UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trajectories of childhood socioeconomic status and markers of cardiovascular health in adolescence Marin, Teresa J. 2006

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2006-0564.pdf [ 2.17MB ]
Metadata
JSON: 831-1.0092719.json
JSON-LD: 831-1.0092719-ld.json
RDF/XML (Pretty): 831-1.0092719-rdf.xml
RDF/JSON: 831-1.0092719-rdf.json
Turtle: 831-1.0092719-turtle.txt
N-Triples: 831-1.0092719-rdf-ntriples.txt
Original Record: 831-1.0092719-source.json
Full Text
831-1.0092719-fulltext.txt
Citation
831-1.0092719.ris

Full Text

Trajectories of Childhood Socioeconomic Status and Markers of Cardiovascular Health in Adolescence by Teresa J. Marin A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R OF A R T S i n T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Psychology) T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A August, 2006 ©Teresa J . Marin, 2006 11 A B S T R A C T Objective: The current study examined trajectories of socioeconomic status (SES) throughout childhood and their relationship to markers of cardiovascular health in adolescence. The goal was to determine whether early life SES, current SES, cumulative SES, and/or social mobility best explained the relationship between SES experiences across an adolescent's lifespan and current blood pressure, heart rate (HR), and body mass index ( B M i ) . Design: 102 medically healthy adolescents completed cardiovascular health assessments. Adolescents enrolled in the study with a parent. Parents reported on family SES, indicating the number of bedrooms in the family home for each year of the child's life. Main Outcome Measures: Systolic blood pressure (SBP), Diastolic blood pressure (DBP), H R , and B M I . Results: Using Jones, Nagin, and Roeder's (2001) semiparametric group-based method, four distinct trajectories of childhood SES were identified. Trajectory groups were differentially related to adolescents' S B P and D B P . Early life SES explained trajectory group differences in adolescents' SBP and D B P . Cumulative SES also contributed to differences in adolescents' D B P . Trajectories of childhood SES were unrelated to H R and B M I . Trajectory findings were confirmed in follow-up analyses. Conclusions: O f the life-course models that we tested, an early life SES model best explained adolescents' current blood pressure. These findings point toward early life developmental processes as potential candidates for explaining the relationship between SES and risk factors related to C V D . Interventions designed to reduce SES health disparities should take place early in a child's life. i i i T A B L E O F C O N T E N T S Abstract i i Table of Contents i i i List of Tables iv List of Figures v Acknowledgments v i Introduction 1 Method 5 Results 10 Discussion 18 References 25 iv LIST OF T A B L E S Table 1. Demographic and health characteristics of the sample 31 Table 2. Indices of socioeconomic status through child's life 32 Table 3. Selecting the number of trajectories 33 Table 4. Model selection for the shape of each trajectory 34 Table 5. Partial correlations between SES indices and adolescents' biological outcomes 35 V L I S T O F F I G U R E S Figure 1. Estimates of SES trajectories 36 Figure 2. Estimates of average S B P by trajectory group 37 Figure 3. Estimates of average D B P by trajectory group 38 A C K N O W L E D G M E N T S Thank you to Dr. Edith Chen and Dr. Gregory Mi l le r for their guidance and support throughout the preparation of this manuscript. 1 I N T R O D U C T I O N Socioeconomic status (SES) is an important determinant of health status at each phase of the life-cycle. Thus, among children, adults, and the elderly, individuals in lower SES groups experience higher rates of morbidity and mortality due to a wide range of medical conditions (Adler et al., 1994; Anderson & Armstead, 1995; Chen, Matthews, & Boyce, 2002). Life-course models of SES have proposed various pathways through which SES at different stages can influence health. A t least four major life-course models exist: critical period, current SES, cumulative SES, and mobility. The critical period and current SES models emphasize the importance of timing of SES exposure, whereas cumulative SES and mobility models emphasize dynamic aspects of SES across time. According to critical period models, there is a window of time in which SES exerts its most profound effects on the body. This critical period is thought to be in early childhood, when important developmental processes are underway. Early life environmental conditions may program a pattern of biological and behavioural responses that have a long-term impact (Hertzman, 1999; Barker, 1992). Research has demonstrated the significance of early life experiences in long-term health. For instance, early childhood environments predict adult cardiovascular disease ( C V D ) , stomach cancer, and hemorrhagic stroke (see Galobardes, Davey Smith, & Lynch, 2006; Galobardes, Lynch, & Davey Smith, 2004, for reviews). Importantly, these associations persist after accounting for adult SES. In contrast, current SES models specify paths through which socioeconomic experiences at any point in the life-cycle can influence concurrent health status. Such models suggest that the current living environment can have a relatively immediate impact on health. For instance, current l iving conditions may affect a person's access to healthcare or influence susceptibility to 2 acute medical conditions. Although few studies have examined the impact of current socioeconomic circumstances independent of the childhood environment, research evidence suggests that current SES is an important predictor of self-reported health (Laaksonen, Rahkonen, Martikainen, & Lahelma, 2005) and C V D mortality (Pensola & Martikainen, 2003). Models of cumulative risk focus on the additive effects of SES experience. Thus, individuals who are exposed to low SES for longer durations are thought to be at greater risk. Indeed, research evidence has shown that the amount of time spent in low SES is an important predictor of mortality (Davey Smith & Hart, 2002; McDonough, Duncan, Will iams, & House, 1997) and young adults' self-reported health (Power, Manor, and Matthews, 1999). Finally, mobility models propose that changes in SES over the life course wi l l affect health. Markers of SES like family income can fluctuate from year to year (Duncan, 1988). Thus, over any given period of time, a person may experience upward mobility or downward mobility, and these changes in social status may impact health. A few studies have examined these types of relationships. Income fluctuations have been associated with morality risk among middle-income adults (McDonough, Duncan, Will iams, & House, 1997). More recently, downward mobility in adulthood has been associated with increased risk o f hypertension (Matthews, Kiefe, Lewis, L i u , Sidney, & Yunis, 2002) and poorer self-reported health (Virtanen, Vahtera, Kiv imaki , Liukkonen, Virtanen, & Ferrie, 2005). Conversely, upward mobility in adulthood has been associated with cardiovascular risk reduction, but only among certain race-gender groups (Karlamangla et al., 2005). Previous research on the longitudinal relationship between SES and health has mainly focused on adult populations. Thus, we still know very little about SES processes throughout childhood and adolescence. Findings from the mental health literature suggest that children and adolescents are affected by both timing and dynamic SES indicators (Duncan, Brooks-Gunn, & Klebanov, 1994; Duncan, Yeung, Brooks-Gunn, & Smith, 1998; Mcleod & Shanahan, 1996). Specifically, early childhood experiences of poverty influence future mental health outcomes; regardless of subsequent changes in the child's SES environment (Duncan, Yeung, Brooks-Gunn, & Smith, 1998). Furthermore, the experience of persistent poverty is associated with worse mental health outcomes than the experience of transient poverty or no poverty (Mcleod & Shanahan, 1996). Clearly, patterns of SES over time have implications for children's mental health; and it is also important to test whether these findings extend to markers of children's physical health. This line of research would have important implications for designing effective interventions to minimize health disparities among children. Recent evidence indicates that poor cardiovascular disease begins in the early decades of life. Many young individuals have developed atherosclerotic plaque by the time they reach adolescence (Berenson, Srinivasan, Bao, Newman, Tracy, & Wattigney, 1998; Berenson et al., 1992), and risk factors for C V D , including obesity, high blood pressure, and elevated lipids track across childhood and adolescence and into adulthood (Lauer, Burns, Clarke, & Mahoney, 1991; Mahoney et al., 1996). Importantly, these risk factors create a long-term burden on the cardiovascular system and are predictive of later clinical outcomes such as morbidity and mortality. Thus, it is important to identify environmental and psychosocial factors that contribute to the early progression of C V D (Matthews, 2005). In adults, the life-course approach has been used successfully to examine relationships between childhood and adulthood socioenvironmental circumstances and cardiovascular mortality (Davey Smith, Hart, Blane, & Hole, 1998; Frankel, Davey Smith, & Gunnel, 1999; Davey Smith & Hart, 2002). 4 The current article examines patterns of SES through childhood and their association with systolic blood pressure (SBP), diastolic blood pressure (DBP) , heart rate (HR), and body mass index (BMI) in adolescence. Both blood pressure and B M I in childhood and adolescence are important predictors of C V progression in adulthood; higher levels of these risk factors in early life are associated with premature onset of preclinical atherosclerosis and coronary artery calcification (Raitakari et al., 2003; Mahoney et al., 1996). Our main objective was to determine which life-course model (or combination of models) best explains the impact of SES throughout childhood on adolescent health. We utilized a new statistical approach to identify common trajectories of family SES across an adolescent's lifespan, and then used these trajectories to predict adolescent cardiovascular markers. We expected to identify distinct trajectories of family SES that would represent different combinations of the timing and dynamics of SES experiences. For instance, one trajectory might show low SES in the child's early life and upward mobility over time, whereas another might show high SES in the child's early life that would persist over time. To the extent that distinct trajectories of SES could be identified, we expected that they would be differentially related to adolescents' SBP, D B P , H R , and B M I . We expected that adolescents in families with persistently low and constantly fluctuating SES would have the worst C V outcomes. 5 M E T H O D Participants Public high school students in the St. Louis area were recruited via school flyers, announcements, and classroom presentations. Adolescents were eligible for the study i f they were (a) between the ages of 14 and 18, (b) fluent in the English language, (c) free of chronic medical conditions and mental health problems, and d) not using medication that affected the cardiovascular system. This study was approved by the Institutional Review Board at Washington University. The final sample consisted of 102 adolescents. Ages ranged from 14-18 years (mean 15.61). Fifty-three percent were females; 75% were Caucasian, 24% were African American, and 1% "other." Twenty-four percent of the adolescents' parents had a high school diploma, 11% had some college, and 63% had a college degree or higher. Students came from families where the mean family income category was 3.97 (category 3 corresponds to $50 000-74 999 and category 4 corresponds to $75 000-99 999). Students participated in the study with a parent. Eighty-three percent of parents were mothers. See Table 1 for a summary of descriptive information. Socioeconomic Status To capture SES trajectories over time, parents were asked to indicate the number of bedrooms in the family home during each year of the child's life. Number of bedrooms was used as a marker of SES for a number of reasons. First, retrospective recall of home ownership is more accurate than other asset variables such as family income (Cohen, Doyle, Turner, Alper, & Skoner, 2004). Second, housing is a more dynamic measure than other SES indicators, such as parental education. In our sample, most families (68.6%) had lived in two or more homes through the child's life, allowing us to capture changes in family SES. Based on the number of bedrooms in the family home during each year of the child's life, we calculated trajectories of SES, using procedures described in more detail below. Physiological Measures Blood pressure. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were monitored using a Dinamap Pro 100 automated blood pressure monitor (Critikon, Tampa, F L ) with a standard occluding cuff on the participant's right arm. S B P and D B P measures were taken 3 times during the last 5 minutes of a 10-minute baseline rest period. The averages of the three measures of S B P and D B P were used in statistical analyses. The coefficient alpha was .96 for SBP and .94 for D B P . Heart rate. Heart rate (HR) was measured through E K G monitoring. A n E K G signal was transduced using two active Meditrace SF450 disposable silver/silver chloride electrodes (Kendall-LTP, Chicopee, M A ) placed on each side o f the abdomen and a ground electrode beside the navel. The E K G signal was filtered and amplified by the Biopac M P 100 system (Biopac Systems, Santa Barbara, C A ) . H R was monitored continuously during the last 5 minutes of the 10-minute rest period. Body Mass Index. Height and weight were taken on a standard medical-grade balance beam scale and body mass (BMI) was computed from these two variables ( B M I = (Weight in Kilograms / ( Height in Meters ) x (Height in Meters )). For children and adolescents, the National Center for Health Statistics presents B M I by age and sex, using Z scores. These age-and sex-adjusted Z-scores were used in this study. 7 Potential Confounders. We measured a number of processes that could provide alternative explanations for relations between childhood SES and biological outcomes. We collected demographic information, including participant age and ethnicity. Because the majority of the sample (99%) was of Caucasian or African American descent, we created a dichotomous ethnicity variable coded as 1 for Caucasian and 2 for Other. Parents were asked whether there was a history of heart disease among first-degree relatives of the child. Answers were coded as 1 for "yes" and 0 for "no." We also asked parents to report on the number of people living in the family home through the child's life. Procedures Upon arriving at our laboratory, a research assistant described the study procedures in detail. Written assent was obtained from adolescents and consent from parents. Parents were interviewed about family SES. Adolescents were seated in an individual testing room, and the three E K G electrodes were applied. The blood pressure cuff was placed on the upper aspect of the participant's right arm with the microphone placed above an area where the brachial artery could be palpated. After a 15 min adaptation period while machines were tested, the 10 min baseline period was begun, and blood pressure and heart rate were monitored and averaged during the final five minutes. Adolescents then participated i i i a videotaped 'survival task' with parents, the results of which are described elsewhere (Chen & Berdan, 2006). Statistical Analyses In the first wave of analyses, we examined the distribution of study variables and screened for outliers. In each of the SBP and H R distributions, there was a score greater than 3 SD's from the sample mean. These scores were replaced with the next highest score in their respective distributions. In the second wave of analyses, we conducted bivariate analyses to 8 assess the relationship between study variables and potential confounds. In the third wave of analyses, we modeled trajectories of SES across childhood by using a S A S procedure called T R A J (Jones, Nagin, & Roeder, 2001) that separates individuals into trajectory groups. T R A J is a semiparametric, group-based modeling strategy that identifies clusters of individual trajectories. Model estimation produces posterior probabilities of membership in each trajectory group for each participant. These probabilities are then used to assign individuals to the trajectory group to which they have the highest probability of belonging. Differences in outcomes by group membership can then be modeled (Jones & Nagin, submitted). In the first part of the analysis, we determined the number of trajectories that best represented patterns of SES in our sample. The Bayesian Information Criterion (BIC) was used to determine the optimal number of trajectories, with higher values indicating a better fit. Next, we looked at the parameter estimates to determine the shape of each trajectory. We used these to determine whether each trajectory was linear, quadratic, or cubic. Finally, we examined the relationship between trajectory group membership and adolescents' biological outcomes. After completing the group trajectory analyses, it was quite clear how the timing of SES exposure contributed to adolescents' cardiovascular health outcomes. However, the trajectory analyses did not provide a very direct test of the dynamic models. Thus, the fourth wave of analyses was done as a follow-up to the T R A J analyses. Using specific measures of early life SES, current SES, cumulative SES, and social mobility, we were able to perform a head-to-head comparison of the life-course models, which T R A J did not permit us to do - it only allowed indirect inferences about which model provides the best explanation. We aimed to replicate the group trajectory findings and to further differentiate between the models of SES. First, we calculated the four specific indices of SES: (1) Early life SES was calculated by averaging the 9 number of bedrooms in the family home across the first 3 years of the child's life. (2) Current SES was the number of bedrooms in the family's current home. (3) Cumulative SES was the average number of bedrooms in the family home through the child's life. (4) Slope of SES (an indicator of social mobility) was the slope of the number of bedrooms in the family home through the child's life, calculated by regressing number of bedrooms upon years of life separately for each participant. See Table 2 for a summary and descriptive information. Second, we related early SES, current SES, cumulative SES, and slope of SES to biological outcomes using Pearson correlations. Finally, we compared the relative magnitude of each of these effects by conducting multiple regression analyses in which multiple SES measures were entered simultaneously predicting each biological outcome. 10 R E S U L T S Preliminary Analyses To identify potential confounders, correlations were computed between adolescents' demographic characteristics and study variables. Family history of heart disease was not significantly associated with SES indices or biological outcomes (p's>.10). Each number of bedrooms SES index was significantly related to the corresponding index of number of people in the family home; that is, families who lived in houses with more rooms tended to have more members (r's ranged from .35 to .64,/?<.001). Thus, we statistically controlled for the number of people in the house through childhood in all analyses. Furthermore, we statistically controlled for adolescents' gender, race, and age in all analyses of SBP , D B P , and heart rate, and adolescents' race in analyses of B M I , given that B M I z-scores from the National Center for Health Statistics were already gender and age adjusted. Trajectories of SES and Biological Outcomes Model selection: Identifying the number and shape of trajectories. To test the hypothesis that we would identify distinct trajectories of family SES, we first modeled the trajectory patterns of our sample. The BIC continued to increase as the number of trajectories increased, indicating a better model fit with the addition of each trajectory: The BIC was -1564.86 for three groups, -1406.64 for four groups, and -1310.65 for five groups. To prevent the trajectory group sizes from getting too small, we did not exceed the 5-group model. Both the 4 and 5-group models yielded similar information. For parsimony, we retained the four-trajectory model. The shape of each trajectory was determined by initially including linear, quadratic, and cubic parameters for each trajectory, and then dropping the nonsignificant ones (See Table 2). In 11 Model 1, linear, quadratic, and cubic parameters were included for each of the four trajectories. The cubic parameters were nonsignificant for each. Thus in Model 2, the cubic parameters were dropped from each trajectory. The quadratic parameter was nonsignificant for the first, second, and fourth trajectory. Thus in Model 3, the quadratic parameters were dropped from these trajectories. Because it had the highest B IC coefficient, we adopted Model 3 as our final model. In this model, the first, second, and fourth trajectories were linear, and the third trajectory was quadratic. As shown in Figure 1, the first trajectory group, which comprised 33% of the sample, showed the lowest SES in the child's early life. However, SES increased steadily through childhood and surpassed the next-lowest group by early adolescence. We called this group "low-increasing." The second trajectory group, accounting for 21% of the sample, showed moderate SES at the time of the child's birth and did not change over the course of childhood. We called this group "moderate-persistent." Similar to the moderate-persistent group, a third trajectory group, accounting for 36% of the sample, showed moderate SES at the child's birth; however, SES improved through childhood and leveled off around early adolescence. We referred to this group as "moderate-increasing." Finally, a fourth trajectory group, comprised of 10% of the sample, showed the highest SES at the time o f the child's birth and increased slightly with time. We called this group "high-increasing." It should be noted that trajectory group assignment is based on best fit; so, adolescents in each group do not follow exactly that trajectory. Relating the trajectories to biological outcomes. Next, we tested the hypothesis that SES trajectories would be differentially related to adolescent health outcomes. We first created residualized outcome variables by separating out the shared variance between our outcomes and covariates. For example, we included age, race, 12 gender, and the average number of people in the family home through childhood in a regression equation predicting S B P . We then used the residual scores from the regression analysis as our outcome variable. This variable represented SBP scores minus the shared variance between SBP and the covariates. These steps were repeated to create D B P and H R . Given that B M I Z-scores had already been adjusted for age and gender, B M I was created by regressing B M I on race and average number of people. Scores for each outcome variable were standardized, and these standardized, residualized scores were entered as outcome variables into the T R A J model. To test whether S B P differed by trajectory group, S B P was entered into the 4-group trajectory model specified above. This analysis yielded estimates of the average S B P and their corresponding standard errors for each trajectory group. The estimated average S B P was 114.24 mmHg (££=1.91) for the low-increasing group, 106.80 mmHg (££=1.57) for the moderate-persistent group, 108.34 mmHg (££=1.48) for the moderate-increasing group, and 106.42 mm Hg (££=2.88) for the high-increasing group (see Figure 2). Next, a Wald test was used to test the equality of the trajectory group S B P estimates. The Wald test is a ^-based test conducted on the variance estimates o f the estimated trajectory group averages (Jones & Nagin, submitted). We tested the contrast between the low-increasing and the moderate-persistent groups and found that the estimated average S B P was significantly higher for the low-increasing group, X\ =9.79, /?=.002. Next, we contrasted the low-increasing group with the moderate-increasing and high-increasing groups. Results indicated that the estimated average S B P was significantly higher for the low-increasing group compared to both the moderate-increasing group (jjfi 2=6.89,/»=.01) and the high-increasing group (j£i 2=7.14, p=.007). N o significant differences in the estimated average S B P emerged among the other groups: moderate-persistent versus moderate-increasing (ji 2 =.53, p>20), moderate-increasing versus high-increasing (%\2=.&9, p>.20), and moderate-13 persistent versus high-increasing (%i2=.22, p>.20). Thus, the low-increasing group had higher SBP compared to the moderate-persistent, moderate-increasing, and high-increasing groups, all of which had similar S B P . A comparison of the trajectories suggests that early-life SES contributes to group differences in SBP . Next, we assessed trajectory group differences in D B P by entering D B P into the 4-group trajectory model. The estimated average D B P was 63.43 rnmHg (££=1.34) for the low-increasing group, 59.39 rnmHg (££=1.11) for the moderate-persistent group, 60.17 rnmHg (££=1.02) for the moderate-increasing group, and 56.44 rnmHg (££=1.95) for the high-increasing group (see Figure 2). We tested the contrast between the estimated average D B P for the low-increasing and the moderate-persistent groups. The Wald test indicated that the estimated D B P was significantly higher for the low-increasing group (%\ =5.96, /?=.015). We then compared the low-increasing group with the moderate-increasing and high-increasing groups. Results indicated no significant difference between the low-increasing and moderate-increasing groups (j£i 2=3.05, p>.05). However, the estimated average D B P was significantly higher for the low-increasing group compared to the high-increasing group (x\ =9.45, p=.002). Next, we contrasted the moderate-persistent group with both the moderate-increasing and high-increasing groups and found no significant differences in the estimated average D B P {x\ =.90, p>.20 and %\2=1.64, p>.20, respectively). Finally, we tested the contrast between the moderate-increasing and high-increasing groups. Results indicated that the estimated average D B P was significantly lower for the high-increasing group (jft2=3.81, p=.05). Thus, the low-increasing group showed elevated D B P compared to the moderate-persistent and high-increasing groups. A comparison of the trajectories suggests that early-life SES contributes to these group differences. In addition, the high-increasing group had lower D B P compared to the moderate-increasing 14 group, but because the high-increasing group showed higher early-life, current, and cumulative SES, we cannot differentiate between the effects of early-life, current, and cumulative SES. Next, we tested trajectory group differences in H R and B M I . Results indicated no significant group differences in the estimated averages for these outcomes (p's>. 10). Life-Course Models of SES and Biological Outcomes Group trajectory analyses suggest that early life SES, rather than current SES, contributes to adolescents' SBP and D B P . However, the trajectory analyses did not provide a direct test of the dynamic SES models. Thus, analyses using specific measures o f early life SES, current SES, cumulative SES, and mobility were used to confirm the trajectory findings and to perform a head-to-head comparison of each of the life-course models. These follow-up analyses were intended to confirm our initial findings, and to clarify the role of cumulative SES and social mobility in adolescents' health outcomes. Correlations among SES indices. Early SES, current SES, and cumulative SES were significantly associated with one another (r's ranging from .51 to .84, /?'s<.001). Slope of SES was negatively related to early SES (r=-.44, p<.001), positively related to current SES (r=.49, p<.001), and unrelated to cumulative SES (p>. 10). Correlations between SES indices and biological outcomes. We first tested each type of SES measure separately for its ability to predict biological outcomes. We tested the critical period model by conducting Pearson correlations between early life SES and biological outcomes. Lower early life SES was associated with higher SBP and higher D B P (r=-.25,/K.01 and r=-.26,p<.05, respectively). Early life SES was unrelated to H R and B M I ( r = . 1 0 , / » . 2 0 and r=-.09,/?>.20, respectively). 15 The second timing model proposes that current SES is what is most important to current adolescent health. Thus, we tested the associations between current SES and biological outcomes. Lower current SES was associated with higher D B P (r=-.20, p<.05). Current SES was unrelated to SBP , H R , and B M I (r's ranging from -.07 to .00, p's>.10). To investigate relationships between dynamic SES and health, we first tested the model that cumulative SES would predict biological outcomes. Analyses indicated that lower cumulative SES was associated with higher S B P and D B P (r=-.20, p<.05 and r=-.26, p<.0\, respectively). Cumulative SES was unrelated to H R and B M I (r=.00, p>.20 and r=-.08, p>.\0, respectively). A second conceptualization of relationships between dynamic SES and health is that change in SES w i l l affect adolescent health. To test this hypothesis, we correlated slope of SES with adolescent health outcomes. Analyses indicated that higher slope of SES was associated with higher S B P and B M I (r=.23, p<.05 and r= 20, p<.05, respectively). Slope of SES was unrelated to D B P and H R (r=.05,p>.20 and r=-A3,p>.20, respectively). Simultaneous regression analyses. Simultaneous regression analyses were conducted to compare the effects of early life SES, current SES, cumulative SES, and slope of SES on S B P and D B P . B M I was not included in these analyses because it correlated with only one SES index (slope of SES). H R was not included in these analyses because there were no significant associations. When age, gender, and race were unrelated to the outcome variable, they were dropped from the regression model. Given that early SES, cumulative SES, and slope of SES were all significantly related to SBP, we tested whether one SES index contributed to the prediction of S B P over and above the others. Due to the number of covariates entered in each regression analysis, we did not have the 16 statistical power to include all three predictors in a single equation. Thus, we compared two SES indicators at a time. First, we compared the effects o f early life SES and cumulative SES. Results indicated that early life SES predicted SBP over and above the effects of cumulative SES (P=-.34, f(94)=-1.99, p<.05). In contrast, cumulative SES was not a significant predictor after controlling for early life SES (P=.10, f(94)=.62, p>.20). For early life SES and slope of SES, early life SES was marginally related to S B P after controlling for the effects of slope (P=-.21, /(94)=-1.9, p=.06). In contrast, slope o f SES did not contribute to the prediction o f S B P after controlling for early life SES (P=.12, ?(94)=1.20,/?>.20). Finally, we compared slope of SES and cumulative SES simultaneously. Slope of SES contributed significantly to the prediction of SBP above and beyond cumulative SES (p=.219, r(95)=2.44, p<05). Cumulative SES was marginally related to S B P after controlling for the effects of slope (p=-.19, f(95)=-1.91,_p=06). These analyses indicate that early life SES predicts S B P over and above the effects of both cumulative SES and slope of SES. Next, we compared the effects of early SES, current SES, and cumulative SES on D B P . First, we compared the effects of early SES and current SES. Results showed that early SES predicted over and above the effects of current SES (P=-.27, /(94)=-2.14, p<.05). In contrast, current SES did not predict over and above the effects of early SES (p=-.05, r(94)=-.40, p>.20). We then entered early SES and cumulative SES into a regression equation predicting D B P . Results indicated that neither predictor was significant after controlling for the effects of the other (P=-.23, t(93)=-l.2,p>.20 for early-life SES and p=-.09, f(93)=-.48,p>.20 for cumulative SES). Finally, we compared the effects of current SES and cumulative SES. Neither predictor contributed significantly to D B P after controlling for the effects of the other (p=.074, f(95)=.45, p>.20 for current SES and p=-.32, r(95)=-1.85, p>.05 for cumulative SES). These analyses 17 indicate that early life SES predicts DBP independent of current SES. However, both early SES and cumulative SES may play a similarly important role in adolescents' DBP. 18 D I S C U S S I O N The current study examined patterns of SES throughout childhood and their relationship to adolescents' SBP , D B P , H R , and B M I . The goal was to determine which life-course models (early life, current, cumulative, and/or mobility) best explained the relationship between SES experiences across an adolescent's lifespan and markers of current cardiovascular health. Our first objective was to determine whether distinct trajectories of SES would be evident in our sample of middle-class families. Analyses revealed four distinct trajectories of family SES: low-increasing, moderate-persistent, moderate-increasing, and high-increasing. Adolescents in the low-increasing group experienced low early life SES and upward mobility through childhood. Adolescents in the moderate-persistent group experienced moderate early life SES that remained stable over time. Compared to adolescents in other groups, these adolescents experienced the least social mobility and the lowest current SES . Adolescents in the moderate-increasing group experienced moderate early life SES and upward mobility that leveled off in early adolescence. Finally, adolescents in the high-increasing group experienced high early life, cumulative, and current SES. Our next goal was to determine whether these trajectories differentially related to markers of C V risk. In respect to SBP , adolescents in the low-increasing group had the highest SBP of any of the trajectory groups. The difference between the low-increasing group and each of the other groups was quite large (around 7 mm Hg). In adult populations, S B P differences of this magnitude are clinically relevant. For instance, it has been estimated that a 7 mm H g drop in middle-aged men's S B P would be associated with a 16% reduction in first occurrences of major C V D over the following 10 years (Emberson, Whincup, Morris , Walker, Ebrahim, 2004). Which life-course model best explains this effect? The low-increasing trajectory showed the 19 lowest SES early in life, but SES increased over time so that it was comparable to the moderate-persistent and moderate-increasing groups by adolescence. This indicates that current SES does not differentiate adolescents' current SBP . In addition, the low-increasing trajectory showed similar cumulative SES to the moderate-persistent group, indicating that group differences in S B P could not be predicted by cumulative SES. However, trajectory group comparisons failed to differentiate the effects of " low" versus "increasing" SES. Specifically, the low-increasing group had higher S B P compared to the moderate-increasing and high-increasing groups, but because social mobility differed across groups, it was unclear whether group differences were due to early SES or mobility. However, follow-up analyses indicated that early life SES predicted above and beyond slope of SES, and this strongly suggests that timing, rather than dynamics, is the critical factor. Collectively, these analyses indicate that of the life-course SES models we studied, the critical period account best depicts S B P in adolescence. SES early in life also appears to play a role in adolescents' D B P . Adolescents in the low-increasing group had higher D B P than the moderate-persistent or high-increasing groups. Given that the low-increasing and moderate-persistent trajectories showed similar current SES, current SES does not best discriminate adolescents' D B P . In addition, the similar cumulative SES shown by the low-increasing and moderate-persistent trajectories indicates that cumulative SES does not fully explain group differences in D B P . Furthermore, the regression analyses showed that early life SES predicted D B P over and above current SES. Thus, there is evidence for the importance of early life SES in predicting current adolescent D B P . However, there is also evidence that cumulative SES contributes to adolescents' D B P . Specifically, the high-increasing group had the lowest D B P and was significantly different from both the moderate-increasing and low-increasing groups. The high-increasing group had higher early life, current, and cumulative • 20 SES; so, it is difficult to discern which aspect(s) of S E S account for the low D B P in this group. However, multiple regression analyses indicated that while early life SES predicted over and above current SES, it did not independently predict beyond cumulative SES. These findings suggest that both early life and cumulative SES may play similarly important roles in adolescent D B P . Finally, trajectory analyses did not allow us to tease apart the effects of " low" versus "increasing" SES. Specifically, the low-increasing group had higher D B P compared to the moderate-persistent group, but it was unclear whether this was due to early life SES (low versus moderate) or mobility (increasing versus persistent). However, follow-up analyses indicated that slope of SES, was unrelated to D B P ; so, we were able to rule out mobility as a factor in adolescents' D B P . Taken together, these results point to early life SES exposure as an important determinant of adolescents' D B P , but there is also some evidence for cumulative SES influences. In fact, it may be that timing matters at the low end of SES (accounting for the high D B P in the low-increasing group), but that cumulative SES matters at the high end of the spectrum (accounting for the low D B P in the high-increasing group). These findings provide insights into how SES in early life" comes to influence C V morbidity and mortality. They suggest that by the time children reach adolescence, the detrimental influence of low early-SES on C V health are already apparent. It may be that low SES increases exposure to social and physical "pollutants" - such as crowding, violence, malnutrition, allergens - that have the capacity to program patterns of biological functioning over the long-term (Meaney, 2004; Mi l l e r & Chen, under review). In rodents this biological embedding of social experience results in differential adulthood regulation o f stress-response systems such as the hypothalamic-pituitary-adrenocortical and sympathetic adrenal medullary axes (Hertzman, 1999; Meaney, Aitken, Bhatnagar, & Sapolsky, 1991; Meaney, Aitken, Van 21 Berkel, Bhatnagar, & Sapolsky, 1998). Heightened activity of these systems may increase vulnerability to cardiac disease over the lifespan. Early life experiences may also influence health behaviors through the life course. For instance, early life SES is associated with cardiovascular risk factors such as cigarette smoking, binge drinking, and obesity in adulthood (Lynch, Kaplan, & Salonen, 1997; Lawlor et al, 2005). Thus, adolescents who experienced low early life SES may be more likely to engage in poor health behaviors that have an adverse effect on the cardiovascular system. In contrast, current SES was unrelated to adolescents' blood pressure after controlling for the effects of early SES experiences. This finding is consistent with past research showing that SES is unrelated to blood pressure among adolescents. In general, studies in this area have shown that SES effects on blood pressure are evident in childhood and adulthood, but not in adolescence (Chen, Matthews, & Boyce, 2002). To explain this pattern of relationships, West (1997) suggested that SES differences in children's health diminish in adolescence due to a common youth culture in school. Youth culture is thought to cross family and neighborhood socioeconomic lines in a way that reduces SES health differences during this stage of life. Although more research is needed to test these hypotheses directly, it appears that current SES models may be more useful for explaining the relationship between SES and B P among populations of children and adults, rather than among adolescents. In respect to dynamic aspects of SES, our findings indicate that cumulative SES may play a role in adolescents' D B P . Higher D B P among adolescents who experienced longer durations of low SES may reflect cumulative wear-and-tear on the body (i.e., allostatic load, McEwen, 1998). This is consistent with past research showing that the accumulation of environmental and psychosocial risk associated with low SES is related to children's blood pressure, adrenocortical 22 activity, and B M I (Evans, 2003). In the present study, cumulative SES was associated with D B P , but not S B P . We speculate that this is related to the way low-SES children appraise their social environments. Stressors can be appraised as threatening or challenging, and research evidence suggests that these two responses to stressors have distinct physiological correlates (Tomaka, Blascovich, Kibler , & Ernst, 1997). Threat appraisals are characterized by increases in vascular reactivity, linked to D B P , whereas challenge appraisals are characterized by increases in cardiac reactivity, linked to SBP. Previous research has demonstrated that low-SES children are more likely to interpret ambiguous social situations as threatening (Chen, Langer, Raphaelson, & Matthews, 2004). Thus, higher D B P among adolescents who experienced lower cumulative SES may reflect the accumulation of threat appraisals over time. We found no support for mobility as an important determinant o f B P . In the group trajectory analyses for S B P , the moderate-increasing and high-increasing groups differed from the low-increasing group, but did not differ from the moderate-persistent group. If mobility was an important predictor of SBP , we would expect to see differences between "persistent" and "increasing" trajectory groups. Regression analyses confirmed these findings, as mobility was unrelated to S B P after controlling for the effects of early life SES. We reached similar conclusions for D B P . Specifically, the moderate-increasing and high-increasing groups differed, whereas the moderate-increasing and moderate-persistent groups did not, suggesting that "persistent" versus "increasing" trajectories do not differentiate adolescents' D B P . Furthermore, correlational analyses showed no association between mobility and D B P . Collectively, these findings indicate that social mobility through childhood does not play a role in adolescents' B P . However, our sample showed very little social mobility; so, this dimension might be more important in populations with varying degrees of upward and downward mobility. In addition, it 23 may be important to use measures of SES that are more sensitive to SES fluctuations (e.g., family income), as well as dynamic measures of health. For instance, changes in B P may occur closely in time with fluctuations in SES; so, without a repeated-measures design, increases and decreases in B P would get washed out over time. Counter to our expectations, we found no association between trajectories of SES and adolescents' H R and B M I . These findings are inconsistent with past studies showing that lower SES is associated with higher B M I among adolescents (e.g., Evans, 2003; Goodman, 1999). Recent evidence suggests that family and neighborhood resources have independent effects on adolescents' B M I (Chen & Patterson, in press). Thus, it may be important to take neighborhood characteristics into account when examining the relationship between family housing and B M I . Given that H R and B P are both controlled by the autonomic nervous system, we would expect to find a similar pattern of results for these outcomes. Thus, it is unclear why early-life SES would shape D B P and SBP , but not H R . There are a number of limitations to this study that should be noted. First, assignment into trajectory groups is probabilistic in nature. In each trajectory group, some adolescents had a high probability of membership in that group, whereas others had a lower probability of membership in that group. However, the 95% confidence intervals calculated around the trajectories show little overlap, indicating minimal classification error (See Figure 1). Also , errors in classification would only increase random error, making it more difficult to detect relationships between trajectories and outcomes like S B P and D B P . Second, the study would have benefited from a more thorough examination of C V risk, including measures of cigarette smoking, physical activity, cholesterol, and triglycerides. It is possible that the impact of SES timing and dynamics depends on the C V risk factor in question, and this is an important area for 24 future study. Third, our outcome measures were limited by one-time lab assessment. Resting B P in the lab can be influenced by numerous factors, such as the white-coat-effect. These problems are mitigated to a large extent by ambulatory B P measures which are better predictors of clinical outcomes. For these reasons, ambulatory monitors should be used in the next wave of studies. Fourth, there may be problems associated with the measurement of SES via number of bedrooms. First, parents' retrospective reports on housing may be unreliable. Second, this measure does not take into account whether families rent or own, which neighborhood they live in, or the costs o f renting versus owning in their neighborhood. However, this imperfect indicator of SES should only diminish power to detect relations with B P . Future studies that collect a broad array of SES data prospectively would help resolve these difficulties, especially i f they could be done from the time the child was born. Finally, our measure of social mobility may also reflect the effects of moving on children. Changing homes, neighborhoods, and schools can be stressful for children, and this may account for the association between upward mobility and higher B M I and SBP in this study. In summary, findings from the current study indicate that the timing of SES exposure plays an important role in adolescents' blood pressure. Specifically, SES experiences early in life were associated with adolescents' S B P and D B P , independent of current SES. These findings point toward early life developmental processes as potential candidates for explaining the relationship between SES and risk factors related to C V D . In addition, our results suggest that interventions designed to reduce SES health disparities would be most effective i f they took place within the first few years of a child's life. 25 R E F E R E N C E S Adler, N . E . , Boyce, T., Chesney, M . A . , Cohen, S., Folkman, S., Kahn, R. L . , & Syme, S. L . (1994). Socioeconomic status and health: The challenge of the gradient. American Psychologist, 49, 15-24. Anderson, N . B . , & Armstead, C. A . (1995). Toward understanding the association of socioeconomic status and health: A new challenge for the biopsychosocial approach. Psychosomatic Medicine, 57, 213-225. Barker, D . J. P. (1992). Fetal and infant origins of adult disease. London: British Medical Journal. Belsley, D. , Kuh , E . , Welsch, R. (1980). Regression Diagnostics. N e w York: Wiley. Berenson, G . S., Srinivasan, S. R., Bao, W. , Newman, W . P., Tracy, R. E . , & Wattigney, W. A . (1998). Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. New England Journal of Medicine, 338, 1650-1656. Berenson, G.S. , Wattigney, W . A . , Tracy, R .E . , Newman, W.P. , Srnivasan, S.R., Webber, L .S . , et al. (1992). Atherosclerosis of the aorta and coronary arteries and cardiovascular risk factors in persons aged 6 to 30 years and studied at necropsy (the Bogalusa Heart Study). American Journal of Cardiology, 70, 851-858. Chen, E . & Berdan, L . (2006). Socioeconomic Status and Patterns of Parent-Adolescent Interactions. Journal of Research on Adolescence, 16, 19-27. Chen E. , & Patterson, L . Q. (in press). Neighborhood, family, and subjective socioeconomic status: How do they relate to adolescent health? Health Psychology. Chen, E . , Langer, D . , Raphaelson, Y . , & Matthews, K . (2004). Socioeconomic status and health in adolescents: The role of stress interpretations. Child Development, 75, 1039-1052. 26 Chen, E. , Matthews, K . A . , & Boyce, W . T. (2002). Socioeconomic differences in children's health: H o w and why do these relationships change with age? Psychological Bulletin, 128, 295-329. Cohen, S., Doyle, W . J., Turner, R. B . , Alper, C. M . , & Skoner, D . P. (2004). Childhood socioeconomic status and host resistance to infectious illness in adulthood. Psychosomatic Medicine, 66, 553-558. Davey Smith, G . , & Hart, C. (2002). Life-course socioeconomic and behavioral influences on cardiovascular disease mortality: The collaborative study. American Journal of Public Health, 92, 1295-1298. Duncan, G . (1988). The volatility of family income over the life course. In P. Baltes, D . Featherman, & R. M . Lerner (Eds.) Life-Span Development and Behavior, Vol. 9 (pp. 317-358). Hillsdale, N J : Lawrence Erlbaum Associates. Duncan, G . , Brooks-Gunn, J., & Klebanov, P. (1994). Economic deprivation and early childhood development. Child Development, 65, 296-318. Duncan, G . , Yeung, W . J., Brooks-Gunn, J., & Smith, J. R. (1998). H o w much does childhood poverty affect the life chances of children? American Sociological Review, 63, 406-423. Emberson, J., Whincup, P., Morris , R., Walker, M . , & Ebrahim, S. (2004). Evaluating the impact of population and high-risk strategies for the primary prevention of cardiovascular disease. European Heart Journal, 25, 484-491. Evans, G. W. (2003). A multimethodological analysis of cumulative risk and allostatic load among rural children. Developmental Psychology, 39, 924-933. Frankel, S., Davey Smith, G . , & Gunnel, D. (1999). Childhood socioeconomic position and adult cardiovascular mortality: the Boyd Orr cohort. American Journal of Epidemiology, 150, 1081-1084. Galobardes, B . , Davey Smith, G . , & Lynch, J. W. (2006). Systematic review of the influence of childhood socioeconomic circumstances on risk for cardiovascular disease in adulthood. Annals of Epidemiology, 16, 91-104. Galobardes, B . , Lynch, J. W. , & Davey Smith, G . (2004). Childhood socioeconomic circumstances and cause-specific mortality in adulthood: Systematic review and interpretation. Epidemiological Reviews, 26, 7-21. Goodman, E . (1999). The role of socioeconomic status gradients in explaining differences in U S adolescents'health. American Journal of Public Health, 89, 1522-1528. Hertzman, A . (1999). The biological embedding of early experience and its effects on health in adulthood. Annals of the New York Academy of Sciences, 896, 85-95. Jones, B . , & Nagin, D . Advances in group-based trajectory modeling and a S A S procedure for estimating them. Submitted. Jones, B . , Nagin, D . , & Roeder, K . (2001). A S A S procedure based on mixture models for estimating developmental trajectories. Sociological Methods and Research, 29, 374-393. Karlamangla, A . S., Singer, B . H . , Will iams, D. R., Schwartz, J. E . , Matthews, K . A . , Kiefe, C. I., & Seeman, T. E . (2005). Impact of socioeconomic status on longitudinal accumulation of cardiovascular risk in young adults: the C A R D I A Study (USA) . Social Science and Medicine, 60, 999-1015. 28 Laaksonen, M . , Rahkonen, O., Martikainen, P., & Lahelma, E . (2005) Socioeconomic position and self-rated health: The contribution of childhood socioeconomic circumstances, adult socioeconomic status, and material resources. American Journal of Public Health, 95, 1403-1409. Lauer, R. M , Burns, T. L . , Clarke, W . R., & Mahoney, L . T. (1991). Childhood predictors of future blood pressure. Hypertension, 18, 174-181. Lynch, J. W. , Kaplan, G . A . , & Salonen, J. T. (1997). Why do poor people behave poorly? Variation in adult health behaviours and psychosocial characteristics by stages of the socioeconomic lifecourse. Social Science and Medicine, 44, 809-819. Mahoney, L . T., Bruns, T. L . , Stanford, W. , Thompson, B . H . , Witt, J. D . , Rost, C. A . , & Lauer, R. M . (1996). Coronary risk factors measured in childhood and young adult life are associated with coronary artery calcification in adulthood: The Bogalusa Heart Study. Journal of the American Medical Association, 290, 2271-2276. Matthews, K . A . (2005). Psychological perspectives on the development of coronary heart disease. American Psychologist, 60, 783-79. Matthews, K . A . , Kiefe, C . I., Lewis, C . E . , L i u , K . , Sidney, S., & Yunis , C. (2002). Socioeconomic trajectories and incident hypertension in a biracial cohort o f young adults. Hypertension, 39, 772-776. McDonough, P., Duncan, G . J. , Wil l iams, D . , & House, J. (1997). Income dynamics and adult mortality in the United States, 1872-1989. American Journal of Public Health, 87, 1476-1483. McEwen, B . S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 338, 171-179. Mcleod, J. D . , & Shanahan, M . J. (1996). Trajectories of poverty and children's mental health. Journal of Health and Social Behavior, 37, 209-220. Meaney, M . J. (2004). The nature of nurture: Maternal effects and chromatin remodeling. In: J. T. Cacioppo, & G . G . Berntson (Eds.) Essays in social neuroscience. (pp. 1-14). Cambridge, M A , U S : M I T Press. Meaney, M . J., Aitken, D . H . , Bhatnagar, S., & Sapolsky, R. M . (1991). Postnatal handling attenuates certain neuroendocrine, anatomical, and cognitive dysfunctions associated with aging in female rats. Neurobiology of Aging, 12, 31-38. Meaney, M . J., Aitken, D . H . , Van Berkel, C. Bhatnagar, S., & Sapolsky, R. M . Effect of neonatal handling on age-related impairments associated with the hippocampus. Science, 239, 766-768. Mil ler , G. E . , & Chen, E . (under review). Early-Life Socioeconomic Conditions Predict Expression of Genes Regulating Inflammation During Adolescence. Pensola, T., & Martikainen, P. (2003). Life-courses experiences and mortality by adult social class among young men. Social Science and Medicine, 58, 2149-2170. Power, C , Manor, O., & Matthews, S. (1999). The duration and timing of exposure: Effects of socioeconomic environment on adult health. American Journal of Public Health, 89, 1059-1065. Raitakari, O. T., Juonala, M . , Kahonen, M . , Taittonen, L . , Laitinen, T., Maki-Torkko, N . , et al. (2003). Cardiovascular risk factors in childhood and carotid artery intima-media thickness in adulthood. Journal of the American Medical Association, 290, 2277-2283. Tomaka, J., Blascovich, J., Kibler, J., & Ernst, J. M . (1997). Cognitive and physiological antecedents of threat and challenge appraisal. Journal of Personality and Social Psychology, 73, 63-72. Virtanen, P., Vahtera, J., K iv imak i , M . , Liukkonen, V . , Virtanen, M . , & Ferrie, J. (2005). Labor Market Trajectories and Health: A Four-Year Follow-up Study o f Initially Fixed-Term Employees. American Journal of Epidemiology, 161, 8 4 0 - 8 4 6 . West, P. (1997). Health inequalities in the early years: Is there equalization in youth? Social Science and Medicine, 30, 665-673. Table 1 Demographic and Health Characteristics of the Sample Family Current Income (1-6 scale) 3.97±1.5 Parent Years of Education 15.93±2.45 History of Heart Disease 4% Adolescents Age 15.61±1.05 Female 53% Caucasian 75% African American 24% SBP (mmHg) 108.93±9.17 D B P (mmHg) 60.23±6.24 HR(bpm) 73.11±10.80 B M I 24.17±4.89 Note. For family income, category 3 corresponds to $50 000-74 999 and category 4 corresponds to $75 000-99 999. Table 2 Indices of Socioeconomic Status through Child's Life SES Index Computation Mean±SD Early Life Average # bedrooms in family home 3.07±0.85 from child's birth to age 3 Current Current # bedrooms in family home 3.78±0.84 Cumulative Average # bedrooms in family home 3.46±0.72 from child's birth to present Slope Slope of # bedrooms in family home 0.053±0.064 from child's birth to present Table 3 Selecting the number of trajectories. Number of Groups B I C Change in B I C 1 -2161.97 2 -1616.40 545.57 3 -1564.86 51.54 4 -1406.64 158.22 5 -1310.65 95.99 Note. BIC=Bayesian Information Criterion Table 4 Model Selection for the Shape of each Trajectory Trajectory 1 Trajectory 2 Trajectory 3 Trajectory 4 BIC Model and Parameter Function Tested Parameter Estimate Function Parameter Tested Estimate Function Parameter Tested Estimate Function Tested Parameter Estimate 1 Intercept Cubic 1.84 Cubic 2.85 Cubic 3.25 Cubic 4.31 -1406.64 Linear 0.11 0.039 0.13** 0.082 Quadratic 0.00 -.002 -.002 -.002 Cubic 0.00 0.00 0.00 0.00 2 Quadratic Quadratic Quadratic Quadratic -1392.00 Intercept 1.83 2.85 3.23 4.32 Linear 0.036 0 15*** 0.079** Quadratic 0.00 -.001 -.006*** -.001 3 Linear Linear Quadratic Linear -1382.06 Intercept 1.86 2.92 3.23 4.37 Linear 0.014** 0.15*** -.058*** Quadratic -.006*** Note. BIC=Bayesian Information Criterion. Trajectory l=Low-increasing, Trajectory 2=Moderate-persistent, Trajectory 3=Moderate-increasing, and Trajectory 4=High-increasing. **p<.0\, ***p<.0 Table 5 Partial correlations between SES indices and adolescent biological outcomes. S B P D B P H R B M I Early life -.25** -.26* .10 -.09 Current -.07 -.20* -.05 .00 Cumulative -.20* -.26** .00 -.08 Slope 0.23* .05 -.13 .20* Note. Correlations control for age, race, gender, and the number of people in the family home. *jt?<.05. **p<.0l 36 6 i 0 "I 1 1 1 1 1 1 1 1 — — i 0 2 4 6 8 10 12 14 16 18 Age of Child (Years) Figure 1. Estimates of SES trajectories and their accompanying 95% confidence interval from the child's birth to current age. 37 • Low-increasing • Moderate-persistent O Moderate-increasing i l High-increasing Figure 2. Estimates of average SBP by trajectory group. The error bars represent standard error of the mean. The estimated average S B P for the low-increasing group is significantly higher than the estimated averages for the moderate-persistent, moderate increasing, and high-increasing groups. 38 Figure 3. Estimates of average D B P by trajectory group. The error bars represent standard error of the mean. The estimated average D B P for the low-increasing group is significantly higher than the estimated averages for the moderate-persistent and high-increasing groups. The estimated average D B P is significantly lower for the high-increasing group compared to the moderate- increasing group. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0092719/manifest

Comment

Related Items