Body weight and obesity
Accumulation of childhood poverty on young adult overweight or obese status: race/ethnicity and gender disparities Daphne C Hernandez,1 Emily Pressler2 1
Department of Health and Human Performance, Texas Obesity Research Center, University of Houston, Houston, Texas, USA 2 Department of Human Development Family Studies, The Pennsylvania State University, State College, Pennsylvania, USA Correspondence to Professor Daphne C Hernandez, Department of Health and Human Performance, University of Houston, 3855 Holman Street, 104 Garrison, Houston, Texas 77204-6015, USA; [email protected] Received 3 July 2013 Revised 3 November 2013 Accepted 9 December 2013 Published Online First 3 January 2014
ABSTRACT Background Childhood poverty is positively correlated with overweight status during childhood, adolescence and adulthood. Repeated exposure of childhood poverty could contribute to race/ethnicity and gender disparities in young adult overweight/obese (OV/OB) weight status. Methods Young adults born between 1980 and 1990 who participated in the Young Adult file of the 1979 National Longitudinal Study of Youth were examined (N=3901). The accumulation of childhood poverty is captured via poverty exposure from each survey year from the prenatal year through age 18 years. Body mass index was calculated and categorised into the reference criteria for adults outlined by the Center for Disease Control. Logistic regression models were stratified by race/ethnicity and included a term interacting poverty and gender, along with a number of covariates, including various longitudinal socioeconomic status measures and indicators for the intergenerational transmission of economic disadvantage and body weight. Results Reoccurring exposure to childhood poverty was positively related to OV/OB for white, black and Hispanic young adult women and inversely related for white young adult men. A direct relationship between the accumulation of childhood poverty and OV/OB was not found for black and Hispanic young adult men. Conclusions Helping families move out of poverty may improve the long-term health status of white, black and Hispanic female children as young adults. Community area interventions designed to change impoverished community environments and assist low-income families reduce family level correlates of poverty may help to reduce the weight disparities observed in young adulthood.
INTRODUCTION
To cite: Hernandez DC, Pressler E. J Epidemiol Community Health 2014;68:478–484. 478
In the USA, experiencing poverty or having a low socioeconomic status (SES) during childhood is positively correlated with overweight status during childhood, adolescence1 2 and adulthood.3–7 However, gender and race/ethnic disparities exist in the poverty-overweight status relationship. Specifically, poverty and multidimensional measures of SES have a positive association with obesity for adolescent women, but not men, during the transition to adulthood.5 8 Further, adolescent women exposed to persistent economic adversity are at a greater risk for obesity in adulthood.3 Although the prevalence of overweight status for Black adolescent women remains high even after household SES increases,9 several studies find an inverse association between SES and overweight9 and obesity1 among non-Hispanic white adolescent women. Further, an
inverse association between the poverty-income ratio and obesity has also been observed among adult white women 10; yet, an inverse association between poverty-income ratio has been inconsistent among adult white men.10 The next step is to explore the race/ethnicity and gender differences that evolve in young adult overweight/obese (OV/OB) status from repeated exposures of childhood poverty that occurred between the prenatal period to age 18 years. The life course perspective11 provides a framework for understanding how experiences with childhood poverty influence OV/OB in young adulthood and contribute to race/ethnic and gender disparities. Within the life course perspective, the accumulation model can be applied to explain how adverse events, such as exposure to poverty, relate to developing obesity.12 The accumulation of adversity recognises that the repeated or continued exposure to adversity over the life course relates to an individual’s risk for disease.12 The risk literature supports this model as research has found that children who reside in poverty for long periods face greater risk to their development and health compared with children who reside in poverty for shorter periods or do not experience poverty at all.13 14 Thus, duration of poverty is related to various negative child health outcomes, but more research is needed regarding the association between repeated exposures of childhood poverty and race/ethnic and gender disparities in young adult OV/OB status. The current study applies a life course perspective11 to explore the accumulating effects of childhood poverty from the prenatal period to age 18 years, on OV/OB status in young adulthood (mean=21 years) paying close attention to race/ethnicity and gender differences. To achieve this overarching goal we use data from the National Longitudinal Survey of Youth 1979 (NLSY79) and the Young Adults files (YA-NLSY). We also include measures of poverty closer to the body mass index (BMI) assessment as controls to demonstrate that beyond later poverty exposure the accumulation of childhood poverty can uniquely impact young adults’ OV/OB status. Aside from these various measures of poverty, we include longitudinal measures of childhood SES. Health researchers have suggested that multiple SES indicators, rather than a composite SES measure, should be used in analytical models as education and income are associated with health differently.15 Thus, we include multiple longitudinal indicators of childhood SES (eg, per cent of childhood exposed to maternal unemployment), as well as indicators correlated with poverty (eg, parental relationship instability) as covariates.
Hernandez DC, et al. J Epidemiol Community Health 2014;68:478–484. doi:10.1136/jech-2013-203062
Body weight and obesity Finally, research has demonstrated that there is an intergenerational transmission of economic disadvantage and mobility16 and body weight within families.17 18 The longitudinal, two-generational nature of the data provides a unique opportunity to include variables correlated with economic disadvantage and mobility, such as young adults’ maternal grandmother’s education and mother’s childhood family structure. The data also provides an opportunity to control for the intergenerational transmission of weight status within families by including maternal BMI around the same age as her child’s weight status in young adulthood. Thus, the analyses include sophisticated measures that are rarely available in other datasets, resulting in more precise measures that help further our understanding of race, ethnic and gender disparities in young adult OV/OB status.
METHODS AND PROCEDURES Data from the NLSY79 and the linked YA-NLSY file was used. NLSY data are sponsored by the US Department of Labor and have been compiled through the Center for Human Resource Research at the Ohio State University. The original NLSY79 cohort includes a nationally representative sample of 12 686 men and women followed from 1979–2010. The NLSY has surveyed the biological children of these women every 2 years since 1986 and represents over 90% of all the children born to this cohort.
Sample A total of 11 494 children were born to 4928 NLSY79 mothers between 1970 and 2010. The child sample was restricted to young adults who were at least 20 years old by 2010 (3959 children were excluded because they were born prior to the start of data collection in 1979 or were too young). Of these young adults (n=7535), the sample was further restricted to individuals who had valid BMI data at age 20 years (2367 cases excluded) and complete data needed to calculate poverty or measures correlated with poverty (1168 cases excluded). Young adult women who were pregnant during their BMI assessment were excluded from the analyses (99 cases excluded). The final analytical sample consisted of 3901 young adults who were born between 1980 and 1990 (mean=1985). The Institutional Review Board of the Pennsylvania State University approved the study. Compared with the excluded sample, young adults in the analytical sample had a lower BMI and spent more of their childhood in poverty but not as a young adult. The analytical sample was comprised of young adults who were more likely born at a healthy weight to older, married mothers. The young adults were younger, white, male, single, resided with more siblings, displayed less depressive symptoms but tried more substances as an adolescent, and were exposed to less maternal unemployment but greater family instability.
Measures Body mass index Young adults self-reported their height (feet and inches) and weight (pounds) in the YA-NLSY. BMI was calculated and categorised into the reference criteria for adults outlined by the Center for Disease Control (ie underweight (BMI
Accumulation of childhood poverty on young adult overweight or obese status: race/ethnicity and gender disparities Daphne C Hernandez,1 Emily Pressler2 1
Department of Health and Human Performance, Texas Obesity Research Center, University of Houston, Houston, Texas, USA 2 Department of Human Development Family Studies, The Pennsylvania State University, State College, Pennsylvania, USA Correspondence to Professor Daphne C Hernandez, Department of Health and Human Performance, University of Houston, 3855 Holman Street, 104 Garrison, Houston, Texas 77204-6015, USA; [email protected] Received 3 July 2013 Revised 3 November 2013 Accepted 9 December 2013 Published Online First 3 January 2014
ABSTRACT Background Childhood poverty is positively correlated with overweight status during childhood, adolescence and adulthood. Repeated exposure of childhood poverty could contribute to race/ethnicity and gender disparities in young adult overweight/obese (OV/OB) weight status. Methods Young adults born between 1980 and 1990 who participated in the Young Adult file of the 1979 National Longitudinal Study of Youth were examined (N=3901). The accumulation of childhood poverty is captured via poverty exposure from each survey year from the prenatal year through age 18 years. Body mass index was calculated and categorised into the reference criteria for adults outlined by the Center for Disease Control. Logistic regression models were stratified by race/ethnicity and included a term interacting poverty and gender, along with a number of covariates, including various longitudinal socioeconomic status measures and indicators for the intergenerational transmission of economic disadvantage and body weight. Results Reoccurring exposure to childhood poverty was positively related to OV/OB for white, black and Hispanic young adult women and inversely related for white young adult men. A direct relationship between the accumulation of childhood poverty and OV/OB was not found for black and Hispanic young adult men. Conclusions Helping families move out of poverty may improve the long-term health status of white, black and Hispanic female children as young adults. Community area interventions designed to change impoverished community environments and assist low-income families reduce family level correlates of poverty may help to reduce the weight disparities observed in young adulthood.
INTRODUCTION
To cite: Hernandez DC, Pressler E. J Epidemiol Community Health 2014;68:478–484. 478
In the USA, experiencing poverty or having a low socioeconomic status (SES) during childhood is positively correlated with overweight status during childhood, adolescence1 2 and adulthood.3–7 However, gender and race/ethnic disparities exist in the poverty-overweight status relationship. Specifically, poverty and multidimensional measures of SES have a positive association with obesity for adolescent women, but not men, during the transition to adulthood.5 8 Further, adolescent women exposed to persistent economic adversity are at a greater risk for obesity in adulthood.3 Although the prevalence of overweight status for Black adolescent women remains high even after household SES increases,9 several studies find an inverse association between SES and overweight9 and obesity1 among non-Hispanic white adolescent women. Further, an
inverse association between the poverty-income ratio and obesity has also been observed among adult white women 10; yet, an inverse association between poverty-income ratio has been inconsistent among adult white men.10 The next step is to explore the race/ethnicity and gender differences that evolve in young adult overweight/obese (OV/OB) status from repeated exposures of childhood poverty that occurred between the prenatal period to age 18 years. The life course perspective11 provides a framework for understanding how experiences with childhood poverty influence OV/OB in young adulthood and contribute to race/ethnic and gender disparities. Within the life course perspective, the accumulation model can be applied to explain how adverse events, such as exposure to poverty, relate to developing obesity.12 The accumulation of adversity recognises that the repeated or continued exposure to adversity over the life course relates to an individual’s risk for disease.12 The risk literature supports this model as research has found that children who reside in poverty for long periods face greater risk to their development and health compared with children who reside in poverty for shorter periods or do not experience poverty at all.13 14 Thus, duration of poverty is related to various negative child health outcomes, but more research is needed regarding the association between repeated exposures of childhood poverty and race/ethnic and gender disparities in young adult OV/OB status. The current study applies a life course perspective11 to explore the accumulating effects of childhood poverty from the prenatal period to age 18 years, on OV/OB status in young adulthood (mean=21 years) paying close attention to race/ethnicity and gender differences. To achieve this overarching goal we use data from the National Longitudinal Survey of Youth 1979 (NLSY79) and the Young Adults files (YA-NLSY). We also include measures of poverty closer to the body mass index (BMI) assessment as controls to demonstrate that beyond later poverty exposure the accumulation of childhood poverty can uniquely impact young adults’ OV/OB status. Aside from these various measures of poverty, we include longitudinal measures of childhood SES. Health researchers have suggested that multiple SES indicators, rather than a composite SES measure, should be used in analytical models as education and income are associated with health differently.15 Thus, we include multiple longitudinal indicators of childhood SES (eg, per cent of childhood exposed to maternal unemployment), as well as indicators correlated with poverty (eg, parental relationship instability) as covariates.
Hernandez DC, et al. J Epidemiol Community Health 2014;68:478–484. doi:10.1136/jech-2013-203062
Body weight and obesity Finally, research has demonstrated that there is an intergenerational transmission of economic disadvantage and mobility16 and body weight within families.17 18 The longitudinal, two-generational nature of the data provides a unique opportunity to include variables correlated with economic disadvantage and mobility, such as young adults’ maternal grandmother’s education and mother’s childhood family structure. The data also provides an opportunity to control for the intergenerational transmission of weight status within families by including maternal BMI around the same age as her child’s weight status in young adulthood. Thus, the analyses include sophisticated measures that are rarely available in other datasets, resulting in more precise measures that help further our understanding of race, ethnic and gender disparities in young adult OV/OB status.
METHODS AND PROCEDURES Data from the NLSY79 and the linked YA-NLSY file was used. NLSY data are sponsored by the US Department of Labor and have been compiled through the Center for Human Resource Research at the Ohio State University. The original NLSY79 cohort includes a nationally representative sample of 12 686 men and women followed from 1979–2010. The NLSY has surveyed the biological children of these women every 2 years since 1986 and represents over 90% of all the children born to this cohort.
Sample A total of 11 494 children were born to 4928 NLSY79 mothers between 1970 and 2010. The child sample was restricted to young adults who were at least 20 years old by 2010 (3959 children were excluded because they were born prior to the start of data collection in 1979 or were too young). Of these young adults (n=7535), the sample was further restricted to individuals who had valid BMI data at age 20 years (2367 cases excluded) and complete data needed to calculate poverty or measures correlated with poverty (1168 cases excluded). Young adult women who were pregnant during their BMI assessment were excluded from the analyses (99 cases excluded). The final analytical sample consisted of 3901 young adults who were born between 1980 and 1990 (mean=1985). The Institutional Review Board of the Pennsylvania State University approved the study. Compared with the excluded sample, young adults in the analytical sample had a lower BMI and spent more of their childhood in poverty but not as a young adult. The analytical sample was comprised of young adults who were more likely born at a healthy weight to older, married mothers. The young adults were younger, white, male, single, resided with more siblings, displayed less depressive symptoms but tried more substances as an adolescent, and were exposed to less maternal unemployment but greater family instability.
Measures Body mass index Young adults self-reported their height (feet and inches) and weight (pounds) in the YA-NLSY. BMI was calculated and categorised into the reference criteria for adults outlined by the Center for Disease Control (ie underweight (BMI
