTION C-7 A new institution is designed C-8 An institution designs an activity in a new format C-9 An institution designs a new activity in an established format C-l 0 Two or more institutions design an activity which will enhance their programs of service MASS C-ll An individual, group, or institution designs an activity for a mass audience ^ an audiovisual production about prenatal issues; Category 2, an individual designs an activity for another individual. Individuals designing an activity for pregnant women could include learning opportunities organized by the physician, dietitian/nutritionist, nurse, family members, or friends. A group designing an activity for an individual could include The Healthiest Babies Possible Program whereby a group of health professionals design a program to provide individual counselling to women unable to speak English or those living on a limited income; Category 5, a committee designs an activity for a larger group. This category could include prenatal and childbirth education classes; and Category 11, an individual, group, or institution designs an activity for a mass audience. This could include programs produced for cablevision, or mass media campaigns produced in the United States on drinking during pregnancy. ^ Cyril O. Houle, The Design of Education, (San Francisco: Jossey Bass, 1972), p. 44. 21 A more in-depth description of these prenatal educational design situations, according to Houle's classification, is now presented. The purpose of this description is to demonstrate that any learning episode is influenced by the situation in which it occurs. C-l Individual Learning. To date no evidence exists on the extent of learning during pregnancy where an individual designs an activity for herself. It can be inferred that the woman designs her own learning activity with the use of books, articles, and pamphlets available on the subject of pregnancy, or by viewing audiovisual productions. C-2 Learning with the Physician. The physician is considered the center of prenatal care for the pregnant woman for the following reasons: (1) the physician is responsible for the primary care of the pregnant woman, (2) the physician is the only professional able to admit the woman to the hospital, and (3) the physician sees the pregnant woman on numerous office visits during the prenatal period. Obstetrical visits are usually recommended monthly until the thirty-second week of gestation; biweekly until the thirty-sixth week of gestation; and, weekly until delivery (Backman, 1983). The American College of Obstetricians and Gynecologists recommend that a pregnant woman make about thirteen visits for prenatal care during the course of a normal pregnancy (National Centre for Health Statistics, 1986). The physician may provide information directly to the pregnant woman and/or refer her to other health professionals. Some women lack access to physicians for a variety of reasons, but in the United States only 2 percent of women in 1984 had no prenatal care during their pregnancy (National Centre for Health Statistics, 1986). C-2 Learning with the Dietitian. Outpatient hospital dietitians accept referrals for prenatal nutrition counselling from physicians for the at-risk pregnant woman. This service is free of charge to the woman in British Columbia and is covered by the 22 B.C. Ministry of Health hospital program budget. Sixty-three of the seventy-eight acute care hospitals in British Columbia enlist the services of an outpatient dietitian (Schwartz, Bell, and Webber, 1987). C-2 Learning with the Public Health Nurse. The public health nurse has close contact with the members in her community district and therefore may be in contact with women having financial, family, or health problems. A home visit may occur and learning could potentially take place during this visit. A second way in which learning with the public health nurse may arise is as a follow-up to prenatal classes. Women attending prenatal classes are asked to complete a brief questionnaire in the publication "You and Your Baby," a prenatal assessment form. Questions are related to dietary habits, alcohol consumption, tobacco use, medication use, stress, and other health concerns. Responses are evaluated by the public health nurse and if follow up contact is deemed necessary, a home visit is conducted. C-2 Learning with Family Members and Friends. The influence and extent of learning during pregnancy with family members and friends is not documented. It is known that in many Asiatic cultures the words and advice of mothers, grandmothers, or elder women in the family play significant roles in learning about health issues. Some cultures deal with health problems by first consulting the elders in the family and will only resort to the medical profession if they are unable to resolve their own health problems (Wood, 1985). C-2 Learning at the Healthiest Babies Possible Program. Healthiest Babies Possible is a Vancouver Health Department outreach prenatal program that provides nutrition counselling to pregnant women who live within the Vancouver City limits. The service is aimed at women with a low income and those who do not speak English. Counselling, which takes place in the woman's own home, is conducted by either a dietitian or a nutrition aide and when necessary an interpreter is made available. 23 These counsellors help the pregnant woman decide what and how much to eat during pregnancy, how to plan balanced meals on a limited budget, how to incorporate traditional or ethnic foods into the diet, and how to consider other life style habits which may affect the developing baby (Healthiest Babies Possible, 1979 and 1985). Self referral, physician referral, and other health professional referrals are accepted as long as the woman meets the eligibility criteria of the program, low income or English as a second language. C-5 Learning at Prenatal Classes. The B.C. Ministry of Health offers prenatal classes to pregnant women. The majority of health units in British Columbia conduct two classes for women in their first trimester of pregnancy, often referred to as the "early bird series." Nutrition, alcohol consumption, and tobacco use are the primary topics discussed in most early bird classes. Health units also conduct labor and delivery classes for women during the latter part of their pregnancy, often referred to as the "regular series." The number of classes in the regular series can range from two to six depending on the health district policy. The number of classes as well as the content of the classes varies with each health unit, and therefore the classes are not standardized. During the series, the women are asked to complete a screening pamphlet and if indicated, follow up individual counselling is done or arranged by a public health nurse. If a nutrition problem arises in health units where a nutritionist is available, the woman is referred for nutrition intervention. The Vancouver Childbirth Association is a voluntary nonprofit organization that offers prenatal classes in five areas of the Lower Mainland: Burnaby, Coquitlam, Surrey, North Vancouver, and Vancouver. Their series consists of two "early bird" classes, eight pre-labor classes, and one postpartum class. The instructors do not have to be qualified health professionals but must undergo a training program provided by the Association. Following training, the instructors are required to pass 24 an examination which makes them eligible to become "qualified childbirth educators." There is, however, a prerequisite that all instructors must have given birth (Brown, 1985). Some of the other prenatal classes that are available to pregnant women include: Grace Hospital prenatal classes, Saint Paul's Hospital prenatal classes, Surrey-Memorial Hospital prenatal classes, Kwantlen College prenatal classes, and private prenatal classes organized by midwives. Burnaby Health Department, North Shore Health Department, Richmond Health Department, and the Vancouver City Health Department also offer prenatal classes. There are some differences between services offered by the B.C. Ministry of Health and those of the Burnaby, North Shore, Richmond, and Vancouver Health Departments, as these are primarily under municipal jurisdiction. One of the differences is that the majority of health units within the Vancouver City Health Department do not offer "early bird" prenatal classes. Instead, a kit containing both information pamphlets and a screening pamphlet is available to pregnant women who call the health unit. Women are asked to complete and return the screening pamphlet to the health unit, and follow up intervention is conducted if the woman is identified as being at-risk. C-ll Learning Designed for a Mass Audience. An example of an activity designed for a mass audience is a set of Prenatal Education Videotapes produced by the Vancouver City Health department in Cantonese, Hindi, and Punjabi languages. The programs were broadcast on cablevision from July to December of 1982 (Kendall, 1983). By categorizing learning activities during pregnancy according to Houle's educational design situations, one can obtain a sense of the diversity of the learning opportunities available and identify the different focus each activity has to offer. Much of the research conducted on learning during pregnancy has examined the 25 prenatal class educational design category. Kendall (1985) surveyed twenty-two urban Canadian Health Departments to identify how many provide prenatal programs, and to identify the types of program objectives and evaluation procedures. Kendall found that all the health departments he studied offered prenatal programs. The percentage of pregnant women receiving service ranged from 25 to 75 percent. Ten departments estimated that they saw 25 percent to 50 percent of their pregnant population. Only four of the twenty-two health departments surveyed by Kendall (1985) had specific measurable objectives for their programs. Examples of two measurable objectives cited were: The pregnant woman who attends prenatal class will gain twenty to thirty pounds, and the pregnant woman who attends prenatal class will give birth to an infant weighing 3,300 grams or more. General goal statements for prenatal classes included: 1) to assist expectant families to prepare for and to anticipate labour and delivery with confidence; 2) to encourage expectant parents to prepare physically' for childbearing; 3) to help each woman achieve her optimum physical, mental, and emotional level of health throughout pregnancy; 4) to give parents instructions in methods of relaxation, posture, muscle tone and breathing for a healthier and more comfortable pregnancy, labour and post delivery period; 5) to help both expectant parents acquire an appreciation of the changes each will experience as a result of the pregnancy and birth; 6) to assist expectant parents in understanding the needs of their own newborn and to encourage beginning skills in the care and feeding of their newborn; 7) to reinforce the importance of the husband-wife relationship as it enhances the development of responsible parenthood; 8) to provide the concept of healthful life style. Regarding evaluation of prenatal programs, only seven of the twenty-two health departments had formally assessed their programs. The types of evaluation used were primarily to review attendance percentages of the total population, and to review perceived utility of the information provided. The majority of program evaluations were directed to level three and level four of Bennett's seven level hierarchy for program evaluation (Boyle, 1981). Bennett's level three is directed at Perry R. Kendall, "Survey of Canadian Urban Public Health Prenatal Programs Goals Objects, Populations Served and Outcome Measures," Canadian Journal of Public  Health 76 (July/August 1985): 270. 26 evaluating participation rates; level four is directed at evaluating personal reactions of the participants; level five is directed at evaluating knowledge, attitude, skill, and aspiration change; level six is directed at evaluating practice or behavior change; and, level seven is directed at evaluating end results, an example being birth weight of the infant. Kendall is critical of the lack of measurable objectives and stringent evaluation measures. He comments that existing prenatal classes meet public demand, are exceedingly popular, and may increase parents' confidence levels in preparation for birth. However, at the same time he questions whether these reasons are sufficient justification for resource allocation in financially difficult times. He suggests that resources be redirected to programs with more clearly defined goals which are linked to pregnancy outcome. Kendall's comments about the efficacy of prenatal classes are partly supported in a recent study conducted in Montreal by Robitaille and Kramer (1985). In a prospective epidemiologic survey of 1,676 primiparous women delivering in four Montreal Hospitals during an eight month period, the researchers concluded that participation in prenatal classes was not associated with maternal weight gain or infant birth weight outcome, once adjustments had been made for age and socioeconomic status. On the other hand, reduction in cigarette smoking was greater for prenatal class participants even when age and socioeconomic status had been controlled for in the analysis. However, before concluding that prenatal classes are not efficacious, one needs to question whether learning activities during pregnancy in other settings may have masked the differences observed between participants and nonparticipants. Results of studies which examined the relationship between attendance at prenatal classes with attitudes and confidence levels toward coping with labor and delivery have been mixed. McCraw and Abplanalp (1984) found that participation in 27 a childbirth education class did not affect maternal attitudes toward labor and delivery along with other maternal attitudes as identified in Schaefer and Manheimer's Pregnancy Research Questionnaire; whereas, Walker and Erdman (1984) found that participation in a childbirth education class significantly decreased primiparous women's self-reported ratings of anxiety to cope with labor. Robitaille and Kramer (1985) found that primigravidas who participated in prenatal classes in Montreal were older, of higher socioeconomic status, and less likely to be smokers than were pregnant nonparticipants. These findings are consistent with Thordarson and Costanzo's (1976) examination of participants of prenatal classes in the Vancouver area. Prenatal class attendance in Vancouver was found to be lower among younger, poorer, and less educated women. Vinal (1981), on the other hand, identified prenatal class participants as being young and well educated, while nonparticipants had experienced more pregnancies and had been married considerably longer. This profile of participants in prenatal classes is consistent with the adult education literature on participation in organized learning activities. Darkenwald and Merriam (1982), upon examining the results of the "National Centre for Education Statistics, Participation in Adult Education Final Report," concluded that participants of organized learning activities in contrast to nonparticipants were younger, white, better educated, and more affluent. Selwood (1984), in a study examining the impact of prenatal education on the conduct of second stage labor conducted at Grace Hospital in Vancouver, the largest maternity hospital in British Columbia as well as in Canada, described the fifteen women in her study who had not participated in prenatal classes. Of the nonparticipants, one had been born in North America and the other fourteen individuals had been born in one of the following Asiatic Countries: the Philippines, Hong Kong, Vietnam, and China. Although the study was not designed to provide a profile of participants and nonparticipants attending prenatal classes, this demographic information is of interest 28 since it is a Vancouver based study. Furthermore Selwood noted that individuals who had resided in Canada longer than ten years were more likely to have attended prenatal classes. In summary, the research examining learning that occurs during pregnancy has primarily focused on participation in prenatal classes. Participants of prenatal classes can be described as older women of childbearing age, better educated women of a higher socioeconomic status, and non-smoking women experiencing their first pregnancy. Although most research studies have focused on learning at prenatal classes, if learning activities are categorized according to Houle's educational design situations, it becomes evident that learning during pregnancy is not limited to prenatal classes. There are numerous learning opportunities available to the pregnant population. Description of Health Behaviors and Learning about Health Behaviors Eating, drinking, and smoking constitute the three health behaviors examined. For the purposes of this dissertation, these behaviors are operationalized in terms of their outcomes, that is: weight gain, alcohol consumption, and tobacco use of pregnant women. Research studies which have described women's health behaviors during pregnancy are reviewed, and learning about these health behaviors presented. Weight Gain: Behaviors and Learning Research conducted on describing weight gain of pregnant women and learning about weight gain has taken diverse directions. Weight gain has usually not been studied in isolation and has been included in studies which have focused on examining dietary practices of pregnant women. The learning that occurs about weight gain or dietary practices during pregnancy has focused on reporting pregnant women's sources of nutrition information and assessing the effectiveness of 29 intervention strategies to improve birth weight through nutrition counselling. Dietary practices and learning about weight gain are described as follows. Behaviors: Weight Gain and Dietary Practices The Nutrition Canada Survey (Food Consumption Patterns Report, 1977) found that 92 percent of pregnant women participating in the survey had made dietary changes during their pregnancy. Sixty-seven percent of these changes were self-initiated, and only 19 percent of these changes had been made on the advice of a doctor or clinic. Fruit and milk were the foods in which the greatest changes in consumption were seen. Sixty-one percent of pregnant women had increased their fruit consumption and 59 percent had increased their milk consumption. The pregnant women's self-reported dietary changes were confirmed by comparing the mean dietary intake from data collected on pregnant women with those on nonpregnant twenty to thirty-nine year old females participating in the study. A total of 894 pregnant women across Canada were involved in the survey, and it must be noted that those pregnant women who participated had been referred by local health authorities. As such, the data may be biased to show a superior picture of health and motivation compared to a random sample of the pregnant population. Johnston, Hyson, and Blackmere (1985) recently conducted a study to determine weight gain and related characteristics of pregnant Nova Scotia women. Five hundred and two women participated in the study. Questionnaires were voluntarily completed by maternity patients in eighteen hospitals throughout Nova Scotia for a one month period in 1982. Increased consumption for all the basic food groups was reported, with milk and milk products, and fruits and vegetables showing the greatest increases. Specifically, 70 percent of the sample had increased their milk and dairy intake and 62 percent had increased their fruit and vegetable intake. Sixty-one percent of the women indicated they had eaten to satisfy hunger, 38 30 percent had reported experiencing cravings, and 22 percent had indicated aversions to particular foods. Furthermore, approximately 80 percent of those responding indicated that the information they had received during pregnancy had influenced their dietary habits. In contrast, Halliday (1978) reported that only 32 percent of seventy-three pregnant women attending a two hour prenatal nutrition class in Saskachewan had changed their dietary habits as a result of the prenatal class. Orr and Simmons (1979), on the other hand, interviewed ninety-two pregnant women in an outpatient hospital setting in Boston. Approximately 66 percent of these women had made dietary changes attributable mainly to dietary advice given by a dietitian. Although no statistically significant relationship was found, a not surprising trend was observed, indicating that women who had expressed a desire for nutrition advice were more likely to make changes in their diets. In a Vancouver study, Kiss (1983) described the health behaviors of prenatal class participants. Smoking, alcohol, nonprescription drug use, caffeine, and diet were examined. Women registered at "early bird" classes at six health units with the Vancouver City Health Department were approached to participate in the study. The sample consisted of two hundred and twenty women. Women completed the prenatal assessment form provided at the classes as well as a health habits questionnaire designed for the study. Ninety-nine percent of the women had reported changing their dietary intake. Of these women, 94 percent had increased their milk intake, 84 percent had increased their fruit and vegetable intake, 49 percent had increased their meat intake, 45 percent had increased their bread and cereal intake, and 41 percent had decreased their intake of sweets. Although the majority of women had reported increasing their food intake, actual number of servings consumed from each of the food groups in Canada's Food Guide were less than adequate. Only 37 percent reported having four or more servings of milk and milk products. Eighty-three percent reported having at least two servings of meat or alternates, 79 percent 31 reported having at least four servings of fruits and vegetables, and 65 percent reported having at least four servings of breads and cereals. When all the food groups were combined, 81 percent of women had diets which did not meet the minimum requirements of Canada's Food Guide. Women in the Kiss (1983) study were also asked to check the factors which had influenced them the most in making their dietary changes. A list of influencing factors was provided on the questionnaire and the following results were obtained: 67 percent stated that their own personal knowledge had influenced their dietary changes; 50 percent stated that books-magazines-pamphlets had influenced their dietary changes; 48 percent stated that cravings and appetite had influenced their dietary changes; 42 percent stated that their doctor had influenced their dietary changes; 36 percent stated that attending prenatal classes had influenced their dietary changes; 25 percent stated that the public health nurse had influenced their dietary changes; 25 percent stated that their tolerance to food had influenced their dietary changes; 19 percent stated that their family members and friends had influenced their dietary changes; 11 percent stated that a nutritionist had influenced their dietary changes; and, 1 percent stated that the media had influenced their dietary changes. Although 67 percent of the women had stated that their own personal knowledge had influenced their dietary changes, it is not known from which resources they had obtained their information. Print material, cravings and appetite, and the physician had also been cited as influencing women's dietary changes during pregnancy, however public health personnel had played secondary roles. A study of attitude toward weight gain was conducted by Palmer, Jennings, and Massey (1985) to determine whether attitudes toward slimness affect weight gain during pregnancy. The concern of women to retain a slim figure during pregnancy may prejudice prospective mothers against normal prenatal weight gain. The sample 32 of this study consisted of twenty-nine pregnant women enrolled in childbirth classes in Washington State. The women were of a middle class background with at least high school education. Although the sample was not representative of the pregnant population, a statistically significant relationship was found between weight gain and: (1) a positive attitude towards weight gain, (2) physician recommendations, and (3) knowledge of appropriate weight gain. It is of interest to note that 41 percent of the women in the sample had a negative attitude toward weight gain during pregnancy. Since attitude and knowledge were both found to be associated with appropriate weight gain, the authors recommend that professionals dealing with pregnant women include an assessment of attitudes and knowledge. Pomerance et al. (1980) surveyed 195 pregnant women at two medical centers in California to assess attitudes towards recommended amounts of weight gain. Thirty-seven percent of patients believed they should gain less than twenty pounds during their pregnancy, and 4 percent believed they should gain more than thirty pounds. Eighty-eight percent of the patients thought their doctors would be concerned with too much weight gain, and 39 percent thought their doctors would be concerned with too little weight gain. These results are contrary to the medical literature which indicate that too little weight gain is associated with having a low birth weight infant, rather than too much weight gain. Patients' perceptions about weight gain were not compared to actual weight gain. A different pattern of perceptions about prenatal weight gain emerged from the Johnston, Hyson, and Blackmere study (1985). Although only 8 percent of 502 pregnant women believed that they should gain fewer than twenty pounds, 13 percent of the pregnant women in the study actually gained less than twenty pounds. The B.C. Ministry of Health (McCarthy and Mackay, 1984) weight gain recommendations are twenty-five to twenty-seven pounds for the normal weight 33 healthy pregnant woman, thirty to thirty-two pounds for the underweight woman, and sixteen to twenty pounds for the overweight woman. The studies of Pomerance et al. (1980), and Johnston, Hyson, and Blackmere (1985) did not ascertain whether the women who perceived that they should gain less than twenty pounds were actually overweight women. If the women were overweight, then these perceptions would be consistent with health professionals' recommendations about weight gain, otherwise the women's perceptions would be contrary to health professional recommendations and potentially place the fetus at risk. A discrepancy between perceived self evaluation and health professional evaluation was found in the Orr and Simmons (1979) study. Those individuals who stated that they had no need for nutrition advice during pregnancy were those individuals who tended to be underweight, as defined by the Chicago Build and Blood Pressure Study, and had low hematocrit readings, as defined by the Interdepartmental Committee on Nutrition for National Defense (Orr and Simmons, 1979). Taffel and Keppel (1986) examined data of 7,704 pregnant women who had participated in the 1980 National Natality Survey conducted in the United States. All these women were married and had received prenatal care during their pregnancy. Sixty percent of these women reported that their doctors had suggested a weight gain limit during pregnancy. Forty-eight percent of the doctors had suggested a twenty-two to twenty-seven pound limit, 22 percent had suggested a twenty-eight to thirty-four pound limit, 23 percent had suggested a sixteen to twenty-one pound limit, 4 percent had suggested a gain of less than sixteen pounds, and 3 percent had suggested a gain of more than thirty-four pounds. If all the women who had received recommendations to gain less than twenty-one pounds had been obese, then these recommendations would have been appropriate. Otherwise, these recommendations would have been contrary to weight gain recommendations for underweight women and women of normal weight. The association between reported weight gain advice 34 and actual weight gain was also examined. Women who had reported receiving no weight gain limit were more likely to gain less than twenty-two pounds. Furthermore, the more stringent the weight limit reported, the more likely a woman would gain less than twenty-two pounds. The results suggest that reported weight gain advice and actual weight gain are associated. In summary, research which examines weight gain during pregnancy and dietary practices indicate that the majority of women attempt to change their dietary habits with increases in milk, fruits, and vegetables showing the greatest improvements. Key factors which are reported to influence dietary change include: women's knowledge, print material, hunger, and the physician. Physician recommendations were found to be associated with appropriate weight gain. However in one study, 27 percent of women had received advice from physicians to gain less than twenty pounds during pregnancy, an amount contrary to recommended weight gain for average healthy women. As well, between 8 and 37 percent of pregnant women believed they should gain less than twenty pounds during pregnancy, and 41 percent had a negative attitude toward weight gain during pregnancy. Perceptions, attitudes, and knowledge about weight gain, as well as advice from health professionals are potential influencing factors on whether women achieve ideal weight gain during their pregnancy. Learning about Weight Gain: Information Sources and Intervention Strategies The research related to learning about weight gain during pregnancy has been primarily focused on reporting pregnant women's sources of nutrition information. In a study conducted at the Vancouver General Hospital by Schwartz and Barr (1977), 150 new mothers volunteered to participate. It was found that the physician was named by 63 percent of the women to be the most helpful human source of nutrition information during their pregnancy. The physician was followed by the 35 husband at 35 percent, and the prenatal class instructor at 30 percent. The authors classified past education and experience to be the most helpful material source of nutrition information during pregnancy. Past education and experience were named by 63 percent of the women, followed by prenatal class booklets at 44 percent, and pocket books at 28 percent. Because more than one source of information was given by some of the pregnant women, the total percentages exceed 100. Johnston, Hyson, and Blackmere (1985) also reported that the family doctor was the most frequently mentioned source of nutrition information at 58 percent. This was followed by prenatal classes at 47 percent, books at 40 percent, public health nurses at 28 percent, and family members at 27 percent. The total percentages exceed 100 because more than one major source of information was listed by these pregnant women. Five percent of the women in the Johnston study indicated they had not received any nutrition information during their pregnancy. Whether these women had previously obtained adequate information remains unknown. Orr and Simmons (1979), on the other hand, reported that the dietitian was the major source of nutrition information during pregnancy. The subjects of their study, however, were women being seen in an outpatient clinic of a large obstetrical hospital. Certification of prenatal dietary instruction by the dietitian was necessary for payment of additional welfare allotments. Since welfare recipients comprised 55 percent of the subjects in this study, it is understandable that the dietitian was cited as the primary source of nutrition information. Furthermore, 50 percent of the pregnant women in the Orr and Simmons study said they felt they had not needed professional dietary instruction. The reasons women gave for not needing the instruction were as follows: 39 percent felt that they had sufficient information from previous prenatal information received; 35 percent felt that their general knowledge was sufficient; 11 percent said they had obtained their information from reading; and, 36 15 percent reported they had obtained their information from advice of mothers or friends. Palmer, Jennings, and Massey (1985) had asked the twenty-nine participants in their study how much weight should be gained during pregnancy. Sixty-six percent of these pregnant women responded that weight gain should be between twenty and twenty-six pounds, an answer deemed by the authors to be the "correct" response. Research examining the effect of nutrition counselling on maternal weight gain has usually included measurement of the efficacy of nutrition counselling in relation to birth weight. Orstead et al. (1985) studied the effect of intensive nutrition counselling on weight gain and low birth weight incidence, in comparison to the effect of a single nutrition class on weight gain and low birth weight incidence. Although the authors conclude that intensive nutrition counselling results in a superior pregnancy outcome, there is a research design flaw in the study. The control group attended one prenatal class lecture and consisted of pregnant patients attending a Chicago hospital clinic between 1975 and 1977, whereas the test group attended one prenatal class along with intensive nutrition counselling, and consisted of pregnant patients attending the same hospital clinic between 1979 and 1980. The fact that the control group data came from the 1975-1977 period and the test group data came from the 1979-1980 period makes it difficult to establish whether these two groups can be considered comparable. Differences in instructors conducting prenatal classes could potentially exist and the extent of knowledge and information available to the medical community between these two time periods lead one to question the validity of the results. McDonald and Newson (1986) examined the dietary changes made by 430 nutritionally high risk pregnant women who received nutrition intervention from the Prince Edward Island Department of Health and Social Services. Dietary histories 37 were taken at entrance to and throughout the counselling program. Average daily energy and protein intakes were determined by averaging the diet histories taken during the entire intervention period. Energy and protein intakes during the intervention period were significantly higher during intervention in comparison with initial energy and protein intakes. What is of particular importance to note is that those women who entered the program after twenty weeks gestation, classified as late referrals, had higher initial energy and protein intakes than women referred prior to twenty weeks gestation. This difference suggests that some form of change occurred during early pregnancy. Whether these changes were due to physiological increases in appetite, self-initiated change, or change recommended by a health professional remains unknown. Even though initial energy and protein intakes were higher for late referrals, a significant difference was found between initial intake of late referrals and average intake during late intervention. The results of this study support the notion that nutrition intervention can significantly increase caloric and protein intake, however the research design of the study did not examine the effect of nutrition intervention on birth weight outcome. Ershoff et al. (1983) evaluated a prenatal health education program conducted within a Health Maintenance Organization (HMO) in Southern California. The experimental group consisted of fifty-seven pregnant smokers who received individual nutrition counselling and a home correspondence smoking cessation program. These women's behavior was evaluated against outcomes of the control group, seventy-two pregnant smokers receiving standard prenatal care. The control group had a slightly higher percentage of white women, women with less than high school education, and women with family incomes of less than $15,000 per year. The experimental group, on the other hand, had a higher percentage of women who had had one or more miscarriages, as well as a higher incidence of toxemia and premature deliveries during previous pregnancies. The results indicated that a significantly greater 38 percentage of women in the experimental group had adjusted their diets than women in the control group (91 percent versus 68 percent, p<0.01). As well, 90 percent of the women in the experimental group had gained at least twenty-four pounds during pregnancy, in contrast to 77 percent of women in the control group (p<0.10). Analyses of birth weight data revealed that infants born to women in the experimental group had a significantly higher mean birth weight than infants born to women in the control group (p<0.05). Furthermore, low birth weight incidence was 7.0 percent in the experimental group and 9.7 percent in the control group (statistically non-significant). The Montreal Diet Dispensary program (Higgins, 1976) and the Special Supplemental Food Program for Women, Infants, and Children (WIC) in the United States (Collins, Demellier, and Leeper, 1985) have also attempted to examine the effect of nutrition intervention strategies coupled with food supplementation on maternal weight gain and low birth weight incidence. The authors however were unable to separate the effects of nutrition counselling from those caused by the food supplementation on the final outcomes. In summary, research on learning about weight gain has primarily focused on enumerating the sources of information utilized by pregnant women. The physician dominates as the major source of nutrition information during pregnancy. Past education and experience as well as prenatal classes were also cited as information sources. Reports of the effect of nutrition counselling on maternal weight gain and infant birth weight have been mixed. Because numerous factors can potentially influence maternal weight gain and infant birth weight outcome, methodologic problems exist in this type of research. Despite these problems, intervention strategies related to nutrition counselling show an association with improved pregnancy outcome. 39 Alcohol Consumption: Behaviors and Learning Alcohol intake of the pregnant population received little attention prior to the interest shown in Fetal Alcohol Syndrome, a syndrome which was first described by Jones and Smith in 1973 (Cushner, 1981). It was after this time that researchers focused on describing drinking behaviors of pregnant women. Learning about alcohol during pregnancy has primarily focused on media campaigns targeting the general population, and intervention strategies targeting pregnant problem drinkers. Very few studies have reported women's sources of information on the topic. Behaviors and learning about alcohol are described as follows. Behaviors: Alcohol Consumption One of the earlier studies which examined drinking patterns during pregnancy was conducted by Little, Schultz, and Mandell in 1976 in Seattle, Washington. One hundred and sixty-two women were interviewed during their fourth month of pregnancy, and of these, 156 women were available for a second interview during their eighth month of pregnancy. Ninety-five percent of the women were white and of a middle class background. The interviews produced estimates of alcohol consumption during three consecutive time periods: six months before pregnancy, the first four months of pregnancy (early pregnancy), and the last four months of pregnancy (late pregnancy). Fifteen percent of the women had abstained from consuming alcohol before pregnancy, 19 percent had abstained from consuming alcohol during early pregnancy, and 23 percent had abstained from consuming alcohol during late pregnancy. The magnitude of the decrease in alcohol consumption during the early pregnancy period was directly proportional to the level of consumption before pregnancy period (V=0.71, p<0.001). Approximately 66 percent had consumed less alcohol during their first four months of pregnancy in comparison to prepregnant levels. However during the last four months of pregnancy, 36 percent of the women 40 were drinking more and 40 percent were drinking less than the amounts reported during the first four months of pregnancy. Although it is not reported in the article, it is assumed that 24 percent of the women had consumed the same amounts of alcohol during the last four months as in the first four months. These data tell us that a decrease in alcohol consumption had occurred from before pregnancy to early pregnancy, and an increase in alcohol consumption had occurred from early pregnancy to late pregnancy. Women were also asked why they had changed their alcohol consumption during pregnancy. Over half of the women who reported a decrease in the use of alcoholic beverages during pregnancy had related the change to a physiologic effect such as nausea, stomach irritation, headache, or the bad taste of alcohol rather than to health reasons such as fetal welfare or other health considerations. A comparison of drinking patterns among pregnant women in Seattle, Washington was conducted by Streissguth et al. (1983) who evaluated changes in drinking and smoking patterns from 1974/75 to 1980/81. The subjects were two cohorts of pregnant women who were interviewed at six months gestation at two Seattle hospitals. Response rates were not reported. Sample size for 1974/75 was 1,529 women, and for 1980/81 was 1,413 women. Alcohol consumption levels were reported in ounces of absolute alcohol. One ounce of absolute alcohol is equivalent to approximately two drinks of wine, beer, or liquor. In 1974/75 prior to pregnancy, 20 percent had abstained from drinking; 73 percent had consumed less than an average of one ounce of absolute alcohol per day; and, 7 percent had consumed greater than an average of one ounce of absolute alcohol per day. In 1980/81 prior to pregnancy, 35 percent had abstained from drinking; 59 percent had consumed less than an average of one ounce of absolute alcohol per day; and, 6 percent had consumed greater than an average of one ounce of absolute alcohol per day. In 1974/75 during pregnancy, 19 percent had abstained from drinking; 79 percent had consumed less 41 than an average of one ounce of absolute alcohol per day; and, 2 percent had consumed greater than an average of one ounce of absolute alcohol per day. In 1980/81 during pregnancy, 58 percent had abstained from drinking; 41 percent had consumed less than an average of one ounce of absolute alcohol per day; and, 1 percent had consumed greater than an average of one ounce of absolute alcohol per day. Although the proportion of heavy drinkers who consumed alcohol during pregnancy had decreased over the six year period from 2 percent in 1974/75 to 1 percent in 1980/81, the proportion of heavy drinkers among those who consumed alcohol had not decreased. For women who had consumed more than an average of one ounce of absolute alcohol per day, the most dramatic decreases were found in higher educated and older women. There was a significant reduction in the number of drinkers who reported consuming five or more drinks on a single occasion during pregnancy, referred to as "binge drinkers." In 1974/75, 19 percent had engaged in binge drinking prior to pregnancy and this figure decreased to 12 percent during pregnancy. In 1980/81, 17 percent had engaged in binge drinking prior to pregnancy and this figure decreased to 8 percent during pregnancy. Streissguth and associates (1983) also reported the type of alcoholic beverages consumed. In 1974/75 the beverages of choice were wine and liquor, whereas in 1980/81 wine was the most popular beverage consumed. Further examination of the types of beverages consumed by women who stopped drinking after learning they were pregnant in 1980/81 revealed that the use of liquor had dropped by 50 percent, the use of wine had dropped by 40 percent, whereas the use of beer had dropped by only 30 percent. The authors raise the question of whether beer drinkers are a different population. In this study beer drinkers were poorly educated and of lower social class. Those women who had decreased their alcohol consumption were more likely to be highly educated and older women. The decrease in alcohol consumption over the six year interval is attributed by the authors to the large two 42 year media campaigns and research conducted on Fetal Alcohol Syndrome in the Seattle, Washington area in the late seventies. The authors state that one could not conclude that the low incidence of heavy use of alcohol reported in 1980/81 in Seattle was typical of other cities in the United States. Prager et al. (1984) examined data from the National Natality and Fetal Mortality Surveys in the United States and described drinking patterns of married pregnant women who delivered live births in 1980. A sample of 4,405 women responded to a mailed questionnaire sent to them six months after delivery. The response rate was 56 percent. Those who consumed alcoholic beverages were more likely to be white women, older than twenty-five, and to have more than a high school education. Before pregnancy, 55 percent of the women in this study had consumed alcoholic beverages. More specifically, 58 percent of white women, 40 percent of hispanic women, and 39 percent of black women had consumed alcoholic beverages prior to becoming pregnant. The amount of alcohol consumed also had decreased during pregnancy. Prior to pregnancy, 45 percent of the sample had abstained from drinking or had consumed less than one drink per month, 39 percent had consumed less than an average of three drinks per week, and 16 percent had consumed an average of three or more drinks per week. During pregnancy, 61 percent had abstained from drinking or had consumed less than one drink per month, 36 percent had consumed less than an average of three drinks per week, and 3 percent had consumed an average of three or more drinks per week. Nearly 30 percent of white and hispanic women had stopped drinking during pregnancy, and almost 40 percent of black women had stopped drinking during pregnancy. Older women were less likely to stop drinking than younger women. However no significant differences by education were observed in the portion who had stopped drinking. 43 Lillien, Huber, and Rajala (1982) reported on dietary and ethanol intakes of 889 pregnant women who gave birth in a Massachusetts maternity hospital during a nine month period in 1979-80. Almost all the women participating in this study, 91 percent, had consumed alcohol before becoming pregnant. Alcohol intake was reported as number of ounces of ethanol consumed per month. One ounce of ethanol is equivalent to approximately twenty-two ounces of beer, eight ounces of wine, or two and one-half ounces of an 80-proof liquor. Prior to pregnancy, 9 percent had abstained from consuming alcohol, 61 percent had consumed between 0.1-9.9 ounces of ethanol per month, 14 percent had consumed between 10.0-19.9 ounces of ethanol per month, and 16 percent had consumed 20.0 or more ounces of ethanol per month. During pregnancy, 18 percent had abstained from consuming alcohol, 76 percent had consumed between 0.1-9.9 ounces of ethanol per month, 3 percent had consumed 10.0-19.9 ounces of ethanol per month, and 3 percent had consumed 20.0 or more ounces of ethanol per month. The authors reported that: 27 percent of teenagers had abstained from consuming alcohol during pregnancy, whereas only 17 percent of older women had abstained from consuming alcohol during pregnancy (p<0.10); 28 percent of unmarried women had abstained from consuming alcohol during pregnancy, whereas only 15 percent of married women had abstained from consuming alcohol during pregnancy (p< 0.005); and, 26 percent of women with less than twelve years of schooling had abstained from consuming alcohol during pregnancy, whereas 16 percent of women with more than twelve years of schooling had abstained from consuming alcohol during pregnancy (p< 0.025). Teenagers, unmarried women, and women with less than grade twelve education were more likely to abstain from drinking alcohol during pregnancy. Total amount of monthly alcohol consumption was related to age and income in the Lillien, Huber, and Rajala study (1982). Fourteen percent of women thirty years or older had consumed seven or more ounces of ethanol per month, whereas 44 only 8 percent of women under thirty years of age had consumed seven or more ounces of ethanol per month (p<0.10); and, 16 percent of women with yearly incomes of $20,000 or more had consumed seven or more ounces of ethanol per month, whereas only 8 percent of women with yearly incomes of less than $20,000 had consumed seven or more ounces of ethanol per month (p<0.025). Large monthly ethanol volumes were common among older and wealthier women. Kruse, Le Fevre, and Zwieg (1986) examined changes in smoking and drinking behaviors during pregnancy of 255 married women residents of Callaway County, Missouri. The population of the study was obtained from birth certificate data provided by the State of Missouri. A questionnaire was mailed to these women three to twelve months following delivery. A 69 percent response rate was obtained. Nonrespondents were younger and had less education. As part of the study, women had been asked whether they had consumed alcoholic beverages before and during pregnancy. Forty-nine percent of the respondents reported consuming at least one drink before pregnancy. Prior to pregnancy, 51 percent of the women had abstained from consuming alcohol, 31 percent had consumed one drink per week, 13 percent had consumed between two and four drinks per week, 4 percent had consumed between five and nine drinks per week, and 1 percent had consumed more than nine drinks per week. During pregnancy, 77 percent of the women had abstained from consuming alcohol, 19 percent had consumed one drink per week, and 4 percent had consumed between two and four drinks per week. No one in the sample reported having consumed more than four drinks per week during pregnancy. In comparison to other research studies reviewed in this dissertation, a larger percent of the married women in the Kruse study had abstained from drinking during pregnancy. Of the women who drank prior to pregnancy, eighty-nine individuals said they had decreased their alcohol consumption during pregnancy. Of 45 these eighty-nine women, sixty-six individuals or 53 percent of drinkers said they had stopped drinking completely. No association was found between decreasing alcohol intake and education level. In comparing women who drank before pregnancy with non-drinkers, no significant association was found between alcohol consumption and age, education level, or income. In the Kiss, Le Fevre, and Zweig study women were also asked to rate on a scale of one to five the importance of various factors which had influenced their reduction of alcohol consumption during pregnancy (one = not at all important, two = not important, three = neutral, four = important, and five = very important). Fear for infant's health was cited as important or very important by 95 percent of the women; advice from the doctor was cited as important or very important by 67 percent of the women; printed media was cited as important or very important by 53 percent of the women; advice of family was cited as important or very important by 47 percent of the women; and, alcohol making patients sick was cited as important or very important by 17 percent of the women. The authors conclude that fear for their infant's health was the most important factor in the woman's decision to curtail drinking and various sources of information seemed less important. This conclusion may in fact be misleading since the authors have combined health beliefs (fear for infant's health) and sources of information (advice from physicians, family, and friends) in their measurement tool of examining the importance of various factors which influence women's reduction in alcohol intake. Without a more in-depth examination of these factors, it is not known whether fear for the infant's health originated from the various sources of information or came from some other undesignated source. Kiss (1983) examined alcohol consumption of 220 prenatal class participants in Vancouver, and reported that 58 percent of participants had consumed alcohol before 46 becoming pregnant. Prior to pregnancy, 42 percent had abstained from consuming alcohol, 11 percent had consumed less than one drink per week, 33 percent had consumed between one and six drinks per week, 12 percent had consumed between seven and fourteen drinks per week, and 2 percent had consumed more than fourteen drinks per week. During pregnancy, at the time of completion of the health habits questionnaire, 52 percent had abstained from consuming alcohol, 18 percent had consumed less than one drink per week, 28 percent had consumed between one and six drinks per week, 2 percent had consumed between seven and fourteen drinks per week. No one had consumed more than fourteen drinks per week during pregnancy. Overall, 16 percent of drinkers had quit consuming alcoholic " beverages during pregnancy, 82 percent had reduced their alcohol intake, and 2 percent had not changed their alcohol intake. Women had also been asked to check off from a given list, the factors which had influenced a change in their alcohol intake. Ninety percent had checked their own personal knowledge, 43 percent had checked books-magazines-pamphlets, 36 percent had checked a change in cravings, 22 percent had checked the doctor, 17 percent had checked prenatal classes, 12 percent had checked the public health nurse, 7 percent had checked the media, 6 percent had checked social pressure, and 3 percent had checked a change in personal stress levels. Knowledge and print materials were important factors which had influenced women to change their alcohol consumption during pregnancy. In summary, research about alcohol consumption behaviors during pregnancy indicates that 49 to 91 percent of pregnant women had consumed alcohol prior to becoming pregnant. During pregnancy, 15 to 77 percent of women had abstained from consuming alcohol and 18 to 98 percent had reduced their intake. The amount of alcohol consumed prior to and during pregnancy varies from study to study. Up to 18 percent of women had consumed large amounts of alcohol prior to pregnancy, a period of time when women usually do not know they are pregnant, and up to 47 12 percent of women had consumed large amounts of alcohol during pregnancy. Women who were older and of a higher education level were more likely to consume large amounts of alcohol prior to pregnancy. However, mixed results were noted for association between demographic information and reduction of or abstention from alcohol during pregnancy. One study reported no associations between demographics and reduction, another study reported that better educated and older women were more likely to reduce their alcohol consumption during pregnancy, and yet another study reported that teenagers, unmarried women, and women with less education were more likely to abstain from consuming alcohol during pregnancy. Some studies found that women who consumed alcohol prior to pregnancy were more likely to be white, older, and of a higher education level, whereas other studies found no association between demographics and alcohol consumption. Factors which had influenced women to change their alcohol intake during pregnancy included: infant health, personal knowledge, reading material, cravings, and the physician. All in all the research results in this area are inconclusive. Learning about Alcohol: Media Campaigns, Intervention Strategies, and Information  Sources In 1984, "National Fetal Alcohol Syndrome Awareness Week" was declared in the United States in recognition of the potential for serious consequences of fetal exposure to alcohol, and in the interest of increasing both public and professional awareness of the preventability of these consequences (U.S. Department of Health and Human Services, 1984). In 1985, The National Institute on Alcohol Abuse and Alcoholism conducted a U.S. nationwide media campaign based on the message that the safest choice was not to drink during pregnancy (Worthington-Roberts, Vermeersch, and Williams, 1985). Another major campaign launched in New York, sponsored by the governor's task force, was built on the message that to be 48 perfectly safe women should abstain from alcohol consumption during pregnancy. Television and radio public service announcements and printed materials as well as suggested content for local radio and television shows were distributed (Worthington-Roberts, Vermeersch, and Williams, 1985). In Seattle, a two year media campaign was conducted in the early eighties. The campaign had included access to a twenty-four hour hot line on fetal alcohol and drug effects and an open clinic for pregnant women and mothers to discuss offspring problems related to alcohol (Streissguth et al., 1983). Currently in Ontario (Newbery, 1987) the Addiction Research Foundation has a toll free number to call for confidential information about alcohol or drugs such as cocaine and marijuana. There are more than sixty audio tapes to listen to and two of these tapes include "The Effects of Alcohol on an Unborn Child" and "Drugs and Pregnancy." Public awareness and knowledge about the risks of drinking during pregnancy were examined in Multnomah County, Oregon (Little et al., 1981). The study consisted of a telephone survey of 550 households. A 78 percent response rate was obtained. One of the research questions asked was "What beverages might a pregnant woman drink which could have an undesirable effect on her unborn child?" Ninety percent included one type of alcoholic beverage as their response. One fourth of all respondents who felt alcohol was harmful said that women should abstain from drinking during pregnancy. The other respondents stated that a safe level of alcohol was three drinks per day (mean value). Younger respondents, under twenty years of age, were more likely to endorse higher levels of drinking. Male respondents recommended significantly higher levels of alcohol than female respondents. Seventy-three percent of respondents who felt alcohol was harmful could name at least one effect on the fetus. There was, however, no significant relationship between the ability to name an effect of alcohol on the fetus and the ability to identify a safe level of use. 49 In 1985, the National Health Interview Survey was conducted in the United States (Williams, Dufour, and Bertolucci, 1986) on a probability sample of civilian noninstitutionalized individuals. The sample consisted of 87,649,000 females and 78,142,000 males (weighted data). It was found that 80 percent of female respondents, aged eighteen to forty-four years, knew that heavy drinking increased the chance of adverse pregnancy outcomes, and 62 percent of female respondents had heard of fetal alcohol syndrome. However a little more than 70 percent of those who had heard of Fetal Alcohol Syndrome described it as a newborn addicted to alcohol rather than a child born with certain birth defects. With regard to intervention strategies, Larsson (1983) conducted a study at four Maternal Health Clinics in Sweden to identify maternal alcoholic abusers, and to offer treatment for problem drinkers. Three of the four clinics were situated in socially deprived suburbs in Stockholm. Results showed that 89 percent of pregnant women were classified as occasional drinkers, defined as drinking an average of less than 30 grams of pure alcohol per day; 7 percent of the pregnant women were classified as excessive drinkers, defined as drinking an average of 30 to 125 grams of pure alcohol per day; and, 4 percent of the pregnant women were classified as alcohol abusers, defined as drinking an average of greater than 125 grams of pure alcohol per day. In contrast to Lillien, Huber, and Rajala (1982), Larsson (1983) states that excessive drinkers cannot be identified by socioeconomic characteristics, and therefore a profile on drinking habits of all pregnant women should be obtained. The author also concludes that a discussion about alcohol and the effects on the fetus was sufficient to persuade women who were not addicted to alcohol to change their drinking habits. Alcoholic women, on the other hand, required intensive counselling. Percentages of women who had stopped drinking or had reduced their intake during pregnancy as a result of attending the clinics was, unfortunately, not reported. 50 Therapy for women drinking heavily during pregnancy was integrated with routine prenatal care at the Boston City Hospital women's clinic (Rosett, Weiner, and Edelin, 1983). Heavy drinking was defined as consuming more than forty-five drinks a month with at least five drinks on some occasions. From the survey conducted between 1974 and 1979 at Boston City Hospital, 162 of the 1,711 women in the study reported heavy drinking. Of these 162 women, forty-nine continued to receive prenatal care at the Boston City Hospital. Of the forty-nine women in the treatment program, thirty-three women (67 percent) abstained or markedly reduced alcohol intake before the third trimester. Of these thirty-nine women, nineteen attained total abstinence. In this study, women who reduced their alcohol intake were younger and nulliparous. Frequency and quantity of alcohol intake were not predictive of therapeutic success. The authors classified problem drinkers as those with a social problem, symptom problem, or dependence problem. Women experiencing social drinking problems drank primarily due to social pressures from friends and relatives. Women experiencing symptom problems drank primarily to relieve a range of psychological symptoms and to alter mood and perceptions. For many, pregnancy was a time that exacerbated stressful situations in their lives. The added physicial and social responsibilities created ambivalent feelings toward motherhood. Women experiencing alcohol dependency problems drank between one-half and one liter of liquor or its equivalent per day. Rosett, Weiner, and Edelin (1983) state that intensive counselling is needed for alcoholics since few alcoholics appear to change their patterns in response to mass media campaigns. With regard to women's sources of alcohol information, Black (1983) interviewed twenty-five women attending a clinic for pregnant women in Leicester, England. The interview took place during the fourth month of pregnancy. Sixteen of the twenty-five women had consumed an alcoholic beverage during the week prior to the interview. Thirteen of the twenty-five women stated that they had received 51 information about alcohol during their pregnancy. Their sources of alcohol information were newspapers, magazines, television, radio, friends, doctors, midwives, and health visitors. The majority had obtained their information from the media, and only two women reported having obtained information from a physician or other health professional. Kruse, Le Fevre, and Zwieg (1986) in their study examining smoking and drinking behaviors of 255 married women in Callaway County, Missouri, reported that 75 percent of the women had discussed their smoking and drinking habits at some point during pregnancy with their physician. In summary, learning about alcohol has primarily focused on media campaigns encouraging pregnant women not to consume alcohol during their pregnancy. Intervention strategies for pregnant problem drinkers include intensive counselling sessions as it appears alcoholic women do not respond to media campaigns in the same way as non-alcoholic women. From the few studies which have examined sources of information on alcohol, it appears that health professionals do not dominate as a principal resource reaching the total pregnant population. Tobacco Use: Behaviors and Learning Research studies which have examined smoking behaviors among pregnant women, and identified effectiveness of intervention strategies in assisting pregnant women to quit smoking are not new. As with alcohol, very few studies have reported women's sources of information about smoking. Behaviors and learning about tobacco use are described as follows. Behaviors: Tobacco Use The incidence of smoking among pregnant women varies with their ethnic origin, age, marital status, and level of education. Mcintosh (1984) conducted a literature review of smoking incidence during pregnancy including twenty-eight 52 published research studies from 1953 to 1976 from Canada, U.S.A., Great Britain, Europe, Australia, and Israel. He reported that of the total 225,025 women in these studies, 42 percent smoked during their pregnancy. The percentage of smokers ranged from 14 percent in American Orientals to 61 percent in U.S. whites. The quantity of cigarettes smoked was not reported. Overall, the percentage of pregnant women who smoke is lower than the percentage of pregnant women who drink. Prager et al. (1984) examined data from the 1980 National Natality and Fetal Mortality Surveys in the United States and described smoking behaviors before and during pregnancy of married women older than twenty years of age giving birth to live infants. A sample of 4,405 women responded to a mailed questionnaire sent to them six months after delivery. The response rate was 56 percent. Smokers were typically white, under twenty-five years of age, and had a high school education or less. The prevalence of smoking before pregnancy was highest among white mothers at 32 percent; compared with black mothers at 25 percent; hispanic mothers at 23 percent; and, other races at 20 percent. Overall, 69 percent of the women participating in the survey had abstained from smoking before pregnancy, and 31 percent had smoked. More specifically, 9 percent had smoked between one and ten cigarettes a day, and 22 percent had smoked eleven or more cigarettes a day. During pregnancy, 75 percent of the women in the sample had abstained from smoking, 12 percent had smoked between one and ten cigarettes a day, and 13 percent had smoked eleven or more cigarettes a day. Over 10 percent of women are heavy smokers during their pregnancy. Of those women who smoked one to ten cigarettes a day prior to their pregnancy, 31 percent had stopped smoking during pregnancy, 66 percent had continued to smoke between one and ten cigarettes a day during pregnancy, and 3 percent had increased their smoking to eleven or more cigarettes a day during pregnancy. Of those women who smoked eleven or more cigarettes a day prior to their pregnancy, 12 percent had stopped smoking during 53 pregnancy, 27 percent had reduced their cigarette smoking to less than eleven cigarettes a day during pregnancy, and 61 percent continued to smoke eleven or more cigarettes a day during pregnancy. Fewer women who smoked more than ten cigarettes a day before pregnancy were able to abstain or reduce their cigarette smoking during pregnancy. Overall 18 percent of smokers had stopped smoking during pregnancy. The tendency to stop smoking during pregnancy was directly related to educational level. In 1980 Cresswell-Jones (1983) conducted a study of smoking patterns among pregnant women in Simcoe, Ontario. An anonymous self-reporting questionnnaire was completed by 132 volunteer pregnant women who were eighteen to thirty-five years of age, of high educational level, and from a wide range of occupations. Forty-nine percent of women had smoked before pregnancy, and this figure decreased to 35 percent during pregnancy. The two major reasons given by the pregnant women for the decrease in cigarette smoking were: 1) less desire, and 2) concern for the baby. Perceptions of the women about smoking were also investigated. The women were asked "In your opinion, can your baby be affected by your tobacco use?" The response was as follows: 80 percent said yes, 11 percent said no, 6 percent said maybe, and 3 percent didn't know. Fairly similar responses were obtained in a 1978 study by Luce and Schweitzer (Fielding and Yankauer, 1978b), who reported that two thirds of young nonpregnant women smokers believed that smoking can harm the newborn. Although the majority of women are aware that smoking can affect the newborn, over 20 percent are not aware of the hazards of smoking during pregnancy. Smoking incidence during pregnancy in Seattle, Washington decreased slightly from 1974/75 to 1980/81. Streissguth et al. (1983) evaluated changes in smoking and drinking patterns of Seattle women over this six year interval. The subjects were 54 two cohorts of pregnant women who were interviewed at six months gestation at two Seattle hospitals. Response rates were not reported. Sample size for 1974/75 was 1,529 women, and for 1980/81 was 1,413 women. In 1974/75 the incidence of smokers was 25 percent and in 1980/81 the incidence of smoking had decreased to 22 percent. During pregnancy in 1974/75, 75 percent had not smoked, 14 percent had smoked less than sixteen cigarettes a day and 11 percent had smoked sixteen or more cigarettes a day. During pregnancy in 1980/91, 78 percent had not smoked, 16 percent had smoked less than sixteen cigarettes a day and 6 percent had smoked sixteen or more cigarettes a day. The authors report that the incidence of smokers in this study is low in comparison with other previously reported research studies. A 1983 population based hospital survey of 3,628 women in their postpartum period was conducted in the Ottawa-Carleton region of Ontario (Stewart and Dunkley, 1985). The authors reported that this was the first population based study to be conducted in Canada to determine: the incidence of smoking before and during pregnancy; the demographic and socioeconomic characteristics of smoking pregnant women; and, the use of health care services by pregnant women. The sample included 3,296 women, and a 91 percent response rate was obtained. A self-report questionnaire was used which had a test-retest reliability coefficient value of 0.90. The percentage of smokers before pregnancy was 37 and at the time of delivery was 26 percent. Of the smokers, 39 percent reported that they had not changed their smoking habits, 31 percent reported that they had stopped during their pregnancy, 28 percent said they had decreased the amount smoked, and 2 percent stated they had increased the amount smoked. Nearly 40 percent of smokers reported that they had not changed their level of cigarette use during pregnancy. Changing smoking behaviors during pregnancy appears to be a problem among pregnant women. The highest cessation rates were among women who had smoked only one to five cigarettes per day before pregnancy. Most women who stopped smoking did so 55 as soon as they found out they were pregnant. Of the women who smoked prior to pregnancy, 13 percent had reported smoking between one and five cigarettes a day, 31 percent had reported smoking between six and fifteen cigarettes a day, 47 percent had reported smoking between sixteen and thirty cigarettes a day, and 9 percent had reported smoking more than thirty cigarettes a day. During pregnancy, 31 percent of smokers had reported quitting smoking, 4 percent of smokers had reported smoking between one and five cigarettes a day, 20 percent of smokers had reported smoking between six and fifteen cigarettes a day, 38 percent of smokers had reported smoking between sixteen and thirty cigarettes a day, and 7 percent of smokers had reported smoking more than thirty cigarettes a day. A large percentage of women continued to smoke heavily during pregnancy, more specifically, 45 percent continued to smoke over fifteen cigarettes per day. Assisting women to change their smoking behaviors during pregnancy remains a challenge for health professionals. A step-wise regression analysis was used to identify factors associated with cessation of smoking during the first trimester of pregnancy. Fewer number of cigarettes smoked before pregnancy, higher level of education, less use of alcohol before pregnancy, multiparity, and less use of marijuana before pregnancy were found to be associated with smoking cessation. Subgroups which had particularly high rates of smokers were teenagers, single women, women living common law, and women with lower education levels. Among the primiparous women, prenatal classes were attended by 62 percent of smokers, compared with 86 percent of non-smokers. On a smaller sample of 430 pregnant women classified as nutritionally high risk referred to the Prince Edward Island Department of Health and Social Services, McDonald and Newson (1986) reported that 60 percent were smokers. Reduction of the incidence of smoking was not reported. Nutritionally high risk women had been defined as women who had had one of the following conditions: underweight status at conception; low weight gain during pregnancy; poor obsterical record during a previous 56 pregnancy; pernicious vomiting; medical problems with nutritional implications (e.g. diabetes, heart disease); personal stress; or, expecting twins. Kruse, Le Fevre, and Zwieg (1986) examined changes in smoking and drinking behaviors during pregnancy of 255 married women residents of Callaway County, Missouri. The population of the study was obtained from birth certificate data provided by the State of Missouri. A questionnaire was mailed to these women three to twelve months following delivery. A 69 percent response rate was obtained. Nonrespondents were younger and had less education. Twenty-eight percent of the respondents had reported smoking at least one cigarette a day before pregnancy. Prior to pregnancy, 72 percent of the women in the sample had not smoked, 10 percent had smoked between one and nineteen cigarettes a day and 18 percent had smoked more than nineteen cigarettes a day. During pregnancy, 11 percent of the women in the sample had not smoked, 13 percent had smoked between one and nineteen cigarettes a day, and 10 percent had smoked more than nineteen cigarettes a day. Of the women who smoked before pregnancy, 33 percent had decreased their cigarette smoking during pregnancy and only 17 percent had stopped completely. It was not reported whether the remaining 50 percent of smokers had changed their smoking habits. As has been previously noted, cigarette smoking is a major health concern during pregnancy. Women more likely to smoke were under thirty years of age, had lower incomes, and had lower educational levels. Women who were more likely to decrease their cigarette smoking were those who had had some education beyond high school. In this study women were also asked to rate on a scale of one to five the importance of various factors which had influenced their reduction in cigarette smoking during pregnancy (one = not at all important, two = not important, three = neutral, four = important, and five = very important). Fear for infant's health was cited as important or very important by 96 percent of the women; advice from the doctor was cited as important or very important by 67 percent of the women; 57 advice of family was cited as important or very important by 50 percent of the women; printed media was cited as important or very important by 46 percent of the women; media was cited as important or very important by 38 percent of the women; and, smoking making patients sick was cited as important or very important by 33 percent of the women. Health concern was a major factor in influencing women to reduce their cigarette smoking during pregnancy. Advice from the physician had also played a major role. Kiss (1983) examined smoking behaviors of 220 prenatal class participants in Vancouver, and reported that 31 percent of participants had smoked before becoming pregnant. Prior to pregnancy, 69 percent had not smoked, 7 percent had smoked between one and six cigarettes a day, 16 percent had smoked between seven and twenty-four cigarettes a day, and 8 percent had smoked more than twenty-four cigarettes a day. During pregnancy, at the time of completion of the health habits questionnaire, 85 percent had not smoked, 5 percent had smoked between one and six cigarettes a day, 9 percent had smoked between seven and twenty-four cigarettes a day, and 1 percent had smoked more than twenty-four cigarettes a day. Overall, 91 percent of smokers had decreased their cigarette smoking, 6 percent had not changed their cigarette smoking, and 3 percent had increased their cigarette smoking. In comparison to other research studies which examined smoking behaviors during pregnancy, women participating in the Kiss study had substantially changed their smoking behaviors. Women had also been asked to check off from a given list, the factors which had influenced a change in their cigarette smoking. Eighty-seven percent had checked their own personal knowledge, 46 percent had checked books-magazines-pamphlets, 46 percent had checked the doctor, 30 percent had checked family members and friends, 24 percent had checked a change in cravings, 22 percent had checked prenatal classes, 19 percent had checked social pressures, 14 percent had checked the public health nurse, and 3 percent had checked stress. 58 Knowledge and print material had primarily influenced women to change their smoking behavior. Although alcohol and smoking behaviors are reported in this dissertation separately, there is some relationship between the two behaviors. Prager et al. (1984) reported that before pregnancy 35 percent of women were both non-smokers and non-drinkers; 21 percent were both smokers and drinkers; 34 percent were drinkers only; and, 10 percent were smokers only. During pregnancy 47 percent of women were both non-smokers and non-drinkers; 12 percent were both smokers and drinkers; 27 percent were drinkers only; and, 14 percent were smokers only. Although the prevalence of drinking was higher than the prevalence of smoking, reduction in drinking was more pronounced than reduction in smoking. Norman (1985) studied the inter-relationships among health behaviors of 412 university students at Memorial University in Newfoundland. Statistically significant correlations between smoking and alcohol behaviors were found. Similar results were also found in the National Health Interview Survey conducted in the United States (Williams, Dufour, and Bertolucci, 1986). Although a relationship between smoking and drinking exists, Fox, Sexton, and Hebel (1987) reported that enrollment in a smoking cessation intervention program for pregnant smokers had not influenced these women's alcohol intake, although changes in their smoking habits were documented. In summary, the results of the studies which have investigated prenatal smoking behaviors indicate that smoking during pregnancy remains a challenge for the health professional. Approximately 30 to 40 percent of pregnant women smoke prior to pregnancy, and the incidence varies with ethnic background, education level, marital status, and age. Approximately 15 to 35 percent of women continue to smoke during their pregnancy. The number of cigarettes smoked was also high, and varied from study to study. Up to 56 percent of smokers smoked more than sixteen 59 cigarettes a day before pregnancy and up to 45 percent of smokers smoked more than sixteen cigarettes a day during pregnancy. Smoking is a major health concern during pregnancy. Those women who are most likely to abstain or decrease their cigarette smoking are those who smoked fewer cigarettes at the beginning of pregnancy, and women of a higher education level. Smokers were more likely to be of a lower education level, lower socioeconomic level, white, unmarried, and less likely to attend prenatal classes. Factors which had influenced women to change their cigarette smoking included: concern for infant health, reading print material, and receiving advice from a doctor, family members, and friends. Learning about Tobacco: Intervention Strategies and Information Sources Although the negative effects of smoking during pregnancy had been documented prior to 1970, in 1978 Fielding and Yankauer (1978b) reported that relatively few efforts focused on the pregnant woman as a target group for smoking intervention. One early smoking intervention study was conducted by Donovan (1977) who measured the effects of intensive individual anti-smoking advice given by physicians during prenatal visits at maternity units in the hospital setting. Women who were eligible for the study had to smoke more than five cigarettes a day, had to be less than thirty weeks pregnant, and had to be less than thirty-five years of age. These women were randomly assigned to a control group, which received antenatal care usually provided for in the hospital and which included anti-smoking advice routinely given; and, to a treatment group, which received intensive individual anti-smoking advice from a physician. Two hundred and sixty-three women in the treatment group had received intensive anti-smoking advice and this group's mean number of cigarettes decreased from twenty per day before pregnancy to nine per day by the end of the pregnancy. The mean number of cigarettes smoked in the control group was eighteen per day before pregnancy and sixteen per day at the end 60 of the pregnancy. The author, however, reports that any interpretation of these findings was made uncertain by the biased recall identified in the treatment group. One hundred and seventy-four women in the treatment group had been interviewed postnatally, and of these women fifty-four admitted smoking at the end of their pregnancy although they had originally reported that they had quit. Another early smoking cessation program was conducted by Danaher, Shisslak, Thompson, and Ford (1978) with eleven volunteer pregnant women. Participants were involved in six, two-hour group sessions over a seven week period. All women received: a document outlining the risks of smoking during pregnancy; self monitoring tasks; deep muscular relaxation techniques; and, coping strategies for smoking urges. The range of cigarettes smoked at the onset of pregnancy was fifteen to thirty-five cigarettes daily. The minimum and the maximum number of cigarettes smoked daily were reduced to zero to twenty-five cigarettes by the first week of the program, and to zero to twenty cigarettes by the seventh week of the program. Nine of the eleven women had reported reducing the number of cigarettes smoked once they became pregnant, and further reductions were observed during the course of the program, except for three participants. The authors acknowledged that these results are based on a small number of volunteers, that there was no control to measure "spontaneous" quitting during pregnancy, and that there was no control to assess effects of the intervention strategy. More recently, Langford, Thompson, and Tripp (1983) evaluated the impact of health education on smoking during pregnancy among smokers attending prenatal classes in Metropolitan Toronto between 1977 and 1978. A total of 116 women participated in the study, and this figure represents 79 percent of the total number of women attending the classes. The participants were in their seventh month of pregnancy. The control group attended the regular prenatal class series. The 61 experimental group attended the regular prenatal class series and received an additional one-half hour presentation on smoking and a pamphlet on smoking during pregnancy. The experimental group had originally consisted of two experimental groups. In addition to the aforementioned activities, women in the second experimental group had been scheduled to receive a follow-up home visit by the public health nurse to reinforce the class presentation. However it was difficult to ensure that all members in the second experimental group had received the home visit and therefore the researchers decided to amalgamate the two experimental groups for data analyses. Therefore some of the women in the experimental group had a follow-up home visit by the public health nurse whereas others did not. The results of the study did not reveal a significant difference between the control and treatment groups until one year after delivery when 23 percent of the experimental group were non-smokers, compared to 5 percent of the control group. The authors tried to account for the fact that no significant differences were found between the control and the experimental groups by stating that the women who were still smoking by the seventh month of pregnancy appeared to be recalcitrant smokers. Giving these women information about the health effects of smoking during the seventh month of pregnancy had provoked angry reactions from women or their partners because they had not been given the information earlier. During pregnancy, approximately three-quarters of all women had reported some smoking reduction or cessation, and most of these women had reduced their smoking by at least 50 percent. Recently, encouraging results on smoking intervention strategies have been obtained in well controlled studies by Sexton and Hebel (1984), and Windsor et al. (1985). Sexton and Hebel (1984) conducted a prospective, randomized, controlled experiment that was designed to examine the effect of smoking on birth weight 62 outcome. Women eligible for the study were those who had smoked more than ten cigarettes a day before their pregnancy and had not passed their eighteenth week of gestation. The majority of women were recruited from fifty-two private obstetrical practices and a large university hospital in Baltimore. A total of 935 women were enrolled over a two and one-half year period, and of those enrolled 463 were randomly assigned to an intervention treatment group. The intervention strategy described was intensive, and included numerous activities such as: personal contact; group contact; mail contact; telephone contact; monetary rewards; and, receiving "quit packets" containing: information on smoking cessation techniques and health risks associated with smoking during pregnancy, chewing gum as an oral substitute for cigarettes, and a pencil to record number of cigarettes smoked daily. No comparative evaluation of the different kinds of activities was made. A complete account of the intervention strategy was described by Nowicki et al. (1984). The mean number of cigarettes smoked before pregnancy was reported to be twenty-one cigarettes per day for the total sample, and eleven cigarettes per day for the total sample at the time of randomization. Therefore a reduction in cigarette consumption was noted prior to the intervention program. The results of the intervention strategy indicated that 20 percent of the control group had quit smoking and 43 percent of the treatment group had quit smoking. Furthermore 14 percent of the control group had reported smoking more than twenty cigarettes a day; whereas, only 4 percent of the treatment group had reported smoking more than twenty cigarettes a day. The mean number of cigarettes smoked per day in the control group was thirteen and the mean number of cigarettes smoked in the treatment group was six (p<0.01). Self-reports of smoking behaviors were confirmed by salivary thiocyanate levels. Windsor et al. (1985) conducted a study to determine the effectiveness of three smoking cessation methods on women attending Public Health Maternity Clinics in Birmingham, Alabama. Of the 1,838 women attending the clinics between October 63 1983 and September 1984, 25 percent were identified as smokers. Eighty percent of the smoking pregnant women agreed to participate in the study. The pregnant smokers were randomly assigned to three groups. Group one was the control, receiving smoking cessation advice routinely given at maternity clinics. In addition to routine advice given about the dangers of smoking during pregnancy, group two received a booklet on the dangers and risks of smoking during pregnancy; attended a ten minute counselling session; and, were taught how to use the "Freedom from Smoking Program Manual" of the American Lung Association. This manual consisted of a seventeen day self-directed plan to quit smoking. Group three received the same information booklet and counselling session as group two, but were taught to use a different self-help manual "A Pregnant Woman's Guide to Quit Smoking." A statistically significant difference was found between quitting rates and decreases in cigarettes smoked with the control group and groups two and three; however, no difference was found between the groups using different self-help manuals. Nine percent of group one (control) had quit or had reduced their cigarette smoking; 20 percent of group two had quit or had reduced their cigarette smoking; and, 31 percent of group three had quit or had reduced their cigarette smoking. Oral reports of smoking behavior were confirmed by salivary thiocyanate levels and only 3 percent of the women provided inaccurate information. Very few studies have addressed women's sources of information about tobacco use. Kruse, Le Fevre, and Zweig (1986) documented that 75 percent of 255 married women in Callaway County, Missouri had discussed their smoking and drinking habits at some point during pregnancy with their physician. Lincoln (1986) however noted that women are getting mixed messages about smoking. He refers to Virginia Ernsler's comments: On one hand, women are told by health professionals that cigarettes are deleterious to health and reproduction, yet on the other hand, cigarette advertisements are associated with seductiveness and liberation. The Canadian Tobacco 64 Maufacturer's Council has voluntarily withdrawn broadcast advertising and adopted an industry code of practice, including health hazard warning labels. However, there is increased pressure that advertisement of tobacco be restricted by legislation. The steering committee for the National Program to Reduce Tobacco in Canada (1987) is advocating regulatory control over tobacco promotion and distribution. In the United States, it is required that four rotating warnings be included on cigarette packages and all advertising. One of these warnings includes the Surgeon General's Warning that smoking by pregnant women may result in fetal injury, premature birth, and low birth weight (Edwards, 1986). In summary, learning about tobacco has primarily focused on the effectiveness of smoking intervention strategies for pregnant women. An intensive comprehensive approach assisting pregnant women to quit smoking has been found to be effective. However messages from health professionals compete with advertising by tobacco manufacturers. Intensive comprehensive strategies are required to assist women to stop smoking as well as to stop drinking during pregnancy. Health Behaviors and Perinatal Outcomes It is well documented that inadequate weight gain, alcohol consumption, and tobacco use are associated with poor pregnancy outcomes (Institute of Medicine, 1985). The relationship between these three factors and perinatal outcome, including low birth weight, is presented so as to provide for a better understanding of the serious risks attached to women's health practices that are less than optimal, and to provide the basis for what pregnant women should know about weight gain, alcohol consumption, and tobacco use. 65 Weight Gain and Perinatal Outcome Prior to 1970 it was common medical practice to minimize weight gain during pregnancy so as to promote an easy delivery. Furthermore it was believed that excessive weight gain led to the development of toxemia and other obstetrical problems. However, by the seventies, it had become apparent that severe dietary restrictions during pregnancy resulted in a higher risk of having a low birth weight infant (Taffel and Keppel, 1986). A major turning point was the review of the literature conducted by the Committee on Maternal Nutrition of the National Research Council (1970). The committee members in their report concluded that a gain of twenty to twenty-five pounds was associated with the most favorable outcome of pregnancy and that there was no scientific justification for routinely limiting weight gain to less than twenty pounds. The association between pregravid weight status and birth weight outcome was first documented by Tompkins, Wiehl, and Mitchell in 1955. If women at the beginning of pregnancy were 20 percent or more underweight, the incidence of low birth weight was 15.4 percent; if women at the beginning of pregnancy were 5 to 19 percent underweight, the incidence of low birth weight was 8.3 percent; if women at the beginning of pregnancy were at a weight appropriate for their height, the incidence of low birth weight was 5.7 percent; if women at the beginning of pregnancy were 5 to 19 percent overweight, the incident of low birth weight was 4.4 percent; and, if women at the beginning of pregnancy were greater than 20 percent overweight, the incidence of low birth weight was 3.8 percent. The classification of women according to their weight status at the beginning of pregnancy was based on the Metropolitan Life Insurance Height and Weight Tables of 1942. By examining pregravid weight status, maternal weight gain, and pre and postpartum weight status, Tompkins and associates found that underweight women had a greater 66 net body increment in weight after delivery than normal weight women. As such, underweight women had added weight to their own tissue mass at some sacrifice to the fetus. Simpson, Lawless, and Mitchell (1975) found similar associations. Naeye (1979) examined data from the Collaborative Perinatal Project which prospectively followed 53,518 pregnancies in twelve U.S. hospitals between 1959 and 1966. Women's prepregnant weight for height status was calculated using the Metropolitan Life Insurance Tables of 1959. If prepregnant body weight was less than 90 percent of the Metropolitan Life Insurance desirable weight values, the optimal weight gain for these underweight women was found to be thirty pounds to achieve the fewest fetal and neonatal deaths. If prepregnant body weight was between 90 and 135 percent of the desirable values, the optimal weight gain for these normal weight women was found to be twenty-four to twenty-seven pounds to achieve the fewest fetal and neonatal deaths. If prepregnant body weight was greater than 135 percent of the desirable values, the optimal weight gain for these overweight women was found to be fifteen to sixteen pounds to achieve the fewest fetal and neonatal deaths. Maternal caloric intake had the greatest effect on prenatal weight gain. Brown et al. (1981) examined 399 women's health records at a major hospital in St. Paul's Minnesota. The mean birth weight of infants born to mothers who were "very underweight," women weighing less than 80 percent of their standard weight, was 2,976 grams; low birth weight incidence was 12.9 percent. The mean birth weight of infants born to mothers who were "moderately underweight," women weighing between 80 and 90 percent of their standard weight, was 3,020 grams; low birth weight incidence was 16.5 percent. The mean birth weight of infants born to mothers who were of "normal weight," women weighing between 90 and 120 percent of their standard weight, was 3,234 grams; low birth weight incidence was 6.9 67 percent. All three groups of women had equivalent mean prenatal weight gains, 12.6 pounds, 12.9 pounds, and 12.6 pounds. The incidence of low birth weight was approximately twice as high among infants born to underweight women than to women of normal weight with the same weight gain. The incidence of low birth weight was 7.2 percent among infants born to women gaining more than twenty pounds, whereas the incidence of low birth weight was 27.8 percent among infants born to mothers gaining less than twenty pounds. An average weight gain of thirty-five pounds for the underweight woman was associated with a birth weight of 3,600 grams or more. The pattern of weight gain is also an important factor to consider. Van den Berg (1981), in examining data on 15,000 women who were members of the Kaiser Foundation Health Plan in California, found that a weekly weight gain during the second half of pregnancy of less than 0.5 pounds more than doubles the likelihood of low birth weight incidence among infants. The effect is even more pronounced when the prepregnant weight is less than optimal. The National Guidelines on Nutrition During Pregnancy (Health and Welfare Canada, 1987) state that optimal weight gain during pregnancy varies and is dependent on a woman's prepregnant weight for height. A practical range of gain for women is ten to fourteen kilograms (twenty-two to thirty-one pounds). A gain of one to three kilograms (two to seven pounds) is recommended by the end of the first trimester. A goal of four kilograms (nine pounds) by the twentieth week gestation is also suggested. Underweight women should be encouraged to attain their standard weight in addition to gaining normal pregnancy weight gain requirements. Weight gains of thirteen to sixteen kilograms (twenty-nine to thirty-five pounds) have been associated with good perinatal outcomes. For overweight women a gain of seven to nine kilograms (fifteen to twenty pounds) has been associated with good perinatal 68 outcomes. The B.C. Ministry of Health (McCarthy and Mackay, 1984) stress that weight gain varies with every individual. They state that the average weight gain for healthy primiparous women is 12.5 kilograms (27.5 pounds), and 11.6 kilograms (25.5 pounds) for multiparous women. Underweight women need to correct for their underweight status and a gain of 13.6 to 14.5 kilograms (30 to 32 pounds) is recommended. Overweight women do not need to gain as much weight during pregnancy as women of normal weight, and a gain of 7.3 to 9 kilograms (16 to 20 pounds) is sufficient. Recommendations for the rate of gain throughout the pregnancy are also provided. These recommendations include: a gain of one to two kilograms (two to four pounds) by the end of the first trimester, followed by a gain of 0.40-0.45 kilograms (0.88-0.99 pounds) per week with a deceleration of gain to 0.35 kilograms (0.77 pounds) per week toward the end of the pregnancy. The Society of Obstetricians and Gynaecologists of Canada, (Perinatal Medicine Committee, 1980) state that weight gain should approximate three kilograms (seven pounds) by the twentieth week gestation, followed by one kilogram (two pounds) every two weeks during the latter half of the pregnancy. A range of gain was not provided but rather caloric requirements during pregnancy are given. Restriction of weight gain through nutritional methods is contraindicated. In summary, inadequate maternal weight gain and low prepregnant weight for height are associated with a twofold increase in the risk of having a low birth weight infant. Gains associated with a healthy outcome include: eleven to fourteen kilograms (twenty-four to thirty pounds) for women of normal weight, fourteen to sixteen kilograms (thirty to thirty-five pounds) for underweight women, and seven to nine kilograms (fifteen to twenty pounds) for overweight women. By the end of the first trimester, a gain of one to two kilograms (two to four pounds) is suggested, 69 and by the end of the twentieth week gestation a gain of no less than four kilograms (nine pounds) is recommended. Weekly weight gains during the second and third trimesters range from 0.35 to 0.45 kilograms (0.77-0.99 pounds). A gain of less than 0.22 kilograms (0.5 pounds) per week during the second half of pregnancy is associated with an increased risk of having a low birth weight infant. Calorie restricting diets are not advised during pregnancy. Alcohol Consumption and Perinatal Outcome Although wine and beer have been staples of virtually all human societies, it is interesting that what we know about the effects of alcohol on the fetus has only recently been discovered (Fielding and Yankauer, 1978a). Alcohol has the biochemical potential to cause a wide range of effects on the fetus. It crosses the placenta freely and attains the same concentration in the fetus as in the mother (Rosett, Weiner, and Edelin, 1981). Exposure of the fetus at critical stages of development can result in malformation. During the first eight weeks of pregnancy, the formation and development of body organs from embryonic tissue occurs, and alcohol may alter the cell membranes and the embryonic organization of these tissues. Unfortunately, many women do not know they are pregnant when organ development occurs, a time period in which the embryo is particularly sensitive to the effect of toxic chemicals. Throughout gestation, the drinking of alcohol may retard cell growth and division. The third trimester is the time of the most rapid brain growth. Reduction of heavy alcohol consumption by mid-pregnancy can modify some of the adverse effects. While structural malformations occurring in early pregnancy will still persist, delays in growth may be reversible (Rosett, Weiner, and Edelin, 1983). The problems associated with excessive alcohol consumption are well documented and result in a condition called "Fetal Alcohol Syndrome." Infants born to mothers who abuse alcohol during pregnancy exhibit: anomalies of the eyes, nose, 70 heart, and central nervous system; growth retardation; small head circumference; and, mental retardation (Worthington-Roberts, Vermeersch, and Williams, 1985). The Fetal Alcohol Study Group of the Research Society on Alcoholism (Rosett, Weiner, and Edelin, 1981) concluded that diagnosis of Fetal Alcohol Syndrome should only be made if the infant has signs in each of the following categories: (1) growth retardation when corrected for gestational age (2) central nervous symptoms, and (3) two of three designated signs of facial dysmorphia. Fetal Alcohol Syndrome is not limited to infants of alcoholic women, but has also been identified in infants where the pregnant woman had consumed four to five drinks per day and an average of at least forty-five drinks per month (Iber, 1980). The effect of alcohol is varied by a combination of genetic susceptibility, maternal nutrition, and intensity of the insult. Cushner (1981) attributes the designation of the term Fetal Alcohol Syndrome to Jones and Smith in 1973. Since that time, several hundred cases have been reported in the literature. Depending upon the population studied, the incidence of Fetal Alcohol Syndrome has ranged from 1 in 300 to 1 in 2,000 live births, and 30-40 percent in infants of alcoholic mothers (Worthington-Roberts, Vermeersch, and Williams, 1985). Estimates as high as one in fifty live births have been made for some Indian Reservations in the United States (Rosett, Weiner, and Edelin, 1981). Wright et al. (1983) noted that in a sample of 900 women drinking more than 100 grams of alcohol a week (approximately ten drinks per week) there was a twofold risk of delivering a baby on or below the tenth centile of the Castleman Centile Weight Chart in comparison to those women who drank less than 50 grams a week. The authors recommend decreasing alcohol consumption before pregnancy to reduce the risk of low birth weight. In 1977 Little studied 263 pregnant women and found that ingestion of one ounce of absolute alcohol daily (approximately two drinks per day) one month before 71 pregnancy was associated with an average decrease in birth weight of 91 grams, and the ingestion of one ounce of absolute alcohol during the later part of pregnancy was associated with a decrease in birth weight of 160 grams. These associations were independent of tobacco use. In 1981 the Surgeon General of the United States issued an advisory warning on drinking during pregnancy, recommending abstinence for pregnant women (Streissguth et al., 1983). Guidelines prepared by