Life-Course Socioeconomic Position and Type 2 Diabetes Mellitus: The Framingham Offspring Study (original) (raw)

• Journal List
• [Am J Epidemiol](/pmc/?term=%22Am J Epidemiol%22[journal])
• PMC3032804

Am J Epidemiol. 2011 Feb 15; 173(4): 438–447.

The Framingham Offspring Study

*Correspondence to Brendan T. Smith, Institute for Work and Health, 481 University Avenue, Suite 800, Toronto, Ontario, Canada, M5G 2E9 (e-mail: ac.no.hwi@htimsb).

Received 2010 Jun 2; Accepted 2010 Oct 8.

Copyright American Journal of Epidemiology © The Author 2011. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Supplementary Materials

Web Tables

GUID: 1619CD8B-6579-4285-BC50-0FDFF662A34E

GUID: 474435DA-6AA4-4183-B7FB-AAD975A9E213

Abstract

Evidence is lacking on whether the duration and timing of low socioeconomic position (SEP) across a person's life course may be associated with incidence of type 2 diabetes mellitus (T2D). The authors’ objectives were to investigate associations between cumulative SEP and the incidence of T2D in the Framingham Offspring Study (n = 1,893; 52% women; mean baseline age = 34 years). Pooled logistic regression analyses demonstrated that age-adjusted cumulative SEP was associated with T2D in women (for low vs. high cumulative SEP, odds ratio (OR) = 1.92, 95% confidence interval (CI): 1.08, 3.42). Age-adjusted analyses for young-adulthood SEP (7.85 for ≤12 vs. >16 years of education, OR = 2.84, 95% CI: 1.03), active professional life SEP (for laborer vs. professional/executive/supervisory/technical occupations, OR = 2.40, 95% CI: 1.05, 5.47), and social-mobility frameworks (for declining life-course SEP, OR = 2.99, 95% CI: 1.39, 6.44; for stable low vs. stable high life-course SEP, OR = 1.85, 95% CI: 1.02, 3.35) all demonstrated associations between low SEP and T2D incidence in women. No association was observed between childhood SEP and T2D in women for father's education (some high school or less vs. any postsecondary education, OR = 1.26, 95% CI: 0.72, 2.22). In men, there was little evidence of associations between life-course SEP and T2D incidence. These findings suggest that cumulative SEP is inversely associated with incidence of T2D in women, and that this association may be primarily due to the women's educational levels and occupations.

Keywords: adult, diabetes mellitus, educational status, incidence, occupations, parents, risk factors, socioeconomic factors

In 2007, an estimated 17.9 million Americans had physician-diagnosed diabetes; that number is projected to rise to 48.3 million by 2050 (1–3). Approximately 90%–95% of diabetes cases are type 2 diabetes mellitus (T2D) (4), a disease that disproportionately affects disadvantaged populations (5).

Little is known about how specific socioeconomic trajectories experienced across a person's life course influence T2D incidence. A number of models have been proposed to conceptualize life-course socioeconomic position (SEP). The accumulation-of-risk model represents the summation of SEP effects that interact to increase disease risk across a person's life course (6). In contrast, the critical/sensitive-periods model acknowledges that individuals could have heightened vulnerability to low SEP exposure during specific periods in their life courses, resulting in permanent (critical period) or modifiable (sensitive period) changes in disease risk (6). Finally, the social-mobility model reflects the fact that an individual's SEP is dynamic, and it incorporates the trajectory of socioeconomic mobility across one's lifetime in determining disease risk (6).

With regard to the sensitive-periods SEP model, inverse associations have been found between childhood SEP (measured as parents’ educational level or occupation) and incidence (7, 8) and prevalence (9–11) of T2D. Similarly, the results of prospective (7, 10, 12–15) and cross-sectional (16–20) studies typically have shown inverse associations between adult SEP (measured as educational level, occupation, and income) and T2D. There is some indication that this association may be sex-specific. Women with low SEP in childhood (7–11) and adulthood (7, 10, 13, 15–20) have a consistently increased risk of T2D. However, there have been conflicting results for men. Inverse associations between childhood SEP and T2D were observed in some (8–10) but not all (11) studies. Furthermore, some studies found inverse associations between adult SEP and T2D (14, 20); however, most reported weak or no associations (10, 13, 16–19). With regard to the social-mobility SEP framework, initial studies have found that downward and stable low SEP trajectories from childhood to adulthood were associated with increased T2D risk compared with a stable high SEP trajectory in women (7, 11). Very little is known about the association between cumulative SEP and T2D.

Using a life-course SEP approach allows the examination of both the timing and duration of exposure to socioeconomic environments. We therefore could achieve a better understanding of the mechanisms by which socioeconomic disadvantage across the life course might lead to T2D. This adds to the growing literature that interventions up to young adulthood may be an additional source of effective public health approaches to reducing social disparities in rates of T2D.

The primary objective in the present study was to evaluate the extent to which cumulative life-course SEP is associated with T2D incidence in adulthood and whether this association differs by sex. The secondary objectives were to investigate whether childhood and adulthood are sensitive periods in which low SEP increases T2D risk, as well as to determine the impact of social mobility on T2D incidence.

MATERIALS AND METHODS

Study sample

Data were drawn from the prospective Framingham Offspring Study, which was initiated in 1971 and included over 30 years of follow-up on 5,124 male and female offspring (or spouses of offspring) of participants from the Framingham Heart Study original cohort, which has been described in detail elsewhere (21). Clinical examinations, during which each participant completed standardized questionnaires and a physician-administered medical history and underwent a physical examination, were performed approximately every 4 years. The Framingham Heart Study is annually reviewed by the Boston University Medical Center Institutional Review Board.

There were 4,989 Framingham Offspring Study participants in the National Heart, Lung, and Blood Institute repository data set. We excluded 2,136 participants who did not have a father in the original cohort of the Framingham Heart Study (1,598 participants had no parents in original cohort and 538 participants had a mother but no father in original cohort) in which fathers' educational levels were measured. Of the 2,853 eligible participants, those for whom data on fathers' educational level (n = 119), the participant's educational level (n = 531), or the participant's occupation (n = 234) were excluded from analyses. Reasons for missing data included death between examinations 1 and 2 (n = 98), lack of attendance at examinations 2 and 3, during which educational attainment was measured (n = 299), lack of attendance at examination 2, during which occupation was measured (n = 262), unemployment (n = 7), retirement (n = 2), and lack of response to the educational level or occupation questions (n = 97). To minimize misclassification of educational level, we excluded participants who were <28 years of age when their educational level was recorded (n = 53). We assumed then when an individual reached 28 years of age, he or she had either completed his or her education or would be included in the highest educational category as defined in this study. Participants with baseline diabetes (n = 20) or missing information on diabetes status at all examinations (n = 3) were also excluded. The final sample consisted of 1,893 participants (991 women and 902 men).

Childhood SEP: fathers' educational levels

Fathers' educational levels were measured during the enrollment of the participants’ fathers in the original cohort of the Framingham Heart Study (1948–1950; mean age of the fathers = 44 years; range, 28–62 years). This approach provided a direct measure of childhood SEP, which is a more robust measure than the commonly used historical recall measures (22). The fathers’ educational levels were grouped into 3 categories: low (some high school or less), medium (high school graduate), and high (any postsecondary education). Secondary analyses were conducted using each participant's father's occupation as a measure of childhood SEP. SEP was categorized into 3 categories: low (laborer), medium (clerical or sales occupations), and high (professional, executive, supervisory, or technical occupations).

Young-adulthood SEP: participants’ educational levels

Educational attainment typically starts early in life and is completed by adolescence or young adulthood. Access to and pursuit of higher education are influenced by both parental characteristics and early life exposures and are strong determinants of adult SEP (e.g., occupation and income) (23). Participants’ educational levels were measured at examination 2 (1979–1982; mean age = 44 years; range, 21–68 years) and examination 3 (1984–1987; mean age = 48 years, range: 24–72 years). To allow participants to reach their maximal educational attainment, the examination 3 data were prioritized. However, if examination 3 educational level data were missing, examination 2 educational level data were used (n = 547, 29% of sample). This likely resulted in minimal misclassification, as education assessed in adulthood remains stable throughout the life course. Educational level-based SEP was grouped into 3 categories based on years of education: low (≤12 years), medium (13–16 years), and high (≥17 years).

Active professional life SEP: participants’ occupations

Participants’ occupations were assessed at examination 2 (1979–1982) and classified into 4 groups: low (laborer), medium (clerical or sales occupations), or high (professional, executive, supervisory, or technical occupations), or homemaker. The homemaker category was considered separate, as this occupation could be found across all family socioeconomic strata.

Life-course SEP frameworks

To test the accumulation-of-risk model, a cumulative SEP measure was created by summing scores for childhood SEP (fathers' education: low = 0, medium = 1, and high = 2), young-adulthood SEP (participant's education: low = 0, medium = 1, and high = 2), and active professional-life SEP (participant's occupation: low = 0, medium or homemaker = 1, and high = 2). The cumulative SEP score ranged from 0 to 6, with higher scores reflecting increased exposure to high SEP. Cumulative SEP was categorized into low (score = 0–1), medium (score = 2–3), and high (score = 4–6) exposure to high life-course SEP, as in other studies (24, 25). These categories were based on theoretical considerations of how cumulative SEP might represent an accumulation of exposures throughout the life course (6). The homemaker category was combined with the closest other SEP category (clerical/sales occupations) when calculating the effect size of the relation between cumulative SEP and T2D incidence. This enabled us to use a 3-level summary score for each life-course period. Higher educational thresholds were used to classify SEP for a participant's, compared with a father's, educational level to account for secular trends in educational attainment across generations of the Framingham Heart Study.

The sensitive-periods framework calculated the effect of SEP measured during each period in the life course (childhood, young adulthood, and active professional life) individually while simultaneously adjusting for other SEP measures.

The social-mobility framework was tested using life-course SEP trajectories. Each SEP measure was dichotomized as follows: fathers' educational level, low (some high school or less) vs. high (high school graduate or above); participant's educational level, low (≤12 years) vs. high (>12 years); and participant's occupation, low (laborer) vs. high (homemaker/clerical/sales/supervisory/technical/professional/executive), as was done in other studies (24, 25). These categorizations were consistent with the theoretical underpinnings of trajectories in life-course epidemiology (6). The analyses used fathers' educational level as a measure of childhood SEP and compared trajectories with both the participant's educational level and occupation (in separate analyses). Four potential SEP trajectories were investigated: stable high SEP (high childhood and high adulthood), declining SEP (high childhood and low adulthood), increasing SEP (low childhood and high adulthood), and stable low SEP (low childhood and low adulthood).

T2D ascertainment

T2D was diagnosed on the basis of the American Diabetes Association criteria (4), which define T2D as having fasting plasma glucose levels ≥126 mg/dL (7.0 mmol/L) at any examination. Participants who reported fasting for <10 hours or whose fasting information was uncertain were classified as having T2D if they were receiving treatment with either insulin or a hypoglycemic agent; had a nonfasting plasma glucose concentration ≥200 mg/dL (11.1 mmol/L); or had a nonfasting plasma glucose concentration ≥126 mg/dL (7.0 mmol/L) with a diabetes diagnosis at the subsequent Framingham Offspring Study examination. We expected minimal inclusion of type 1 diabetes cases, as baseline diabetes cases were excluded (mean baseline age = 34.4 years) and no participants were <30 years of age at the examination during which diabetes was first diagnosed (mean age = 58 years, with 3 participants <40 years of age) and required subsequent continuous insulin therapy (indicative of type 1 diabetes). Plasma glucose was measured with a hexokinase reagent kit (Agent glucose test, Abbott, South Pasadena, California). Assays were run in duplicate, and the intraassay coefficients of variation ranged from 2% to 3%. For sensitivity analyses, a more conservative T2D classification was defined. It required a diagnosis at a minimum of 2 Framingham Offspring Study examinations or at the last examination attended. This allowed the assessment of potential outcome misclassification that might result from only using 1 plasma glucose measurement to diagnose T2D.

Covariates

Updated data on age, body mass index (BMI), cigarette smoking, and alcohol consumption were used and were represented in the analyses by time-dependent covariates for Framingham Offspring examinations 1 (1971–1975) through 7 (1998–2001). Age and BMI (calculated as measured weight in kilograms divided by height in meters squared) were included as continuous variables. Cigarette smoking was self-reported (classified as never, past, or current smoker) and defined as smoking regularly in the year before the examination. Alcohol consumption was self-reported as the average number of alcoholic drinks (e.g., wine, beer, or cocktails) consumed per week (continuous variable). Components of adult height are associated with development of insulin resistance and T2D in adulthood (26–29). Given that adult height is an established marker of early life factors (30, 31) and other aspects of childhood environment (30, 31) that may not be fully encompassed by father's SEP, we adjusted for baseline height in our analyses. Baseline height and BMI were not strongly correlated in our data set (Pearson's r = 0.19).

Statistical analyses

Age-adjusted means (continuous measures) and proportions (categorical measures) were calculated for baseline covariates and compared across cumulative SEP categories. Multivariable pooled logistic regression adjusted for time-dependent covariates was used to estimate adjusted odds ratios and 95% confidence intervals. In this approach, we created a record for every 4-year interval between examinations for all participants. The binary dependent variable indicated whether a given subject had developed T2D by the end of the interval. Individuals with T2D diagnosed at a previous examination were not included in the population at risk in subsequent examinations. For examinations with known T2D status, any missing covariate data were replaced with data from the closest previously attended examination (32). For age, 4 years was added for each missing examination. If a participant's T2D status was missing and there was no diagnosis of diabetes at a subsequently attended examination, T2D status was coded as no diabetes and covariate data were replaced as described. Of the 11,718 person-examinations included in the analysis, 389 (3%) included updated data. Observations from all examinations were then pooled and analyzed using the generalized estimating equations approach, implemented in PROC GENMOD in SAS (SAS Institute, Inc., Cary, North Carolina) (33). In generalized estimating equations analyses, we specified a binomial distribution and the logit link function, and accounted for the clustering of outcomes by family by assuming compound symmetry covariance structure of residuals. Pooled logistic regression has been shown to be equivalent to using Cox proportional hazards models (34). All analyses were conducted using SAS, version 9.1.

We used multivariable pooled logistic regression analyses to evaluate the associations between cumulative SEP and T2D incidence. In all multivariable analyses, the associations between SEP and T2D were adjusted for age, and subsequent models were adjusted for BMI, smoking, alcohol consumption, and height. No collinearity was detected among covariates when variance inflation factors were used (data not shown). To determine potential sensitive periods, the association between SEP during each life-course period and T2D risk was tested in separate pooled logistic regression analyses. Inclusion of fathers' educational levels, the participants’ educational levels, and the participants’ occupations in the same multivariable model resulted in minimal variance inflation, permitting simultaneous adjustment for all 3 SEP measures. Finally, pooled logistic regression analyses were performed to evaluate the association between social mobility and T2D risk. The interaction between cumulative SEP and sex was tested in the sex-pooled sample by using a 2-df likelihood ratio test, comparing the multivariable models with and without interaction terms between categories of SEP and sex. The interaction between cumulative SEP and sex was statistically significant (P = 0.03); consequently, sex-specific analyses were performed. Sensitivity analyses were completed using a more stringent T2D definition (defined above).

RESULTS

The analyses included 1,893 participants (991 women and 902 men) from the Framingham Offspring Study. Excluded participants were on average older (37.3 vs. 34.4 years of age, P < 0.0001) and more likely to smoke (45.4% vs. 43.2%, _P_ = 0.02), but they did not differ with regard to other covariates or diabetes incidence (all _P_ values > 0.05).

Age-adjusted baseline participant characteristics, stratified by cumulative SEP, are presented in Table 1. In women, cumulative SEP was directly associated with increased alcohol consumption and height and inversely associated with BMI and smoking. In men, cumulative SEP was directly associated with height and inversely associated with BMI, smoking, and alcohol consumption.

Table 1.

Age-Adjusted Baseline Characteristics According to Cumulative Socioeconomic Position Score, Framingham Offspring Study, 1971–2003

In total, 217 incident T2D cases were observed (95 women and 122 men). Results of pooled logistic regression analyses suggested that age-adjusted cumulative SEP was associated with an almost 2-fold increase in T2D risk in women (for low vs. high cumulative SEP, age-adjusted odds ratio (OR) = 1.92, 95% confidence interval (CI): 1.08, 3.42) (Table 2). When we adjusted individually for T2D risk factors, we found that BMI resulted in the greatest decrease in T2D risk. Simultaneous adjustment for all covariates further attenuated the odds ratios (OR = 1.56, 95% CI: 0.85, 2.88). In men, a nonmonotone U-shaped curvilinear association between cumulative SEP and T2D risk was suggested. Specifically, compared with men with a high cumulative SEP, men with a medium cumulative SEP had a lower T2D risk. This association remained statistically significant in the fully adjusted model (OR = 0.53, 95% CI: 0.32, 0.90).

Table 2.

Odds Ratios for Associations Between Cumulative Socioeconomic Position Score and Incidence of Type 2 Diabetes Mellitus in Pooled Logistic Regression Models, Framingham Offspring Study, 1971–2003

To investigate potential periods of sensitivity to low SEP exposure, we performed sex-specific pooled logistic regression analyses of relations between the fathers' educational levels and occupations and the participants’ educational levels and occupations with T2D incidence (Table 3). In women, age-adjusted models showed strong inverse associations of participant educational level and occupation with T2D risk (for ≤12 years vs. >16 years of education, OR = 2.84, 95% CI: 1.03, 7.85; and for laborer vs. professional/executive/supervisory/technical occupation, OR = 2.40, 95% CI: 1.05, 5.47), whereas there was no effect for fathers' educational and T2D (for some high school or less vs. any postsecondary education, age-adjusted OR = 1.26, 95% CI: 0.72, 2.22). These associations were attenuated after adjusting D risk factors and all SEP measures simultaneously. In men, no associations were observed between SEP measures and T2D.

Table 3.

Odds Ratios for the Association Between Socioeconomic Position and Incidence of Type 2 Diabetes in Pooled Logistic Regression Models, Framingham Offspring Study, 1971–2003

For the social-mobility framework, women with a declining SEP (OR = 2.99, 95% CI: 1.39, 6.44) and a stable low SEP (OR = 1.85, 95% CI: 1.02, 3.35) demonstrated increased age-adjusted T2D risk compared with women with a stable high SEP (Table 4). After adjustment for all covariates, increased T2D risk remained strong in the declining SEP category (OR = 2.84, 95% CI: 1.29, 6.25) but not in the stable low SEP category. Similar trends, although with generally weaker associations, were observed when participant occupation was used instead of educational level as the adulthood measure of SEP. In men, no associations were observed between social mobility and T2D risk.

Table 4.

Odds Ratios for the Association Between Social Mobility of Socioeconomic Position and Incidence of Type 2 Diabetes Mellitus in Pooled Logistic Regression Models, Framingham Offspring Study, 1971–2003

Sensitivity analyses in which we used a more conservative measure of T2D incidence (described above) showed generally similar trends, although effect sizes were attenuated (see Web Tables 1–3, available at http://aje.oxfordjournals.org/). Generally similar results were achieved for analyses in which we used fathers' occupation as a measure of childhood SEP instead of fathers' educational level, although effect sizes tended to be somewhat lower (Web Tables 4–6).

DISCUSSION

The present study provides evidence of an inverse association between cumulative SEP and T2D risk in women. Results from sensitive-periods and social-mobility analyses indicated that women who experienced low SEP in young adulthood (i.e., low educational attainment) and active professional life (i.e., low occupation achievement) were at a higher risk of developing T2D than were their higher SEP counterparts, but there was little evidence for an association between low childhood SEP and T2D. Adjustment for T2D risk factors attenuated these associations. Overall, there was no association in men.

Prior literature

Very little is known about associations between cumulative SEP and T2D incidence. We found that higher lifetime exposure to low SEP was related to increased T2D risk in women, which supported the accumulation-of-risk SEP framework. In our sensitive-periods analyses, strong inverse associations were observed between participant education and occupation and T2D risk in women. These findings were consistent with previous prospective studies of SEP and T2D (7, 10, 13, 15). Furthermore, women with decreasing and stable low SEP trajectories from childhood to young adulthood had increased T2D risk compared with women with stable high trajectories, which also supported previous findings (7, 11). Combined, results from the sensitive-periods and social-mobility frameworks suggested that educational level and occupation had the largest impact on T2D incidence in women. Overall, there was little evidence of an association between SEP and T2D incidence in men. Results from previous studies evaluating associations between adult SEP and T2D in men are conflicted; some showed inverse associations (14, 20), and others showed weak or no associations (10, 13, 16–19).

Results from the Nurses’ Health Study (in women, n = 100,330; relative risk = 1.50, 95% CI: 1.32, 1.69 (7)) and the Alameda County Study (n = 5,913; in women, hazard ratio = 1.8, 95% CI: 1.3, 2.6 and in men, hazard ratio = 1.4, 95% CI: 1.0, 2.0 (8)) reported increased T2D risk associated with low versus high childhood SEP in age- and race-adjusted models. The effect sizes in our study showed a similar direction but a nonsignificant association for women (e.g., for fathers' occupation, laborer vs. professional/executive/supervisory/technical, OR = 1.25, 95% CI: 0.80, 1.93).

Potential mechanisms

Childhood SEP is inversely associated with birth weight (35) and smoking (36–38) and (in women) obesity (39) in adulthood. Adult SEP has been shown to be inversely associated with obesity (in women (37, 38, 40)), smoking (41), leisure-time physical activity (42–44), and unhealthy diet (45, 46). Chains of risk are also likely at play, resulting from transgenerational effects, because childhood SEP is predictive of adult SEP (47) and risky health behaviors can be modeled as normative to offspring (48). In our study, adjusting for obesity resulted in the greatest reduction in the association between cumulative SEP and T2D in women. Obesity could be an important mechanism that mediates the sex-specific association between life-course SEP and T2D. Obesity, a significant T2D risk factor (49–51) that is thought to raise risk through increasing insulin resistance (52), is a condition that is also socially patterned (53). Childhood obesity has been shown to predict lower adult SEP in women (54). An important mechanism could be obesity-related discrimination, which limits upward social mobility, particularly in white women (55). This mechanism might explain some of the sex-specific association between SEP and T2D that was observed in women but not in men. It should be noted that adjusting for mediators on the causal pathway is known to induce bias toward the null (56, 57). Thus, to the extent to which obesity is considered a potential mediator of the impact of low SEP, our results that adjusted for BMI may be overly conservative.

Strengths and limitations

There were several limitations to this study. The study population largely included individuals of European descent (representative of the general population of Framingham, Massachusetts, at the study onset). The generalizability of findings to other races and ethnicities and the populations of other countries, some of which have a high prevalence of T2D, is limited (22). Additionally, the study population represented a highly selected sample of Framingham Offspring Study participants (n = 1,893 of 5,124). However, the largest exclusion criterion was not having a father in the original cohort of the Framingham Heart Study (n = 2,136, 66% of excluded participants). This exclusion criterion allowed for a direct and more robust measure of childhood SEP. A limitation of the accumulation-of-risk model, as described in this study, is that it assumes that SEP exposure during each period of the life course has an equal effect on T2D risk. This approach simplifies a complex pathway that links cumulative SEP and T2D, and it alone cannot provide information on the association between SEP and T2D risk during specific periods in the life course. Therefore, to complement this model, we investigated sensitive-periods and social-mobility frameworks to help triangulate the effect of SEP at different stages of the life course and its corresponding risk for development of T2D. Finally, data on family history of diabetes and childhood obesity were not available in the limited access data set for the Framingham Offspring Study.

Strengths of the study include the direct assessment of childhood SEP from participants’ fathers. Furthermore, we investigated the association between life-course SEP and T2D in a sample with clinically determined T2D, limiting misclassification from the undiagnosed and self-reported T2D that is commonly used in other studies.

CONCLUSION

In summary, the present study found that cumulative SEP was inversely associated with T2D risk in female participants of the Framingham Offspring Study. The cumulative association between SEP and T2D risk appeared to be driven particularly by the participants’ educational levels and occupations. There was little evidence of an association in men.

Supplementary Material

Acknowledgments

Author affiliations: Division of Epidemiology, Dalla Lana School of Public Health, University of Toronto, Toronto, Canada (Brendan T. Smith); Institute for Work and Health, Toronto, Canada (Brendan T. Smith); Sansom Institute, School of Health Sciences, University of South Australia, Adelaide, Australia (John W. Lynch); Department of Social Medicine, University of Bristol, Bristol, United Kingdom (John W. Lynch); Boston University and the National Heart, Lung, and Blood Institute, Framingham, Massachusetts (Caroline S. Fox); Department of Endocrinology, Hypertension, and Diabetes, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts (Caroline S. Fox); National Heart, Lung, and Blood Institute, Bethesda, Maryland (Caroline S. Fox); Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, Canada (Sam Harper, Michal Abrahamowicz, Nisha D. Almeida); Division of Clinical Epidemiology, McGill University Health Centre, Royal Victoria Hospital, Montreal, Canada (Michal Abrahamowicz); and Department of Community Health, Epidemiology Section, Center for Population Health and Clinical Epidemiology, Brown University, Providence, Rhode Island (Eric B. Loucks).

The work was supported by Canadian Institutes of Health Research operating grant MOP81239; Canadian Institutes of Health Research Interdisciplinary Capacity Enhancement grant HOA80072; and a Canadian Institutes of Health Research New Investigator Award (to E.B.L.). The Framingham Heart Study is supported by National Institutes of Health/National Heart, Lung, and Blood Institute contract N01-HC-25195.

The authors are grateful to the National Heart, Lung, and Blood Institute for providing access to the limited-access database and to the Framingham Heart Study investigators, whose work provided data on the variables used in the analyses.

Conflict of interest: none declared.

Glossary

Abbreviations

References

1. Centers for Disease Control and Prevention. Number (in Millions) of Civilian, Non-Institutionalized Persons With Diagnosed Diabetes, United States, 1980–2008. Atlanta, GA: Centers for Disease Control and Prevention; 2010. ( http://www.cdc.gov/diabetes/statistics/prev/national/figpersons.htm). (Accessed September 7, 2009) [Google Scholar]

2. American Diabetes Association. Economic costs of diabetes in the U.S. in 2007. Diabetes Care. 2008;31(3):596–615. [PubMed] [Google Scholar]

3. Narayan KM, Boyle JP, Geiss LS, et al. Impact of recent increase in incidence on future diabetes burden: U.S., 2005–2050. Diabetes Care. 2006;29(9):2114–2116. [PubMed] [Google Scholar]

4. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2004;27(suppl 1):S5–S10. [PubMed] [Google Scholar]

5. Brown AF, Ettner SL, Piette J, et al. Socioeconomic position and health among persons with diabetes mellitus: a conceptual framework and review of the literature. Epidemiol Rev. 2004;26:63–77. [PubMed] [Google Scholar]

6. Kuh D, Ben-Shlomo Y, Lynch J, et al. Life course epidemiology. J Epidemiol Community Health. 2003;57(10):778–783. [PMC free article] [PubMed] [Google Scholar]

7. Lidfeldt J, Li TY, Hu FB, et al. A prospective study of childhood and adult socioeconomic status and incidence of type 2 diabetes in women. Am J Epidemiol. 2007;165(8):882–889. [PubMed] [Google Scholar]

8. Maty SC, Lynch JW, Raghunathan TE, et al. Childhood socioeconomic position, gender, adult body mass index, and incidence of type 2 diabetes mellitus over 34 years in the Alameda County Study. Am J Public Health. 2008;98(8):1486–1494. [PMC free article] [PubMed] [Google Scholar]

9. Agardh EE, Ahlbom A, Andersson T, et al. Socio-economic position at three points in life in association with type 2 diabetes and impaired glucose tolerance in middle-aged Swedish men and women. Int J Epidemiol. 2007;36(1):84–92. [PubMed] [Google Scholar]

10. Best LE, Hayward MD, Hidajat MM. Life course pathways to adult-onset diabetes. Soc Biol. 2005;52(3-4):94–111. [PubMed] [Google Scholar]

11. Regidor E, Banegas JR, Gutiérrez-Fisac JL, et al. Socioeconomic position in childhood and cardiovascular risk factors in older Spanish people. Int J Epidemiol. 2004;33(4):723–730. [PubMed] [Google Scholar]

12. Maty SC, Everson-Rose SA, Haan MN, et al. Education, income, occupation, and the 34-year incidence (1965–99) of type 2 diabetes in the Alameda County Study. Int J Epidemiol. 2005;34(6):1274–1281. [PMC free article] [PubMed] [Google Scholar]

13. Robbins JM, Vaccarino V, Zhang H, et al. Socioeconomic status and diagnosed diabetes incidence. Diabetes Res Clin Pract. 2005;68(3):230–236. [PubMed] [Google Scholar]

14. Kumari M, Head J, Marmot M. Prospective study of social and other risk factors for incidence of type 2 diabetes in the Whitehall II study. Arch Intern Med. 2004;164(17):1873–1880. [PubMed] [Google Scholar]

15. Krishnan S, Cozier YC, Rosenberg L, et al. Socioeconomic status and incidence of type 2 diabetes: results from the Black Women's Health Study. Am J Epidemiol. 2010;171(5):564–570. [PMC free article] [PubMed] [Google Scholar]

16. Rathmann W, Haastert B, Icks A, et al. Sex differences in the associations of socioeconomic status with undiagnosed diabetes mellitus and impaired glucose tolerance in the elderly population: the KORA Survey 2000. Eur J Public Health. 2005;15(6):627–633. [PubMed] [Google Scholar]

17. Robbins JM, Vaccarino V, Zhang H, et al. Socioeconomic status and type 2 diabetes in African American and non-Hispanic white women and men: evidence from the third National Health and Nutrition Examination Survey. Am J Public Health. 2001;91(1):76–83. [PMC free article] [PubMed] [Google Scholar]

18. Tang M, Chen Y, Krewski D. Gender-related differences in the association between socioeconomic status and self-reported diabetes. Int J Epidemiol. 2003;32(3):381–385. [PubMed] [Google Scholar]

19. Hazuda HP, Haffner SM, Stern MP, et al. Effects of acculturation and socioeconomic status on obesity and diabetes in Mexican Americans. The San Antonio Heart Study. Am J Epidemiol. 1988;128(6):1289–1301. [PubMed] [Google Scholar]

20. Agardh EE, Ahlbom A, Andersson T, et al. Explanations of socioeconomic differences in excess risk of type 2 diabetes in Swedish men and women. Diabetes Care. 2004;27(3):716–721. [PubMed] [Google Scholar]

21. Kannel WB, Feinleib M, McNamara PM, et al. An investigation of coronary heart disease in families. The Framingham Offspring Study. Am J Epidemiol. 1979;110(3):281–290. [PubMed] [Google Scholar]

22. Galobardes B, Lynch JW, Davey Smith G. Childhood socioeconomic circumstances and cause-specific mortality in adulthood: systematic review and interpretation. Epidemiol Rev. 2004;26:7–21. [PubMed] [Google Scholar]

23. Galobardes B, Shaw M, Lawlor DA, et al. Indicators of socioeconomic position (part 1) J Epidemiol Community Health. 2006;60(1):7–12. [PMC free article] [PubMed] [Google Scholar]

24. Loucks EB, Pilote L, Lynch JW, et al. Life course socioeconomic position is associated with inflammatory markers: the Framingham Offspring Study. Soc Sci Med. 2010;71(1):187–195. [PMC free article] [PubMed] [Google Scholar]

25. Loucks EB, Lynch JW, Pilote L, et al. Life-course socioeconomic position and incidence of coronary heart disease: the Framingham Offspring Study. Am J Epidemiol. 2009;169(7):829–836. [PMC free article] [PubMed] [Google Scholar]

26. Asao K, Kao WH, Baptiste-Roberts K, et al. Short stature and the risk of adiposity, insulin resistance, and type 2 diabetes in middle age: the Third National Health and Nutrition Examination Survey (NHANES III), 1988–1994. Diabetes Care. 2006;29(7):1632–1637. [PubMed] [Google Scholar]

27. Lawlor DA, Ebrahim S, Davey Smith G. The association between components of adult height and type II diabetes and insulin resistance: British Women's Heart and Health Study. Diabetologia. 2002;45(8):1097–1106. [PubMed] [Google Scholar]

28. Smith GD, Greenwood R, Gunnell D, et al. Leg length, insulin resistance, and coronary heart disease risk: the Caerphilly Study. J Epidemiol Community Health. 2001;55(12):867–872. [PMC free article] [PubMed] [Google Scholar]

29. Njølstad I, Arnesen E, Lund-Larsen PG. Sex differences in risk factors for clinical diabetes mellitus in a general population: a 12-year follow-up of the Finnmark Study. Am J Epidemiol. 1998;147(1):49–58. [PubMed] [Google Scholar]

30. Li L, Manor O, Power C. Early environment and child-to-adult growth trajectories in the 1958 British birth cohort. Am J Clin Nutr. 2004;80(1):185–192. [PubMed] [Google Scholar]

31. Wadsworth ME, Hardy RJ, Paul AA, et al. Leg and trunk length at 43 years in relation to childhood health, diet and family circumstances; evidence from the 1946 national birth cohort. Int J Epidemiol. 2002;31(2):383–390. [PubMed] [Google Scholar]

32. Stringhini S, Sabia S, Shipley M, et al. Association of socioeconomic position with health behaviors and mortality. JAMA. 2010;303(12):1159–1166. [PMC free article] [PubMed] [Google Scholar]

33. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics. 1986;42(1):121–130. [PubMed] [Google Scholar]

34. Cupples LA, D'Agostino RB, Anderson K, et al. Comparison of baseline and repeated measure covariate techniques in the Framingham Heart Study. Stat Med. 1988;7(1-2):205–222. [PubMed] [Google Scholar]

35. Valero De Bernabé J, Soriano T, Albaladejo R, et al. Risk factors for low birth weight: a review. Eur J Obstet Gynecol Reprod Biol. 2004;116(1):3–15. [PubMed] [Google Scholar]

36. Jefferis BJ, Power C, Graham H, et al. Effects of childhood socioeconomic circumstances on persistent smoking. Am J Public Health. 2004;94(2):279–285. [PMC free article] [PubMed] [Google Scholar]

37. Lawlor DA, Batty GD, Morton SM, et al. Childhood socioeconomic position, educational attainment, and adult cardiovascular risk factors: the Aberdeen children of the 1950s cohort study. Am J Public Health. 2005;95(7):1245–1251. [PMC free article] [PubMed] [Google Scholar]

38. Power C, Graham H, Due P, et al. The contribution of childhood and adult socioeconomic position to adult obesity and smoking behaviour: an international comparison. Int J Epidemiol. 2005;34(2):335–344. [PubMed] [Google Scholar]

39. Senese LC, Almeida ND, Fath AK, et al. Associations between childhood socioeconomic position and adulthood obesity. Epidemiol Rev. 2009;31:21–51. [PMC free article] [PubMed] [Google Scholar]

40. McLaren L. Socioeconomic status and obesity. Epidemiol Rev. 2007;29:29–48. [PubMed] [Google Scholar]

41. Gilman SE, Martin LT, Abrams DB, et al. Educational attainment and cigarette smoking: a causal association? Int J Epidemiol. 2008;37(3):615–624. [PMC free article] [PubMed] [Google Scholar]

42. Li F, Fisher KJ, Bauman A, et al. Neighborhood influences on physical activity in middle-aged and older adults: a multilevel perspective. J Aging Phys Act. 2005;13(1):87–114. [PubMed] [Google Scholar]

43. Kuh DJ, Cooper C. Physical activity at 36 years: patterns and childhood predictors in a longitudinal study. J Epidemiol Community Health. 1992;46(2):114–119. [PMC free article] [PubMed] [Google Scholar]

44. Pekkanen J, Tuomilehto J, Uutela A, et al. Social class, health behaviour, and mortality among men and women in eastern Finland. BMJ. 1995;311(7005):589–593. [PMC free article] [PubMed] [Google Scholar]

45. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr. 2004;79(1):6–16. [PubMed] [Google Scholar]

46. Turrell G, Hewitt B, Patterson C, et al. Socioeconomic differences in food purchasing behaviour and suggested implications for diet-related health promotion. J Hum Nutr Diet. 2002;15(5):355–364. [PubMed] [Google Scholar]

47. Beller E, Hout M. Intergenerational social mobility: the United States in comparative perspective. Future Child. 2006;16(2):19–36. [PubMed] [Google Scholar]

48. Hanson MD, Chen E. Socioeconomic status and health behaviors in adolescence: a review of the literature. J Behav Med. 2007;30(3):263–285. [PubMed] [Google Scholar]

49. Willi C, Bodenmann P, Ghali WA, et al. Active smoking and the risk of type 2 diabetes: a systematic review and meta-analysis. JAMA. 2007;298(22):2654–2664. [PubMed] [Google Scholar]

50. Hart CL, Hole DJ, Lawlor DA, et al. How many cases of type 2 diabetes mellitus are due to being overweight in middle age? Evidence from the Midspan prospective cohort studies using mention of diabetes mellitus on hospital discharge or death records. Diabet Med. 2007;24(1):73–80. [PubMed] [Google Scholar]

51. Narayan KM, Boyle JP, Thompson TJ, et al. Effect of BMI on lifetime risk for diabetes in the U.S. Diabetes Care. 2007;30(6):1562–1566. [PubMed] [Google Scholar]

52. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes. 1988;37(12):1595–1607. [PubMed] [Google Scholar]

53. Lawlor DA, Davey Smith G, Ebrahim S. Life course influences on insulin resistance: findings from the British Women's Heart and Health Study. Diabetes Care. 2003;26(1):97–103. [PubMed] [Google Scholar]

54. Conley D, Glauber R. Gender, body mass, and socioeconomic status: new evidence from the PSID. Adv Health Econ Health Serv Res. 2007;17:253–275. [PubMed] [Google Scholar]

55. Puhl RM, Heuer CA. The stigma of obesity: a review and update. Obesity (Silver Spring) 2009;17(5):941–964. [PubMed] [Google Scholar]

56. Cole SR, Hernán MA. Fallibility in estimating direct effects. Int J Epidemiol. 2002;31(1):163–165. [PubMed] [Google Scholar]

57. Kaufman JS, Maclehose RF, Kaufman S. A further critique of the analytic strategy of adjusting for covariates to identify biologic mediation. Epidemiol Perspect Innov. 2004;1(1):4. (doi: 10.1186/1742-5573-1-4) [PMC free article] [PubMed] [Google Scholar]

Articles from American Journal of Epidemiology are provided here courtesy of Oxford University Press