Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Jan 1.
Published in final edited form as: Biodemography Soc Biol. 2017;63(2):87–103. doi: 10.1080/19485565.2017.1279536

Early Life Socioeconomic Status and Adult Physiological Functioning: A Life Course Examination of Biosocial Mechanisms

Yang Claire Yang *, Karen Gerken **, Kristen Schorpp **, Courtney Boen **, Kathleen Mullan Harris **
PMCID: PMC5439296  NIHMSID: NIHMS851192  PMID: 28521624

Abstract

A growing literature has demonstrated a link between early-life socioeconomic conditions and adult health at a singular point in life. No research exists, however, that specifies the life course patterns of socioeconomic status (SES) in relation to the underlying biological processes that determine health. Using an innovative life course research design consisting of four nationally representative longitudinal datasets that collectively cover the human life span from early adolescence to old age (Add Health, MIDUS, NSHAP, and HRS), we address this scientific gap and assess how SES pathways from childhood into adulthood are associated with biophysiological outcomes in different adult life stages. For each dataset, we constructed standardized, composite measures of early-life SES and adult SES and harmonized biophysiological measurements of immune and metabolic functioning. We found that the relative importance of early-life and adult SES varied across young-, mid-, and late-adulthood, such that early-life SES sets a life course trajectory of socioeconomic well-being and operates through adult SES to influence health as adults age. We also documented evidence of the detrimental health effects of downward mobility and persistent socioeconomic disadvantage. These findings are the first to specify the life course patterns of SES that matter for underlying biophysiological functioning in different stages of adulthood. The study thus contributes new knowledge critical for improving population health by identifying the particular points in the life course in which interventions might be most effective in preventing disease and premature mortality.

INTRODUCTION

The impacts of social status on physiological, cellular, and molecular processes have been widely documented across species from rodents to nonhuman primates (McClintock et al. 2005, Sapolsky 2005, Tung et al. 2012). In humans, a large body of work indicates the detrimental consequences of adverse social conditions and social stress for disease susceptibility and survival. Socioeconomic status (SES) is among the most important determinants of the quality of social environment that strongly conditions one’s exposures to chronic stress and other risk factors for health (Link and Phelan 1995, Pearlin et al. 1981). Research across disciplines consistently documents an SES gradient in health, with individuals of lower SES having higher rates of disability, disease, and death than higher SES individuals (Adler and Ostrove 1999, Marmot 2006). The ubiquity of SES influences on morbidity and mortality brings into sharp focus the need to reduce status-based health disparities.

SES-related disadvantage is not an isolated point-in-time experience. It reaches across time in individuals’ lives. There has been growing interest in examining the early-life origins of SES disparities in health using a life course perspective (Gluckman and Hanson 2004). A dominant paradigm to explain the early- and later-life connection is the sensitive period model, which posits that exposures during sensitive periods of development (e.g., gestation, birth, childhood, and adolescence) induce enduring structural and functional changes in organisms through biological programming that are difficult to reverse and, in turn, affect later disease risk (Barker 1998, Ben-Shlomo and Kuh 2002, Guo and Harris 2000). Such a model suggests that early-life conditions have stronger effects on individual outcomes than conditions experienced at subsequent time points. Empirical evidence would support this model if low early-life SES increases the risk of disease in later life, independent of adult SES and other risk factors.

Although the sensitive period model has gained wide recognition and stimulated much research, the causal relationship it implies (i.e., early life exerts direct and permanent impacts on later-life adult health) has never been properly tested or firmly established. Critical gaps remain in our knowledge about mechanisms underlying the early- and later-life links that are critical for understanding the pathogenesis of diseases of aging. First, it is not clear that the sensitive period model accounts for any observed associations between early SES exposures and later health because most studies have been heavily biased towards examining only early-life periods without consideration of social conditions during the intervening adult years between early and late-life periods. Second, most studies rely on single indicators of SES (typically occupation) that fail to capture the complex and multidimensional nature of socioeconomic well-being and its potentially varying health impacts across the life course. Third, studies of association between life course SES and later life morbidity and mortality risks lack examination of underlying biological mechanisms. While evidence has emerged on links between social disadvantage and markers of physiological stress response such as immune compromise (Yang et al. 2013, McDade, Lindau, and Wroblewski 2011) and metabolic dysregulation (Sweet et al. 2013, Yang, Li, and Ji 2013), few extant studies move beyond static snapshots of biomarkers in relation to SES at a point in time and cannot inform the dynamics of their interplay as individuals age. The temporal properties as well as specificity of biological processes involved in the multidimensional SES-health link are important for identifying points of intervention but are largely unknown.

This study addresses these fundamental deficits in past research and provides new data on how early-life SES is related to later-life health. We move beyond the notion of early-life sensitive periods and towards a more comprehensive and integrative framework that considers alternative models and biosocial mechanisms that operate across the human life span.

Besides the sensitive period model, three additional conceptual models have been proposed to explain the early SES and later health link. The accumulation of risks model holds that deleterious prior exposures associated with low SES accumulate to compound the adverse effects over time (Willson, Shuey, and Elder Jr 2007, Ferraro and Shippee 2009, O’Rand 2009). Statistical evidence for this model is based on the additive effects of early- and later-life SES factors (Hallqvist et al. 2004). The pathway model posits that childhood circumstances affect adult health risk indirectly, such that early experiences set individuals on divergent socioeconomic trajectories that differentially expose them to stressors that impinge on biological functioning and health (Marmot et al. 2001, Hayward and Gorman 2004, Pudrovska and Anikputa 2014). Statistically, this model predicts that the effects of early-life adversity on later-life disease risk is mediated or transmitted through SES at each subsequent life stage. Lastly, the social mobility model posits that movement across levels of SES may affect disease risk (Hallqvist et al. 2004, Luo and Waite 2005). This model predicts that the health effects of early-life exposures can be modified by later-life SES such that upward mobility may mitigate negative impacts of early-life adversity, while downward mobility from prosperity to hardship can be a perniciously stressful experience.

These models are often positioned as conceptually distinct, but research documents that the life course processes underlying these models are likely to be complimentary rather than mutually exclusive (Hallqvist et al. 2004). Extant studies show mixed support for different conceptual models and highlight the complexity of interrelated life-course processes (Luo and Waite 2005, Lynch et al. 1994, Lynch, Kaplan, and Shema 1997, Osler et al. 2003, Power, Manor, and Matthews 1999). Because prior studies often use a cross-sectional design in which early-life SES (typically reported retrospectively) is related to adult health at a point in time, the timing and pattern of SES across the life course in relation to underlying biological mechanisms that determine health cannot be examined. As a result, previous research cannot adjudicate among different models or explicate their interrelations and our current understanding of the early-life impact on adult health remains largely speculative. It remains unknown whether early or contemporaneous SES is more important or whether there are windows of particular vulnerability throughout life. Studies have generally failed to consider simultaneously the length of exposures to social conditions, timing of the manifestation of their health impact, and change in social exposures in relation to disease risk measured in multiple life stages as aging progresses. There has been no research we are aware of that offers a robust test of the early- and later-life link by accounting for all these problems.

This study breaks new ground by examining the SES-health linkages as they unfold across the life span from young adulthood, to midlife, to old age. Combining four nationally representative, longitudinal data sets, we assess the extent to which the associations between early and adult SES and biomarkers of inflammation and metabolic syndrome are consistent with the sensitive period, accumulation of risks, pathway, and/or social mobility models at various stages in the life course. We make several unique contributions to research on social disparities in health. First, the use of an innovative longitudinal research design consisting of multiple large, diverse, population-based samples (Yang et al. 2016) enables us to empirically examine the early-life SES health impacts over a far more extensive period of life than any single dataset can allow. The comprehensive longitudinal analyses help to better adjudicate among several life course models. Second, this study utilizes multidimensional measures of SES that are harmonized across data sources to provide a more complete view of socioeconomic impacts on health. These measures include parent education, family income and welfare receipt in early life, and individual education, earnings, and wealth in adulthood. Use of multiple SES indicators allows for a more comprehensive assessment of SES within each life stage and facilitates the comparison of SES-health links across the life span. Third, we use objectively measured biomarkers of immune and metabolic functions that are fundamental in physiological regulation underlying the aging processes leading to morbidity and mortality (Finch 2010). This study thus provides a mechanistic view of how early life conditions “get under the skin” to influence disease susceptibility over time. Finally, our study design provides new knowledge about whether and how biological pathways linking SES and disease phenotypes may be specific to each life course stage.

DATA AND METHODS

Data for this study come from four nationally representative NIH studies that collectively span multiple stages of the life course. Data for young adulthood come from 12,237 participants in the National Longitudinal Study of Adolescent to Adult Health (Add Health) aged 12–18 at Wave I (1994–95) and followed up at aged 24–32 in Wave IV (2008–09). Data for mid adulthood come from 908 respondents aged 25–74 in the National Survey of the Midlife Development in the United States (MIDUS) surveyed at Wave I (1995–96) and followed up at Wave II (2004–09). Data for late adulthood come from 10,165 participants aged 50 and older in the Health and Retirement Study (HRS) at baseline in 1998 and followed every two years through 2006, as well as 1,026 respondents aged 57–85 in the National Social Life, Health, and Aging Project (NSHAP) at Wave I (2005–06) and followed up at Wave II (2010–11). More information about the studies used in this analysis can be found in Supplement 1.

Outcome measures represent two key biological pathways underlying the physiological stress process (Finch 2010). For Add Health, HRS, and NSHAP, immune function was measured using C-reactive protein (CRP), an acute phase protein whose elevation in circulating level indicates systemic inflammation. Additional measures of immune function were available in MIDUS, including fibrinogen, interleukin 6 (IL-6), E-selectin, and intracellular adhesion molecule 1 (ICAM-1), and were used to create a composite score of inflammation burden (Yang, Schorpp, and Harris 2014). The cut points for CRP reflect clinical reference ranges for health risk, while the top quartile was used as a cut point for the other markers. Metabolic function was assessed by a composite measure that includes seven biomarkers used in the clinical definition of metabolic syndrome: diastolic blood pressure, systolic blood pressure, HbA1c, waist circumference, high-density lipoprotein (HDL cholesterol), total cholesterol, and triglycerides. For each measure, the cut points for high risk were defined by clinical practice, or empirically defined as the top or bottom decile (see Table 1 for cut points used for each data source). We constructed the index of metabolic syndrome as the sum of the positive indicators. Details of biomarker data collection and assays are provided in Supplement 1.

Table 1.

Sample Characteristics: Weighted Descriptive Statistics

Young Adulthood Young to Mid Adulthood Late Adulthood
Add Health (Age 24–32) MIDUS (Age 34–74) HRS (Age 50–98) NSHAP (Age 57–91)
Wave/Survey Year IV: 2008–09 II: 2004–06 VIII/IX: 2006/2008 II: 2010–11
Biomarkers
 Inflammation
  C-Reactive Protein (CRP, mg/dl), Mean(SD) 3.8 (5.2) 4.5 (8.3) 4.1 (8.4)
   <1 (normal) 32.7% 25.6% 23.1%
   1–3 (low chronic inflammation) 29.2% 36.4% 39.7%
   3–10 (high chronic inflammation) 29.2% 28.4% 29.1%
   >10 (very high inflammation) 9.0% 9.6% 8.1%
   N 12,252 11,251 1,026
  Inflammation burden (MIDUS only)
   0 38.3%
   1 30.1%
   2 16.8%
   3 9.6%
   4 3.8%
   5 1.3%
   N 856
 Metabolic Syndrome
   0 25.5% 10.1% 1.0% 10.0%
   1 34.8% 25.6% 8.3% 24.7%
   2 24.9% 26.8% 18.2% 33.4%
   3 11.1% 25.6% 25.1% 31.9%
   4 3.4% 8.0% 26.9%
   5 0.3% 4.0% 20.5%
   N 12,133 852 10,793 988
  Hypertensiona 26.9% 49.7% 67.9% 70.7%
  Abdominal obesityb 49.1% 50.9% 63.1% 59.3%
  HbA1cc 30.1% 72.3% 52.4% 57.2%
  HDL cholesterold 14.3% 19.7% 59.8%
  Total cholesterole 88.0%
  Triglyceridesf 10.2% 15.1%
Socioeconomic Status
 Early life SES (Mean (SD) composite of standardized SES items) −0.003(0.62) −0.001 (0.57) 0.04 (0.55) −0.05 (0.71)
  Parent educational attainment
   Less that high school 12.1% 19.2% 45.2%
   High school graduate 32.3% 35.5% 33.2%
   Some college 21.8% 17.9% 9.47%
   College or more 33.8% 27.4% 12.1%
  Mother’s educational attainment
   Less that high school 44.3%
   High school graduate 35.5%
   Some college 10.9%
   College or more 9.3%
  Father’s educational attainment
   Less that high school 44.3%
   High school graduate 28.3%
   Some college 13.6%
   College or more 13.9%
  Parent socioeconomic index 46.3 (47.1) 73.2 (29.7)
  Ever received welfare in childhood 17.3% 5.1%
  Rate childhood SES
   Poor 31.6% 41.2%
   Average 61.9% 46.8%
   Pretty well off 6.5% 12.0%
  Father unemployed 21.5%
 Adult SES (Mean (SD) composite of standardized SES items) −0.003(0.52) 0.003 (0.53) 0.07 (0.72) 0.15 (0.80)
  Educational attainment
   Less that high school 8.4% 3.1% 16.5% 8.0%
   High school graduate 17.7% 23.5% 53.1% 56.4%
   Some college 43.7% 28.4% 5.4%
   College or more 30.1% 45.0% 24.9% 35.6%
  Household income (thousands) 63.8 (44.4) 84.2 (61.7) 71.42 (106.64) 64.7 (88.6)
  Ever received welfare in adulthood 23.6% 10.2%
  Household assets 439.5 (1,003.3) 815.3 (2,082.9)
Covariates
 Age 28.3 (1.9) 44.2 (10.0) 63.2 (9.5) 66.6 (7.0)
 Female 51.1% 55.4% 53.9% 48.8%
 Race/ethnicity
  White 68.6% 93.6% 87.2% 85.7%
  Black 15.0% 2.9% 8.2% 6.4%
  Hispanic 11.9% 3.5% 4.6% 5.3%
  Other 4.5% 2.6%
 Marital status 50.2% 67.3% 65.9%
 Self-rated childhood health 3.9 (0.9) 4.5 (0.8) 4.1 (0.9)
 Social isolation 25.4% 0.3 (0.2) 17.4%
 CES-D 5.3 (4.2) 1.3 (1.8) 4.1 (3.9)
 Ever cigarette smoker 25.3% 56.4% 59.5%
 Physical activity 14.5% 1.0 (1.3) 69.0%
 Obese 37.8% 39.4% 37.7%
Open in a new tab
a

Hypertension =1 if systolic >=140 mmHg or diastolic >=90 mmHg or on anti-hypertensive medication, or ever diagnosed hypertensive.

b

Abdominal obesity =1 if waist circumference >102 cm in males, >88 cm in females.

c

HbA1c =1 if glucose >5.6 % or ever diagnosed diabetic.

d

HDL cholesterol =1 if <=40 mg/dl or on cholesterol medication. For Add Health, = 1 if in the lowest decile for females or lowest two deciles for males.

e

Total cholesterol =1 if >=180 mg/dl or on cholesterol medication.

f

Triglycerides =1 if >=200 mg/dl. For Add Health, = 1 if in the highest decile.

For each data set, measures of both early-life and adult SES were used to examine life course patterns of SES. Early-life SES measures include parental education, household income, welfare receipt, and subjective financial wellbeing. Parental education was coded as 1=less than high school, 2=high school graduate, 3=some college, and 4=college graduate or more. For Add Health, MIDUS, and HRS, parent education was measured as the maximum of mother and father education in two-parent households. For NSHAP, mother and father education were incorporated as separate SES indicators. Parent-reported household income was included as an indicator of early-life SES for Add Health, and was coded as a continuous measure. While household income in early life was not available for the other data sources, MIDUS included a continuous parent socioeconomic index (SEI) that was based on parent education and occupation, which we used as a proxy for household income. In addition, HRS includes an item that asked respondents if their father was unemployed for a period of several months or more when the respondent was under the age of 16, which was reverse coded to reflect father employment. Welfare receipt was used as an SES indicator for Add Health and MIDUS, and was reverse coded to indicate higher SES. Finally, both HRS and NSHAP included a measure of perceived financial well-being, in which respondents were asked whether their family was “pretty well off,” “about average,” or “poor” when the respondent was before the age of 16. Items were recoded so higher values reflect higher SES. Adult SES measures include respondent education, household income, welfare receipt, and household assets. Respondent education was included in all four data sources and was coded similarly to parent education, with 1=less than high school, 2=high school graduate, 3=some college, and 4=college graduate or more. Household income was also included for all four data sources and was coded as a continuous measure. For Add Health and MIDUS, welfare receipt was used as an indicator of adult SES, and was recoded to reflect higher SES. Finally, NSHAP and HRS included a continuous measure of total household assets, which is an especially important indicator of SES in late adulthood.

Previous research suggests that multivariate scales of SES are more reliable than single measures (Walsemann, Bell, and Goosby 2011, Goosby and Walsemann 2012) and reduce problems related to missing data on the SES variables. We constructed composite measures of early-life SES and adult SES that are composed of several indicators of SES available within each data source, resulting in eight composite SES scores (one early-life SES measure and one adult SES measure for all four data sources). For each composite SES score, three individual SES items were first standardized by calculating the z-scores for each item. The mean of the three z-scores was then calculated to produce a continuous, composite SES score. Positive values for the SES scores indicate higher levels of SES. To maximize sample size, composite measures of early-life and adult SES were calculated for all respondents who had data on at least two of the variables used in each scale.

In addition to continuous measures of SES, we also include categorical measures of SES, which indicate social status trajectories using the lowest quartile of each SES composite measure to indicate socioeconomic disadvantage (1 = disadvantaged, 0 = not disadvantaged). Based on the dichotomous SES measure, we constructed measures of change from early to current status that include four categories: persistent disadvantage (disadvantaged – disadvantaged), upward mobility (disadvantaged – not disadvantaged), downwardly mobile (not disadvantaged – disadvantaged) and never disadvantaged (not disadvantaged – not disadvantaged). We include a measure of childhood health in all models to control for health selection early in life. For all four data sources, childhood health was measured using a single item of self-rated health, with response categories of poor, fair, good, very good, and excellent. Add Health childhood health was collected during Wave I of the study, when respondents were adolescents. Childhood health for MIDUS, NSHAP, and HRS was retrospective. Childhood health was operationalized as continuous, with higher values indicating better health. Descriptive statistics on the outcome, SES measures, childhood health and covariates used in analyses for each of the four data sets are shown in Table 1.

For each outcome, we conducted two sets of analyses. First, we examined the associations between early-life SES, adult SES, and biomarkers assessed at follow-up using ordinal logit models both for inflammation and metabolic syndrome. We estimate models in a stepwise fashion: 1) Model I adjusts for early-life SES; 2) Model II adjusts for adult SES; and 3) Model III simultaneously adjusts for both early-life and adult SES. We control for childhood health in all models to account for potential selection of individuals of varying health status early in life into SES classes in adulthood. Second, to assess the utility of the social mobility model, we model the biomarker outcomes as a function of changes across SES categories or movement into and out of SES disadvantage. All models adjust for survey design effects and nonresponse using sampling weights and control for age, sex, race, and marital status. Analyses were conducted in Stata 13.

RESULTS

We found strong associations of SES measures with both indices of physiological functioning across all life stages; however, the relative importance of early-life and adult SES varied across young-, mid-, and late-adulthood, providing evidence consistent with different life course models. As seen in Model I of Table 2, higher early-life SES is associated with a 15% lower odds of elevated CRP (OR = 0.84, CI = [0.77, 0.91]) and a 20% lower odds of metabolic syndrome (OR = 0.80, CI = [0.74, 0.87]) in the Add Health study. These associations are highly significant and replicated in the MIDUS, NSHAP, and HRS samples with similar magnitude. In Model II, adult SES showed inverse and significant associations with both physiological outcomes across all life stages, with the effect sizes for inflammation increasing from young to old age.

Table 2.

Associations of Early-Life and Current SES with Markers of Physical Health across the Lifespan, Odds Ratios (95% Confidence Intervals)

Young Adulthood (Add Health) Middle Adulthood (MIDUS) Old Age (NSHAP) Old Age (HRS)
I II III I II III I II III I II III
Inflammation
 Early life SES 0.84*** (0.78–0.91) 0.86*** (0.79–0.93) 0.71** (0.58–0.88) 0.78* (0.63–0.97) 0.83* (0.70–0.98) 0.88 (0.75–1.04) 0.86*** (0.80–0.93) 0.94 (0.87–1.02)
 Current SES 0.87** (0.79–0.96) 0.90* (0.82–0.99) 0.68*** (0.56–0.82) 0.72** (0.58–0.88) 0.77** (0.66–0.89) 0.80** (0.68–0.93) 0.69*** (0.63–0.76) 0.70*** (0.64–0.77)
 N 12237 12237 12237 856 856 856 1026 1026 1026 10165 10165 10165
Metabolic Syndrome
 Early life SES 0.80*** (0.74–0.87) 0.83*** (0.77–0.91) 0.68*** (0.56–0.84) 0.74** (0.60–0.91) 0.80* (0.66–0.97) 0.88 (0.70–1.09) 0.87*** (0.81–0.94) 0.94 (0.86–1.01)
 Current SES 0.78*** (0.71–0.85) 0.80*** (0.73–0.88) 0.71*** (0.59–0.86) 0.76** (0.63–0.93) 0.71* (0.54–0.92) 0.73* (0.54–0.99) 0.78*** (0.71–0.84) 0.79*** (0.72–0.86)
 N 12133 12133 12133 852 852 852 988 988 988 9412 9412 9412
Open in a new tab
***

p<0.001;

**

p<0.01;

*

p<0.05;

†

p<0.1

Note: All models adjust for age, sex, race, marital status, and self-rated health in childhood.

The simultaneous adjustment for both early-life and adult SES in Model III yields different results across life stage samples. In young and mid adulthood (Add Health and MIDUS), both early-life and adult SES had significant, negative associations with CRP/inflammation burden and metabolic syndrome, suggesting an accumulation of risks model whereby health risks or benefits associated with SES accumulate over the life time to additively affect physiological functioning. In late adulthood (NSHAP and HRS), current SES completely mediated the associations between early-life SES and both biomarker outcomes, supporting a pathway model whereby the influence of childhood SES on adult physiology operates through adult SES. Supplementary analyses utilizing structural equation models (SEM) were consistent with the results of the regression analyses presented here and offered additional evidence that adult SES fully mediated the association between early-life SES and the biomarkers in the NSHAP and HRS samples (results available upon request). Results from the SEM models showed that, in both older adult samples, early-life SES indirectly affected the biomarker outcomes through its impact on adult SES. We note that the results using the NSHAP study and the HRS are highly consistent, which shows the robustness of the strong and dominant effects of adult SES in late life. Childhood general health status, while significant in predicting biomarkers in young adulthood, is not consistently associated with biomarkers in other later periods of life. The adjustment of childhood health did not affect the SES-biomarker associations assessed at any point in life. It is thus unlikely that health selection explains any of the above findings.

Analysis of SES change over time provides tentative evidence for the social mobility model across life stages. Figure 1 summarizes the results regarding the associations between different mobility categories and biomarker outcomes. Downward mobility was associated with lower risk of metabolic syndrome, compared to those disadvantaged, in young and middle adulthood. In young adulthood, for example, we find a 25% lower risk of metabolic syndrome for downwardly mobile individuals, compared to individuals in persistent disadvantage. Upward mobility was also associated with lower risk of inflammation or metabolic syndrome than the persistently disadvantaged, except in late adulthood. For example, in mid-life, upward mobility was associated with a 39% lower risk of metabolic syndrome. The suggestion of adverse health effects associated with upward mobility documented in late adulthood has emerged in other recent research (Miller et al. 2015). We note, however, the never disadvantaged had lower health risks of all mobility categories compared. Taken together, we find that SES disadvantage at any point in life put individuals at higher physiological risks for health.

Figure 1. Associations of socioeconomic mobility with markers of physical health.

Figure 1

Open in a new tab

p-values indicate overall significance of SES mobility variables within each data set. Reference category is persistent disadvantage in early life and adulthood.

DISCUSSION

Our study contributes new knowledge on the complex temporal dynamics through which SES impacts physiological indicators of health across the life span. Our results demonstrate how SES operates through several life course and biological mechanisms to affect health, and how these mechanisms evolve across young-, mid-, and late-adulthood. In particular, we find that the direct effect of early-life SES on physical health wanes as individuals age, though its indirect effect remains important through its influence on the SES trajectory into adulthood

In early and middle adulthood, we find evidence of an accumulation model, as there are additive and independent associations of both early-life SES and young adult SES with health risks. In old age, early-life SES operates through adult SES in its association with health risks later in the life course, which supports the pathway model. Early-life SES remains salient for health in young to mid-adulthood, over and above current adult SES, perhaps because the resources and support that early-life SES provides continue to play an independent role in adult lifestyles and access to health care. Once adults reach old age, the SES they have established as older adults has been fairly stable over many years and early-life SES is only important through its linkage to this longer-term SES. We also find that social mobility matters across the life course. Interestingly, mobility effects do not exist independently from one’s location on the SES hierarchy. Regardless of the direction of the movement over time from childhood to adulthood, being in the lowest SES quartile at any point is detrimental to physical health.

Our extensive longitudinal life course study design has offered clear advantages in its ability to adjudicate among different life course models compared to previous research. By examining whether and how early-life SES continues to matter in the presence of adult SES across multiple stages of the adult life course, we are less likely to overlook patterns of associations in the data than previous studies, which only evaluate these associations in one particular period of the life course. Findings that have emerged from this study indicate temporal specificity of exposures to social adversity that can better inform disease intervention and control strategies aimed at minimizing socioeconomic disparities in health. We have thus provided an initial framework for future analysis to incorporate additional datasets and explore the behavioral and psychosocial mechanisms that might explain some of the observed linkages between life course SES and health.

Despite these advances, we acknowledge our study has a number of limitations. First, we lack multiple repeated measures of biomarkers and SES across the early life course, beyond childhood and adolescence, to allow for a closer tracking of the SES trajectories across time in relation to subsequent changes in physical indicators. In additional NSHAP analysis where we used two waves of biomarker measures to estimate longitudinal residual change models, we did not find early life SES to be significantly associated with change in biomarkers over a five-year period in late life. While this is consistent with and strengthens the current finding, it is not known whether the association under question varies by interval length of biomarker measurements within and across life course stages. Second, we cannot exclude the possibility that mortality selection may be occurring with older or frailer adults. The results are thus likely conservative regarding the extent to which early-life and adult SES influence physiological functioning in the older adult population in that those most deprived of SES related resources earlier in life for whom the SES-biomarker associations tend to be strongest were less likely to have survived to be observed in old age. Third, while the results presented here offer preliminary evidence of the social mobility model, more research incorporating multiple repeated measures of SES is needed. For each of the samples included in the analysis, we have measures of SES at two time points, which limits our ability to fully understand how changes in SES across the life course affect health trajectories, especially among the older aged cohorts for whom childhood occurred 40–60 years ago. Though this study provides initial evidence of an association between social mobility and health risk, more research incorporating longitudinal SES data is needed.

Nonetheless, we provide new scientific data on when and how SES matters for underlying biophysiological functioning across the human life span, which has important implications for interventions to improve population health. It is clear that early-life SES matters for later life health. This finding suggests that policies and interventions aimed at improving the socioeconomic conditions of children in disadvantaged families will reduce SES disparities in health in all stages of the life span—from early adulthood through middle and late adulthood. Interventions designed to enhance the education, training, income, job skills and job security of parents in low-SES families will directly improve the health of their children as they enter adulthood and set them on an SES trajectory that will continue to benefit their health into old age. Our findings also suggest directions for further research on SES and health. More in-depth study of the timing and patterns of SES across the life course and specific biological response to SES-related stress are needed to better understand the lifetime health impacts of SES. The varying life stage relationship between SES and physical health that our study has shown invites follow-up studies to provide deeper mechanistic explanations for the life course development of social disparities in health.

Supplementary Material

Acknowledgments

This research is supported by the National Institute of Aging grant K01AG036745 and University Cancer Research Funds at the Lineberger Cancer Center (to the first author), and by the Eunice Kennedy Shriver National Institute of Child Health and Human Development grant P01HD31921 (to the last author). We are grateful for training support (T32 HD007168) and for general research support (R24 HD050924) from the Carolina Population Center, University of North Carolina at Chapel Hill. This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. We are grateful to our colleagues, especially Robert Hummer, for helpful suggestions and comments.

References

  1. Adler Nancy E, Ostrove Joan M. Socioeconomic status and health: what we know and what we don’t. Ann N Y Acad Sci. 1999;(896):3–15. doi: 10.1111/j.1749-6632.1999.tb08101.x. [DOI] [PubMed] [Google Scholar]
  2. Barker David J. In utero programming of chronic disease. Clin Sci (Lond) 1998;95(2):115–28. [PubMed] [Google Scholar]
  3. Ben-Shlomo Yoav, Kuh Diana. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. Int J Epidemiol. 2002;31(2):285–93. [PubMed] [Google Scholar]
  4. Brim Orville Gilbert, Ryff Carol D, Kessler Ronald C. How healthy are we?: A national study of well-being at midlife. University of Chicago Press; 2004. [Google Scholar]
  5. Crimmins Eileen, et al. Documentation of Biomarkers in the 2006 and 2008 Health and Retirement Study. 2013 Available at: http://hrsonline.isr.umich.edu/sitedocs/userg/Biomarker2006and2008.pdf.
  6. Dienberg Love Gayle, Seeman Teresa E, Weinstein Maxine, Ryff Carol D. Bioindicators in the MIDUS national study: protocol, measures, sample, and comparative context. J Aging Health. 2010;22(8):1059–80. doi: 10.1177/0898264310374355.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Entzel Pamela, Whitsel Eric A, Richardson Andrea, Tabor Joyce, Hallquist Suzanne, Hussey Jon, Halpern Carolyn T, Harris Kathleen Mullan. Add Health Wave IV Documentation: Cardiovascular and Anthropometric Measures. 2009 Available at: http://www.cpc.unc.edu/projects/addhealth/data/guides/Wave%20IV%20cardiovascular%20and%20anthropometric%20documentation%20110209.pdf.
  8. Ferraro Kenneth F, Shippee Tatyana P. Aging and cumulative inequality: how does inequality get under the skin? Gerontologist. 2009;49(3):333–43. doi: 10.1093/geront/gnp034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Finch Caleb E. The Biology of Human Longevity: Inflammation, Nutrition, and Aging in the Evolution of Lifespans. Academic Press; 2010. [Google Scholar]
  10. Gluckman Peter D, Hanson Mark A. Living with the past: evolution, development, and patterns of disease. Science. 2004;305(5691):1733–6. doi: 10.1126/science.1095292. [DOI] [PubMed] [Google Scholar]
  11. Goosby Bridget J, Walsemann Katrina M. School racial composition and race/ethnic differences in early adulthood health. Health & place. 2012;18(2):296–304. doi: 10.1016/j.healthplace.2011.10.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Guo Guang, Harris Kathleen M. The mechanisms mediating the effects of poverty on children’s intellectual development. Demography. 2000;37(4):431–47. doi: 10.1353/dem.2000.0005. [DOI] [PubMed] [Google Scholar]
  13. Hallqvist Johan, Lynch John, Bartley Mel, Lang Thierry, Blane David. Can we disentangle life course processes of accumulation, critical period and social mobility? An analysis of disadvantaged socio-economic positions and myocardial infarction in the Stockholm Heart Epidemiology Program. Soc Sci Med. 2004;58(8):1555–62. doi: 10.1016/s0277-9536(03)00344-7. [DOI] [PubMed] [Google Scholar]
  14. Harris Kathleen Mullan. An integrative approach to health. Demography. 2010;47(1):1–22. doi: 10.1353/dem.0.0091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hayward Mark D, Gorman Bridget K. The long arm of childhood: the influence of early-life social conditions on men’s mortality. Demography. 2004;41(1):87–107. doi: 10.1353/dem.2004.0005. [DOI] [PubMed] [Google Scholar]
  16. Heeringa Steven G, Connor Judith H. Technical Description of the Health and Retirement Study Survey Sample Design. 1995 Available at: http://hrsonline.isr.umich.edu/sitedocs/userg/HRSSAMP.pdf.
  17. Link Bruce G, Phelan Jo. Social conditions as fundamental causes of disease. J Health Soc Behav. 1995;(Spec No 80–94) [PubMed] [Google Scholar]
  18. Luo Ye, Waite Linda J. The impact of childhood and adult SES on physical, mental, and cognitive well-being in later life. J Gerontol B Psychol Sci Soc Sci. 2005;60(2):S93–S101. doi: 10.1093/geronb/60.2.s93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lynch John W, Kaplan George A, Cohen Richard D, Kauhanen Jussi, Wilson Thomas W, Smith Nicholas L, Salonen Jukka T. Childhood and adult socioeconomic status as predictors of mortality in Finland. Lancet. 1994;343(8896):524–7. doi: 10.1016/s0140-6736(94)91468-0. [DOI] [PubMed] [Google Scholar]
  20. Lynch John W, Kaplan George A, Shema Sarah J. Cumulative impact of sustained economic hardship on physical, cognitive, psychological, and social functioning. N Engl J Med. 1997;337(26):1889–95. doi: 10.1056/nejm199712253372606. [DOI] [PubMed] [Google Scholar]
  21. Marmot Michael G. Status syndrome: a challenge to medicine. JAMA. 2006;295(11):1304–7. doi: 10.1001/jama.295.11.1304. [DOI] [PubMed] [Google Scholar]
  22. Marmot Michael, Shipley Martin, Brunner Eric, Hemingway Harry. Relative contribution of early life and adult socioeconomic factors to adult morbidity in the Whitehall II study. J Epidemiol Community Health. 2001;55(5):301–7. doi: 10.1136/jech.55.5.301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. McClintock Martha K, Conzen Suzanne D, Gehlert Sarah, Masi Christopher, Olopade Funmi. Mammary cancer and social interactions: identifying multiple environments that regulate gene expression throughout the life span. J Gerontol B Psychol Sci Soc Sci. 2005;60(Spec No 1):32–41. doi: 10.1093/geronb/60.special_issue_1.32. [DOI] [PubMed] [Google Scholar]
  24. McDade Thomas W, Lindau Stacy T, Wroblewski Kristen. Predictors of C-reactive protein in the national social life, health, and aging project. J Gerontol B Psychol Sci Soc Sci. 2011;66(1):129–36. doi: 10.1093/geronb/gbq008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Miller Gregory E, Yu Tianyi, Chen Edith, Brody GH. Self-control forecasts better psychosocial outcomes but faster epigenetic aging in low-SES youth. Proc Natl Acad Sci U S A. 2015;112(33):10325–30. doi: 10.1073/pnas.1505063112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. O’Muircheartaigh Colm, Eckman Stephanie, Smith Steven. Statistical design and estimation for the national social life, health, and aging project. J Gerontol B Psychol Sci Soc Sci. 2009;64(Suppl 1):i12–9. doi: 10.1093/geronb/gbp045.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. O’Rand Angela M. Cumulative processes in the life course. In: Elder Glen H, Jr, Giele Janet Z., editors. The craft of life course research. Guilford Press; 2009. [Google Scholar]
  28. Osler Merete, Andersen Anne M, Due Pernille, Lund Rikke, Damsgaard Mogens T, Holstein Bjorn E. Socioeconomic position in early life, birth weight, childhood cognitive function, and adult mortality. A longitudinal study of Danish men born in 1953. J Epidemiol Community Health. 2003;57(9):681–6. doi: 10.1136/jech.57.9.681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pearlin Leonard I, Lieberman Morton A, Menaghan Elizabeth G, Mullan Joseph T. The stress process. J Health Soc Behav. 1981;22(4):337–56. [PubMed] [Google Scholar]
  30. Power Chris, Manor Orly, Matthews Sharon. The duration and timing of exposure: effects of socioeconomic environment on adult health. Am J Public Health. 1999;89(7):1059–65. doi: 10.2105/ajph.89.7.1059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pudrovska Tetyana, Anikputa Benedicta. Early-life socioeconomic status and mortality in later life: an integration of four life-course mechanisms. J Gerontol B Psychol Sci Soc Sci. 2014;69(3):451–60. doi: 10.1093/geronb/gbt122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Radler Barry T, Ryff Carol D. Who participates? Accounting for longitudinal retention in the MIDUS national study of health and well-being. J Aging Health. 2010;22(3):307–31. doi: 10.1177/0898264309358617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sapolsky Robert M. The influence of social hierarchy on primate health. Science. 2005;308(5722):648–52. doi: 10.1126/science.1106477. [DOI] [PubMed] [Google Scholar]
  34. Sweet Elizabeth, Nandi Arijit, Adam Emma K, McDade Thomas W. The high price of debt: household financial debt and its impact on mental and physical health. Soc Sci Med. 2013;91:94–100. doi: 10.1016/j.socscimed.2013.05.009.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Tung Jenny, Barreiro Luis B, Johnson Zachary P, Hansen Kasper D, Michopoulos Vasiliki, Toufexis Donna, Michelini Katelyn, Wilson Mark E, Gilad Yoav. Social environment is associated with gene regulatory variation in the rhesus macaque immune system. Proc Natl Acad Sci U S A. 2012;109(17):6490–5. doi: 10.1073/pnas.1202734109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Walsemann Katrina M, Bell Bethany A, Goosby Bridget J. Effect of school racial composition on trajectories of depressive symptoms from adolescence through early adulthood. Race Soc Prob. 2011;3(3):131–145. [Google Scholar]
  37. Williams Sharon R, McDade Thomas W. The use of dried blood spot sampling in the national social life, health, and aging project. J Gerontol B Psychol Sci Soc Sci. 2009;64(Suppl 1):i131–6. doi: 10.1093/geronb/gbn022.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Willson Andrea E, Shuey Kim M, Elder Glen H., Jr Cumulative advantage processes as mechanisms of inequality in life course health. Am J Sociology. 2007;112(6):1886–1924. [Google Scholar]
  39. Yang Yang C, Boen Courtney, Gerken Karen, Li Ting, Schorpp Kristen, Harris Kathleen M. Social relationships and physiological determinants of longevity across the human life span. Proc Natl Acad Sci U S A. 2016;113(3):578–83. doi: 10.1073/pnas.1511085112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yang Yang C, Li Ting, Ji Yinchun. Impact of social integration on metabolic functions: evidence from a nationally representative longitudinal study of US older adults. BMC Public Health. 2013;13:1210. doi: 10.1186/1471-2458-13-1210.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yang Yang C, McClintock Martha K, Kozloski Michael, Li Ting. Social isolation and adult mortality: the role of chronic inflammation and sex differences. J Health Soc Behav. 2013;54(2):183–203. doi: 10.1177/0022146513485244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Yang Yang C, Schorpp Kristen, Harris Kathleen M. Social support, social strain and inflammation: evidence from a national longitudinal study of U.S. adults. Soc Sci Med. 2014;107:124–35. doi: 10.1016/j.socscimed.2014.02.013. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS851192-supplement-supplement_1.pdf (210.4KB, pdf)

RESOURCES