More schooling is associated with lower Hemoglobin A1c at the high-risk tail of the distribution: An unconditional quantile regression analysis

More schooling is associated with lower Hemoglobin A1c at the high-risk tail of the distribution: An unconditional quantile regression analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article More schooling is associated with lower Hemoglobin A1c at the high-risk tail of the distribution: An unconditional quantile regression analysis Jillian Hebert, Amanda Irish, Aayush Khadka, Abigail Arons, Alicia R. Riley, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5462691/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 03 Jun, 2025 Read the published version in BMC Public Health → Version 1 posted 9 You are reading this latest preprint version Abstract Background: Risk of diabetes increases exponentially with higher levels of glycosylated hemoglobin (HbA1c). Education is inversely associated with average HbA1c, however, differential associations between education and HbA1c across the HbA1c distribution have not been evaluated. Methods: Health and Retirement Study data (N=21,732) was used to evaluate the association between education (linear terms among those with <12 years and ≥ 12 years of education) and first recorded HbA1c (2003-2016) at the mean using linear regression, and at the 1st-99th quantiles of the marginal outcome distribution using unconditional quantile regressions, controlling for birth year, race and ethnicity, gender, birthplace, parent’s education, and year of HbA1c measurement. Results: Mean HbA1c was 5.9%; 16.6% of participants had HbA1c above the diabetes diagnostic threshold of 6.5%. For those with less than 12 years of schooling, there was no association between education and HbA1c at the mean or across the quantiles. For those with 12 or more years of schooling, an additional year of education was negatively associated with mean HbA1c (b OLS =-0.02, 95% confidence interval (CI) -0.03,-0.02); a one-year increase in mean education was associated with lower HbA1c across the distribution, but the magnitude was larger at higher quantiles (b q50 =-0.02, 95%CI -0.02,-0.01; b q90 =-0.06, 95%CI -0.09,-0.04). Conclusions: Educational attainment is inversely associated with HbA1camong those with 12 or more years of schooling, with larger point estimates for those in the high-risk tail of the HbA1c distribution. Unconditional quantile regression Distributional effects Effect heterogeneity Diabetes US Health and Retirement Study (HRS) Figures Figure 1 Figure 2 1. Introduction Prior work, both in the US and globally, reveals a strong inverse relationship between educational attainment and average glycosylated hemoglobin (HbA1c), an index of glucose regulation over the past 2-3 months, such that those with less education are at higher risk of diabetes and diabetes-related complications. 1-6 Among people with diabetes, glycemic control is essential for reducing complications, such as kidney failure, cardiovascular events, and all-cause mortality; 7-11 in parallel, staying within normal ranges of HbA1c is important to prevent diabetes onset and related complications. Risk of diabetes complications increase exponentially with HbA1c, such that a percentage-point increase at a higher level of HbA1c (e.g. 6% to 7%) confers a much greater risk for diabetes-related complications compared to a percentage-point increase at lower HbA1c levels (e.g. 4% to 5%). Therefore, exposures and interventions that have larger impacts for those with higher HbA1c may be important for preventing diabetes-related complications. Education, a main component of socioeconomic status, is consistently associated with better health over the lifecourse. 12-13 Increased educational attainment may impact HbA1c through several pathways, including longer life expectancy, increased income, better access to healthcare, and more health promoting behaviors (e.g., increased physical activity). 1,4,14-16 While evidence suggests a strong protective association between educational attainment and mean HbA1c, no study evaluates the association across the entire HbA1c distribution. Evaluating if the relationship is constant across the HbA1c distribution is important, given the exponential relationship between HbA1c and diabetes onset and related complications. Interventions that specifically impact high levels of HbA1c are of interest given their potential to reduce diabetes onset and related complications. We hypothesize education will have larger associations among those with higher HbA1c as participants belonging to more structurally minoritized subgroups (e.g., minoritized due to race, or poverty status) will be over-represented at the high-risk end of the HbA1c distribution (i.e., higher quantiles of the HbA1c distribution), 17-19 and these same groups also seem to benefit more from education. 45-47 We add to existing literature by evaluating the relationship between education and HbA1c across the HbA1c distribution through a novel application of quantile regression, a modeling technique that evaluates the exposure-outcome relationship across the outcome distribution. Quantile regressions can identify if educational attainment has a heterogeneous effect across the HbA1c distribution, which allows for discovery of whether, and which parts of, the HbA1c’s distribution are differentially impacted by educational attainment. In this way, our application of quantile regression to the education-HbA1c relationship has the potential to deepen our understanding of educational inequities in diabetes by uncovering details hidden by linear regression. This paper empirically evaluates the relationship between educational attainment and late-life HbA1c using linear regression and unconditional quantile regressions (UQR) for US Health and Retirement Study participants ages 50 and older, examining whether the relationship between education and HbA1c varies across the HbA1c distribution. We stratify education at 12 years, where 12 years of schooling typically confers a high school diploma, since prior literature finds evidence for divergent health trajectories of adults with less than a high school education from those with high school or more. 20 2. Methods Data and Analytic Sample Data came from the U.S. Health and Retirement Study (HRS), a national longitudinal sample of non-institutionalized adults 50 years and older, and their spouses of any age, that began in 1992. 21 New cohorts of participants have been added every six years after 1998 to maintain a steady state population, and participants are surveyed biennially. A diabetes sub study collected HbA1c for a subset of HRS participants in 2003; in addition, HbA1c and other biomarker data was collected in 2006 for a randomly selected half of the sample, and in 2008 for the other half; biomarker data were subsequently collected every four years. The eligible sample included all HRS participants with at least one HbA1c measurement between 2003 and 2006 when they were 50 years or older (N = 21,840). Individuals were excluded for missing exposure (N = 89) and covariate data (N = 18); one participant was removed due to their HbA1c measurement being recorded prior to their first HRS interview, resulting in an analytic sample of 21,732 participants (99% of eligible). Exposure Our exposure, educational attainment, was created using self-reported total years of schooling. Education in HRS ranges from 0 to 17 years of schooling, where 17 years includes those with 17 or more years of education (17 or more years: N = 2,327). Due to data sparseness of participants with fewer than 5 years of education, we coded those with less than 5 years of education to 5 years to reduce the impact these outliers may have on estimates (N = 776). To assess possible heterogeneities between participants with different types of credentials, and since educational policies tend to target specific levels of education (e.g. compulsory schooling laws and child labor laws targeted K-12 education, while other policies only addressed college education), education was stratified into two levels: fewer than 12 years of education (N = 4,801; 22%) and 12 or more years of education (N = 16,931; 78%), where completing 12 years of education typically corresponds to earning a high school diploma. Education was modeled linearly within these two educational strata. Outcome Our outcome was the participant’s first recorded HbA1c value (2003-2016) measured at or after age 50. HbA1c is glycosylated hemoglobin and reflects blood glucose over the prior 2-3 months; HbA1c values between 5.71% and 6.49% are consistent with pre-diabetes and values greater than 6.5% are consistent with diabetes. 22-24 HbA1c was measured using an automated ion-exchange high-performance liquid chromatography that recorded the percentage of glycosylated hemoglobin in dried blood spot samples. 21 Covariates All models were adjusted for sex (female; male), race (Non-Hispanic White; Non-Hispanic Black; Latinx/Hispanic; other), birthplace (non-Southern US; Southern US; Foreign), indicator for birth year (1905-1966), indicator for year of HbA1c measurement (2003-2016), mother’s education (5-17(+) years, linear), father’s education (5-17(+) years, linear), as well as missing indicators for mother’s education and father’s education. Sex was included as an indicator of the socially stratifying effects of gender, 25 race as an indicator of the socially stratifying effects of systemic racism, 26-27 and parent’s education as a proxy for childhood socioeconomic status. Chi-square tests were used to evaluate if there were significant differences in categorical covariates by education level (e.g., less than 12 years of education versus 12 or more years of education). See supplemental Table S1 for additional details on covariates. Race was categorized to include an “other” category in all models for precision, but results were not reported for this group due to the heterogeneous composition and consequent lack of interpretability of estimates. Birthplace was classified by location within the US (i.e., non-Southern vs Southern) because studies have found increased risk for adverse later-life health outcomes for those born in the Southern US. 28-31 A subset of participants (N = 414, 2%) were known to be born in the US, but were missing information on the region of birth; participants where the region of birth was unknown were assumed to be born in the Non-Southern US. Birth year was modeled as an indicator variable to capture differences by individual year. Due to a small number of participants falling in the tail ends of the birth year range, values were recoded to facilitate model convergence: those born before 1917 (N = 240, 1%) were recoded as 1917; those born in 1966 (N = 52, 0.2%) were recoded as 1965. Parental education variables (5-17(+) years) were used as a proxy for family socioeconomic status (SES) and modeled continuously. However, HRS participants in the Asset and Health Dynamics Among the Oldest Old (AHEAD) cohort (born 1900-1923) recorded parent’s education as a dichotomized measure (less than 8 years of education, 8 or more years of education) rather than continuous. Dichotomized measures of parent’s education were replaced with continuous values from a previously validated imputation method using measures of childhood socioeconomic status. 32 Additional missingness in parent’s education was imputed using the sample mean (mother’s education: N Missing = 2,101 (10%), sample mean = 10 (SD 3.9); father’s education: N Missing = 3,644 (17%), sample mean = 10 (SD 4.2)) and a missing indicator was added for proper model adjustment. This allowed for retention of participants with missing parental education where missingness is informative (e.g., if the parent was not in the household). 33 Statistical Analysis We used linear regressions and unconditional quantile regressions (UQR) to model the relationship between education and HbA1c. 34-36 UQR evaluates changes in quantiles of the outcome’s unconditional distribution for a one-unit change in the mean of the exposure. We estimated parameters of the linear regression model using ordinary least squares (OLS). We fit UQR models at each unit quantile between the 1st-99th quantiles of the unconditional HbA1c distribution. We used bootstrapping (500 repetitions) to estimate 95% confidence intervals (CIs) for parameters of the linear regression and UQR models; education was modeled as a linear term within both education strata and all models were adjusted for the covariates specified in the preceding section. To visualize the change in the sample distribution of HbA1c implied by UQR results, we created plots to show the counterfactual HbA1c distribution for a one-year increase in the sample’s mean education. First, we binned the factual, or observed, data into quantiles (1st-99th). We then added the UQR estimation of the association between education and HbA1c to the observed data by quantile, creating a potential counterfactual distribution. Finally, we plotted the observed and counterfactual distributions. Further details about constructing these datasets and plots are provided elsewhere. 36 Sensitivity Analysis We conducted three additional analyses to see if results were robust to different analytic decisions. First, to determine if results were sensitive to how the exposure was operationalized (i.e., coded for analysis), we re-coded education as a three category variable (<12 years, 12-15 years, and 16 or more years of education) rather than two. Second, given that the HRS collects data from numerous birth cohorts (from 1905-1966) and given that educational attainment has tended to increase over time, we conducted analyses stratified by HRS entry cohort (which are defined by the HRS using participant’s birth year), to test the sensitivity of the association to secular trends in education. Finally, while medication could be an important contributing factor in the education-HbA1c relationship, medication usage is downstream from education and is therefore a potential mediator of the education-HbA1c relationship. Adjusting for mediators can bias estimates, 37 so the main analyses did not include adjustment for medication. Software and Code All data cleaning and analysis was performed in R. 38 We used the dineq package for fitting UQRs. All code was reviewed by the second author as recommended practice 39 and can be found on GitHub. 3. Results Participants included in analysis (Table 1) had an average of 13 years of education, were predominantly women (57%), White (65%), and born in the non-Southern US (53%). Most participants self-reported having 12 or more years of education (78%) and median HbA1c (Interquartile Range - IQR) in the overall sample was 5.7% (5.3%-6.2%). Compared to participants with 12 or more years of education (henceforth 12+ years), those with less than 12 years of education (henceforth < 12 years) were older (mean age: 67 vs. 65), had higher proportions of Black and Latinx participants (23% vs. 17% Black; 31% vs. 8% Latinx), had a higher proportion of Southern US birthplace (42% vs. 31%), and had lower parental education (mean mother’s education: 8 vs. 11; mean father’s education: 8 vs. 10) with higher rates of missingness (20% vs. 7% missing mothers education; 30% vs. 13% missing father education). Median HbA1c (IQR) for < 12 years was 5.8% (5.5%-6.4%) and 5.6% (5.3%-6.1%) for 12+ years. Chi-square tests found significant differences in race and ethnicity, birthplace, proportion missing parent’s education, and proportion of the sample with HbA1c levels that are consistent with being pre-diebetic or having diabetes. Table 1 Figure 1 displays linear regression and UQR results for both education strata (<12 years; 12+ years). Linear regression results indicate that each additional year of education was not associated with average HbA1c for participants with <12 years (Figure 1, panel a: -0.00, 95% CI: -0.02, 0.02). UQR results for the 1st-99th quantiles suggest that a one-year increase in mean education was not associated with HbA1c across most quantiles, but may have been associated with higher HbA1c at higher quantiles (91-97th quantiles), although the confidence intervals here were wide (Figure 1, panel a: b q5 = -0.02%; 95% CI: -0.03, 0.03, b q50 = 0.00; 95% CI: -0.01, 0.01, b q95 = 0.13%; 95% CI: -0.03, 0.29). For participants with 12+ years, linear regression results indicate that each additional year of education was associated with lower average HbA1c (Figure 1, panel b: -0.02, 95% CI: -0.03, -0.02). UQR results suggest that a one-year increase in mean education was associated with lower HbA1c across almost all quantiles with a larger magnitude at higher quantiles (90-99th quantile) (Figure 1, panel b; b q5 = -0.01%; 95% CI: -0.02, -0.00, b q50 = -0.02%; 95% CI: -0.02, -0.01, b q95 = -0.09%; 95% CI: -0.14, -0.04). Figure 1 Figure 2 illustrates the factual, or observed, distribution of HbA1c in the sample and the predicted reshaping (i.e., counterfactual) of the HbA1c distribution based on the UQR estimates. Figure 2 helps to visualize the potential impact of increased education on the HbA1c distribution (i.e., the counterfactual distribution). Consistent with the magnitude of the UQR estimates in Figure 1, changes in the counterfactual HbA1c distribution are small and therefore difficult to discern; to aid in visualization, Figure 2 zooms in on the peak of the HbA1c factual and counterfactual distributions. A figure including the full HbA1c distribution is included in the appendix (eFigure 1). For those with <12 years, the density at the peak of the counterfactual distribution is slightly lower than the density at the peak of the factual distribution, suggesting a small rightward shift away from the densest point of the distribution, towards higher HbA1c values. For 12+ years, the density at the peak of the counterfactual distribution is slightly higher than the density at the peak of the factual distribution, suggesting a small leftward shift in the direction of the densest part of the distribution, towards lower HbA1c values. Figure 2 Sensitivity analyses Results were attenuated but substantively similar and conclusions were unchanged when stratifying into 3 education strata (<12 years of education, 12-15 years of education, 16+ years of education) (eFigure 2). Results for 12-15 years of education were most similar to results for the 12+ years of education in the main analysis; results for 16+ years of education were null across most quantiles but were associated with lower HbA1c at higher quantiles (90-98th quantiles). When evaluating differences by entry cohorts, most results included the null, but estimates were largely in the same direction (eFigures 3-9). The War Babies (birth years: 1942-1947) and Mid Baby Boomer (birth years: 1954-1959) cohort’s results were most dissimilar to the main results. Results and conclusions were unchanged after additional adjustments for medication (eFigure 10). 4. Discussion In the HRS sample of older US adults, we found that the relationship between educational attainment and HbA1c varies by education level and is heterogeneous across the HbA1c distribution. Overall, associations between education and HbA1c seem to be strongest at the high-risk tail of the HbA1c distribution (i.e., participants with the highest HbA1c levels; 90-99th quantiles). That is, participants with the highest HbA1c levels who are most at risk of developing diabetes or experiencing diabetes-related complications such as cardiovascular events and all-cause mortality had the strongest association with educational attainment. We also found that these associations varied by total years of schooling: among those with less than 12 years of education, a one-year increase in mean education was not associated with HbA1c for participants with low to average HbA1c levels, however, it was weakly associated with higher HbA1c for participants with high HbA1c levels. Conversely, among those with 12 or more years of education, a one-year increase in mean education was associated with lower HbA1c across the distribution, with larger magnitudes in the association at the high-risk tail of the distribution; more education was associated with a leftward shift in the HbA1c distribution, suggesting lower risk of diabetes and diabetes-related complications in the sample population. While the associations appear minimal, our work adds to existing literature by demonstrating a method of evaluating heterogeneous relationships and identifying that the relationship between education and HbA1c differs by education level. There were two analytic decisions that allowed us to identify these novel findings. First, we stratified our exposure at 12 years of education. This stratification was informed by prior research highlighting the divergent health trajectories of adults with less than a high school education from those with high school or more. 20 Second, we used UQRs to evaluate the relationship across the entire HbA1c distribution, as opposed to evaluating the relationship at the mean, where one assumes the change at the mean is constant across the entire HbA1c distribution. Among older adults with more than 12 years of education, estimates at the mean were associated with lower HbA1c, as were estimates for most of the quantiles in the HbA1c distribution; at the highest quantiles of the HbA1c distribution, magnitudes of the estimates were larger, suggesting that there may be a larger association for those with higher levels of HbA1c. This discovery additionally exemplifies that the mean association was not constant across the entire HbA1c distribution. Our findings that increased educational attainment was associated with lower HbA1c among those with 12 or more years of education is consistent with prior literature showing that lower educational attainment is associated with increased risk of diabetes and more broadly that detrimental social factors strongly affect risk of diabetes. 1-17 Our use of quantile regressions revealed a larger magnitude of the education-HbA1c association at the high-risk tail of the HbA1c distribution, where individuals are more likely to have diabetes. We hypothesize two potential mechanisms for this: First, the impact of structural factors (race, poverty, historic systems of marginalization and exclusion) on educational attainment and HbA1c. Evidence shows that education has a larger impact among low childhood socioeconomic status or racial and ethnic minoritized groups, when compared to structurally advantaged groups (e.g., White). 41-44, 47 Second, is that education impacts skill-level, which, in turn, impacts the type of job one attains and their income. Therefore, individuals with diabetes who have higher levels of education likely also have greater access to resources, such as money, health care and medications, or social and behavioral resources that support lifestyle changes to better manage glucose levels . 40, 45-46 Our findings about the relationship between HbA1c and education are relevant for both clinicians and public health policymakers. At the clinical level, this association underscores the role of social determinants of health screening as a part of diabetes risk assessment, and future research should explore the added benefit of more precise screening for education in terms of number of years rather than more commonly used broad categories of educational attainment. Such screening and subsequent linkage to social services has the potential to support clinicians in better identifying and supporting patients to lower their risk of diabetes (e.g., pursuing earlier or more intensive preventive intervention or more aggressive glycemic control once diagnosed with diabetes). Additional studies could explore the effectiveness of interventions for these patients that seek to mitigate the detrimental effect of less education on HbA1c, such as disease education programs, navigational supports, and pharmacy management tools. Finally, our results raise the question of whether increasing educational attainment, even by 1 year, for populations would be an effective intervention to reduce future glucose levels. Future secondary data analyses and intervention studies can further elucidate whether individual-level or population-level interventions that increase post-secondary educational attainment may be an effective strategy for lowering diabetes risk. It is important to contextualize why results for those with less than 12 years of education are different from those with 12 or more years of education, especially because the results for those with less than 12 years of education are contrary to our hypothesis and prior literature. It could be that those with less than 12 years of education are a different, more structurally minoritized group than those with more education, resulting in different relationships between education and HbA1c; we found significant differences in race and ethnicity, birthplace, proportion missing parent’s education, and proportion of the sample with HbA1c levels that are consistent with being pre-diebetic or having diabetes. Our sample of those with less than 12 years of education were more likely to be people of color, born outside of the US or in the Southern US, and their parents had less education or more missing data on education – all potential indicators of increased marginalization. Results could also be due to standard variability in the unconditional quantile regression estimates at the tails of the distribution, where the density is low, and the variance of the RIF is larger. 34 An important strength of this analysis is the additional information provided by the UQR modeling technique. By looking at the exposure-outcome relationship across the entire outcome distribution, estimates can identify heterogeneities in the education and HbA1c relationship. OLS results only suggest lower HbA1c levels given an additional year of education for participants with 12 or more years of education. UQR results further suggest that the magnitude of change in HbA1c is larger for those at the high-risk tail of the distribution. Comparing the OLS and UQR results highlights the limitation of mean models in capturing an exposure’s relationship with the outcome distribution, and the necessity for evaluating the relationship across the entire outcome distribution, especially when the risk could be non-linear. There are limitations in these analyses that should be acknowledged. As with any observational study, residual confounding is a potential problem. We did not have information on whether participants had type 1 or type 2 diabetes; inability to differentiate if participants had type 1 or 2 diabetes was a concern given that they are two distinct diseases with unique risk factors and treatments. Additionally, we evaluated educational quantity, and assumed quality was comparable across respondents, potentially resulting in residual confounding in the relationship between educational attainment and HbA1c. We hypothesize that the education-HbA1c relationship could differ by race and ethnicity and sex, due to socially stratifying effects of gender and systemic racism; understanding if there are differential returns to education by sociodemographic subgroup is an important area for future study. Results from HRS underscores that these analyses should be replicated in other data sources to determine if these results are robust to variations in time, place, and population; evaluating if these relationships are causal is also warranted. Our results suggest the relationship between education and HbA1c is heterogeneous, varying both by education level and across the HbA1c distribution, with the largest associations in the high-risk right tail of the HbA1c distribution where risk of diabetes-related complications is highest. We found a one-year increase in average education for those with 12 or more years of education was associated with lower HbA1c, with larger point estimates for those in the high-risk tail of the HbA1c distribution. Our results add to understanding the education-HbA1c relationship and underscore the importance of evaluating the education-HbA1c relationship across the entire outcome distribution. Our results may also suggest an avenue for intervention: policies to increase education could reduce population-level diabetes complications, such as cardiovascular events and all-cause mortality. Abbreviations HbA1c: Hemoglobin A1c; CI: confidence interval Declarations Ethics approval and consent to participate: This work falls within a study that has been reviewed and approved by the University of California San Francisco’s Human Research Protection Program Institutional Review Board (IRB). The study was granted exempt certification (PI Vable, IRB #23-40341, Reference #408753). Consent for publication: Not applicable. Availability of data and materials: The data that support the findings of this study are available from the U.S. Health and Retirement Study (HRS), but restrictions apply to the availability of these data. Both public (Cross-Wave Tracker File and RAND HRS Longitudinal File 2018) and sensitive (2003 Diabetes Study; 2006-2016 Biomarker Data) datafiles were used in analysis. The HRS is sponsored by the National Institute on Aging (grant number NIA U01AG009740) and is conducted by the University of Michigan. Access to datasets are available through the HRS website (https://hrs.isr.umich.edu). The datasets generated and analyzed during the current study are available in the “More-schooling-is-associated-with-lower-Hemoglobin-A1c-at-the-high-risk-tail-of-the-distribution” GitHub repository. Competing interests: The authors declare that they have no competing interests. Funding: The analyses are supported by R01 AG069092 (PI Vable) from the National Institutes of Health. Elbert Huang is supported by grants from the National Institutes of Health (R01 AG060756, P30 DK092949). The study sponsors had no role in the study design, data collection, analysis, interpretation of results, writing the report, and decision to submit the report for publication. Authors’ contributions: JH cleaned and analyzed the data; they also interpreted the unconditional quantile regression results and wrote the manuscript. AI reviewed the cleaning, analysis, and graphics code and revised the manuscript. AK, ARR, and AMV provided editorial feedback and revised the manuscript. AA and ESH, experts in diabetes and clinical research, provided a thorough review of the manuscript as it related to diabetes and diabetes-related complications. All authors read and approved the final manuscript. Acknowledgements: The authors thank Catherine dP Duarte for her feedback on this manuscript. References Pinchevsky Y, Butkow N, Raal FJ, Chirwa T, Rothberg A. Demographic and Clinical Factors Associated with Development of Type 2 Diabetes: A Review of the Literature. International Journal of General Medicine [Internet]. 2020;13(1):121–9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7127847/ Mirzaei M, Rahmaninan M, Mirzaei M, Nadjarzadeh A, Dehghani tafti AA. Epidemiology of diabetes mellitus, pre-diabetes, undiagnosed and uncontrolled diabetes in Central Iran: results from Yazd health study. BMC Public Health. 2020 Feb 3;20(1). Zeru MA, Tesfa E, Mitiku AA, Seyoum A, Bokoro TA. Prevalence and risk factors of type-2 diabetes mellitus in Ethiopia: systematic review and meta-analysis. Scientific Reports. 2021 Nov 5;11(1). Nilsson PM, Johansson SE., Sundquist J. Low educational status is a risk factor for mortality among diabetic people. Diabetic Medicine. 2004 Jul 19;15(3):213–9. Aamir AH, Ul-Haq Z, Mahar SA, Qureshi FM, Ahmad I, Jawa A, et al. Diabetes Prevalence Survey of Pakistan (DPS-PAK): prevalence of type 2 diabetes mellitus and prediabetes using HbA1c: a population-based survey from Pakistan. BMJ Open [Internet]. 2019 Feb;9(2):1–9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6398762/ Östgren CJ, Sundström J, Svennblad B, Lohm L, Nilsson PM, Johansson G. Associations of HbA1c and educational level with risk of cardiovascular events in 32 871 drug‐treated patients with Type 2 diabetes: a cohort study in primary care. Diabetic Medicine. 2013 Jan;30(5):170–7. Yahyavi SK, Snorgaard O, Knop FK, Schou M, Lee C, Selmer C, et al. Prediabetes Defined by First Measured HbA1c Predicts Higher Cardiovascular Risk Compared With HbA1c in the Diabetes Range: A Cohort Study of Nationwide Registries. Diabetes Care [Internet]. 2021 Oct 21;44(12):2767–74. Available from: https://diabetesjournals.org/care/article/44/12/2767/138475/Prediabetes-Defined-by-First-Measured-HbA1c?searchresult=1 Kuo I-Ching, Lin HYH, Niu SW, Hwang DY, Lee JJ, Tsai JC, et al. Glycated Hemoglobin and Outcomes in Patients with Advanced Diabetic Chronic Kidney Disease. Scientific Reports [Internet]. 2016 Jan 28;6(1). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4730215/ Bots SH, van der Graaf Y, Nathoe HMW, de Borst GJ, Kappelle JL, Visseren FLJ, et al. The influence of baseline risk on the relation between HbA1c and risk for new cardiovascular events and mortality in patients with type 2 diabetes and symptomatic cardiovascular disease. Cardiovascular Diabetology. 2016 Jul 19;15(1). Orozco-Beltrán D, Navarro-Pérez J, Cebrián-Cuenca AM, Álvarez-Guisasola F, Caride-Miana E, Mora G, et al. The influence of hemoglobin A1c levels on cardiovascular events and all-cause mortality in people with diabetes over 70 years of age. A prospective study. Primary Care Diabetes. 2020 Jun;14(6). Jiao X, Zhang Q, Peng P, Shen Y. HbA1c is a predictive factor of severe coronary stenosis and major adverse cardiovascular events in patients with both type 2 diabetes and coronary heart disease. Diabetology & Metabolic Syndrome. 2023 Mar 20;15(1). Miech R, Hauser RM. Socioeconomic Status and Health at Midlife A Comparison of Educational Attainment with Occupation-Based Indicators. Annals of Epidemiology. 2001 Feb;11(2):75–84. Ma J, Pender M, Welch M. Education Pays 2016: The Benefits of Higher Education for Individuals and Society. Trends in Higher Education Series [Internet]. CollegeBoard; 2016 Dec p. 1–44. Available from: https://eric.ed.gov/?id=ED572548 Hales CM, Fryar CD, Carroll MD, Freedman DS, Aoki Y, Ogden CL. Differences in Obesity Prevalence by Demographic Characteristics and Urbanization Level Among Adults in the United States, 2013-2016. JAMA [Internet]. 2018 Jun 19;319(23):2419. Available from: https://jamanetwork.com/journals/jama/fullarticle/2685156 Ogden CL, Fakhouri TH, Carroll MD, Hales CM, Fryar CD, Li X, et al. Prevalence of Obesity Among Adults, by Household Income and Education — United States, 2011–2014. MMWR Morbidity and Mortality Weekly Report [Internet]. 2017 Dec 22;66(50):1369–73. Available from: https://www.cdc.gov/mmwr/volumes/66/wr/mm6650a1.htm?s_cid=mm6650a1_w#F1_down Vierboom YC. Trends in Alcohol-Related Mortality by Educational Attainment in the U.S., 2000–2017. Population Research and Policy Review. 2019 Apr 5;39(1):77–97. Briggs FH, Adler NE, Berkowitz SA, Chin MH, Webb TLG, Acien AN, et al. Social determinants of health and diabetes: A scientific review. Diabetes Care [Internet]. 2020;44(1):258–79. Available from: https://diabetesjournals.org/care/article/44/1/258/33180/Social-Determinants-of-Health-and-Diabetes-A Liu C, He L, Li Y, Yang A, Zhang K, Luo B. Diabetes risk among US adults with different socioeconomic status and behavioral lifestyles: evidence from the National Health and Nutrition Examination Survey. Frontiers in Public Health [Internet]. 2023;11(11):1197947. Available from: https://pubmed.ncbi.nlm.nih.gov/37674682/ Rabi DM, Edwards AL, Southern DA, Svenson LW, Sargious PM, Norton P, et al. Association of socio-economic status with diabetes prevalence and utilization of diabetes care services. BMC Health Services Research [Internet]. 2006;6(1). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1618393/ Case A, Deaton A. The Great Divide: Education, Despair, and Death. Annual Review of Economics. 2022 Aug 12;14(1):1–21. Crimmins E, Faul J, Kim JK, Guyer HM, Langa KM, Ofstedal MB, et al. Documentation of DBS Blood-Based Biomarkers in the 2016 Health and Retirement Study. Institute for Social Research, University of Michigan. 2013; Sherwani SI, Khan HA, Ekhzaimy A, Masood A, Sakharkar MK. Significance of Hba1c Test in Diagnosis and Prognosis of Diabetic Patients. Biomarker Insights [Internet]. 2016 Jul 3;11(11):95–104. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4933534/ American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care [Internet]. 2010 Dec 30;33(Supplement 1):S62–9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2797383/ Gillett MJ. International Expert Committee Report on the Role of the A1C Assay in the Diagnosis of Diabetes. Diabetes Care [Internet]. 2009 Jun 5;32(7):1327–34. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2699715/ Walsemann KM, Gee GC, Ro A. Educational Attainment in the Context of Social Inequality. American Behavioral Scientist. 2013 May 8;57(8):1082–104. Adkins-Jackson PB, Chantarat T, Bailey ZD, Ponce NA. Measuring Structural Racism: A guide for epidemiologists and other health researchers. American Journal of Epidemiology. 2022 Mar 24;191(4). Lett E, Asabor E, Beltrán S, Cannon AM, Arah OA. Conceptualizing, Contextualizing, and Operationalizing Race in Quantitative Health Sciences Research. The Annals of Family Medicine [Internet]. 2022 Jan 19;20(2):157–63. Available from: https://www.annfammed.org/content/annalsfm/early/2022/01/14/afm.2792.full.pdf Howard G, Howard VJ, Katholi C, Oli MK, Huston S. Decline in US Stroke Mortality: An Analysis of Temporal Patterns by Sex, Race, and Geographic Region. Stroke. 2001 Oct;32(10):2213–20. Lackland DT, Egan BM, Jones PJ. Impact of Nativity and Race on “Stroke Belt” Mortality. Hypertension. 1999 Jul;34(1):57–62. Lanska DJ, Peterson PM. Geographic Variation in Reporting of Stroke Deaths to Underlying or Contributing Causes in the United States. Stroke. 1995 Nov;26(11):1999–2003. Fang J, Madhavan S, Alderman MH. The Association between Birthplace and Mortality from Cardiovascular Causes among Black and White Residents of New York City. New England Journal of Medicine. 1996 Nov 21;335(21):1545–51. Vable AM, Gilsanz P, Nguyen TT, Kawachi I, Glymour MM. Validation of a theoretically motivated approach to measuring childhood socioeconomic circumstances in the Health and Retirement Study. Fraser A, editor. PLOS ONE. 2017 Oct 13;12(10):e0185898. Glymour MM, Avendaño M, Haas S, Berkman LF. Lifecourse Social Conditions and Racial Disparities in Incidence of First Stroke. Annals of Epidemiology. 2008 Dec;18(12):904–12. Firpo S, Fortin NM, Lemieux T. Unconditional Quantile Regressions. Econometrica [Internet]. 2009 May;77(3):953–73. Available from: https://www.jstor.org/stable/40263848 Firpo S, Pinto C. Identification and Estimation of Distributional Impacts of Interventions Using Changes in Inequality Measures. Journal of Applied Econometrics. 2015 Feb 24;31(3):457–86. Khadka A, Hebert JL, Glymour MM, Jiang F, Irish A, Duchowny KA, et al. Quantile regressions as a tool to evaluate how an exposure shifts and reshapes the outcome distribution: A primer for epidemiologists. American Journal of Epidemiology. 2024 Aug 3; Victora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach. International Journal of Epidemiology [Internet]. 1997 Feb 1;26(1):224–7. Available from: https://academic.oup.com/ije/article/26/1/224/730584 R Core Team. R: A language and environment for statistical computing. R: A language and environment for statistical computing. 2022; Vable AM, Diehl SF, Glymour MM. Code Review as a Simple Trick to Enhance Reproducibility, Accelerate Learning, and Improve the Quality of Your Team’s Research. American Journal of Epidemiology. 2021 Oct 1;190(10):2172–7. Schillinger D, Grumbach K, Piette J. Association of Health Literacy With Diabetes Outcomes. JAMA [Internet]. 2002 Jul;288(4):475–82. Available from: https://jamanetwork.com/journals/jama/fullarticle/195143 Vable AM, Kiang MV, Basu S, Rudolph KE, Kawachi I, Subramanian SV, et al. Military Service, Childhood Socio-Economic Status, and Late-Life Lung Function: Korean War Era Military Service Associated with Smaller Disparities. Military Medicine. 2018 Mar;183(9-10):e576–82. Vable AM, Canning D, Glymour MM, Kawachi I, Jimenez MP, Subramanian SV. Can social policy influence socioeconomic disparities? Korean War GI Bill eligibility and markers of depression. Annals of Epidemiology. 2016 Feb;26(2):129-135.e3. Vable AM, Kawachi I, Canning D, Glymour MM, Jimenez MP, Subramanian SV. Are There Spillover Effects from the GI Bill? The Mental Health of Wives of Korean War Veterans. PLOS ONE. 2016 May;11(5):e0154203. Meza E, Hebert J, Garcia ME, Torres JM, Glymour MM, Vable AM. First-generation college graduates have similar depressive symptoms in midlife as multi-generational college graduates. SSM - Population Health [Internet]. 2024 Feb;25:101633. Available from: https://www.sciencedirect.com/science/article/pii/S2352827324000338 Khadka A, Pacca L, Glymour MM, Bibbins-Domingo K, White JS, Basu S, et al. Impact of Vietnam-era G.I. Bill eligibility on later-life blood pressure distribution: evidence from the Vietnam draft lottery natural experiment. American Journal of Epidemiology. 2024 Sep; Duarte C, Wannier R, Cohen AK, Glymour MM, Ream RK, Yen IH, et al. Lifecourse Educational Trajectories and Hypertension in Midlife: An Application of Sequence Analysis. The Journals of Gerontology: Series A [Internet]. 2021 Aug;77(2):383–91. Available from: https://academic.oup.com/biomedgerontology/article/77/2/383/6359344 Vable AM, Cohen AK, Leonard SA, Glymour MM, Duarte C d.P., Yen IH. Do the health benefits of education vary by sociodemographic subgroup? Differential returns to education and implications for health inequities. Annals of Epidemiology. 2018 Nov;28(11):759-766.e5. Table TABLE 1. Distribution of Covariates in the Analytic Sample Overall N = 21,732 Less than 12 Years of Education N = 4,801; 22% 12 or More Years of Education N = 16,931; 78% Chi-Squared p-value Educational Attainment (Years) 12.7 (3.0) 8.5 (2.1) 13.9 (1.9) Age (Years) 65.1 (10.6) 67.2 (11.0) 64.5 (10.4) Average Birth Year 1944 (12.2) 1942 (12.5) 1944 (12.0) Proportion Female 57% 57% 57% 0.7295 Proportion in each Race and Ethnicity <0.0001 Non-Hispanic White 65% 43% 71% Non-Hispanic Black 18% 23% 17% Latinx/Hispanic 13% 31% 8% Other 4% 3% 4% Proportion in each Birthplace <0.0001 Non-Southern US 53% 32% 59% Southern US 34% 42% 31% Foreign 13% 26% 10% Average Mother’s Education 10.1 (3.1) 8.1 (2.7) 10.7 (2.9) Proportion Missing Mother’s Education 10% 20% 7% <0.0001 Average Father’s Education 9.9 (3.2) 8.1 (2.6) 10.4 (3.2) Proportion Missing Father’s Education 17% 30% 13% <0.0001 HbA1c Measurement Year 2003 6% 8% 5% 2006 27% 26% 27% 2008 27% 28% 26% 2010 14% 14% 14% 2012 13% 12% 14% 2014 5% 5% 5% 2016 8% 7% 9% Median HbA1c (IQR) 5.7% (5.3%-6.2%) 5.8% (5.5%-6.4%) 5.6% (5.3%-6.1%) Proportion Pre-Diabetic (HbA1c between 5.71%-6.49%) 30% 34% 29% <0.0001 Proportion Diabetic (HbA1c ≥ 6.5%) 17% 23% 15% <0.0001 Medication (Yes) 19% 28% 17% <0.0001 Note: Reported average mother and father’s education includes participants with missing values that were replaced with the sample mean (mean of 10 years for both mother’s and father’s education). Chi-square tests are used to compare significant differences in categorical variables by education level. Abbreviations: IQR = interquartile range Additional Declarations No competing interests reported. Supplementary Files AppendixRevisionAdditionalFile1.pdf Cite Share Download PDF Status: Published Journal Publication published 03 Jun, 2025 Read the published version in BMC Public Health → Version 1 posted Editorial decision: Accepted 06 May, 2025 Reviews received at journal 07 Apr, 2025 Reviews received at journal 07 Apr, 2025 Reviewers agreed at journal 02 Apr, 2025 Reviewers agreed at journal 28 Mar, 2025 Reviewers agreed at journal 24 Mar, 2025 Reviewers invited by journal 24 Mar, 2025 Submission checks completed at journal 24 Mar, 2025 First submitted to journal 21 Mar, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5462691","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":433440452,"identity":"b53268f5-cf08-4d48-90ec-607dd3b7d319","order_by":0,"name":"Jillian Hebert","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYBACxmbmxgMMDAcYDEC8D0DM3sDAIIFfC2MDTAtj4wygCM8BAlqAChFamnmI0cLczthwmIfhjpw5e+/zx7ZtdvY8DMwHb/MQcBhQyzNjy57jhs25bcmJPQxsydZEaDmcuOFGGiNQy4EEewYeM2litNRvuP+Msdmy7QDQYfzfiNKSYHCDDchuO8DYw8DDRlDLwTkGhw139qQxzuw5B/QLM5ux5Rw8Wgz7Dx988KbisLw5+zGGDz/KgCHG3vzwxht8WhoYGJh4DJCFmPEoBwF5kON+EFA0CkbBKBgFIxwAALzXTPRmS5t/AAAAAElFTkSuQmCC","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":true,"prefix":"","firstName":"Jillian","middleName":"","lastName":"Hebert","suffix":""},{"id":433440453,"identity":"2da29b9e-ea0f-4079-9ac3-77b5539ad205","order_by":1,"name":"Amanda Irish","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Amanda","middleName":"","lastName":"Irish","suffix":""},{"id":433440454,"identity":"abd4230e-57e8-44e4-98a9-471bbbf79c25","order_by":2,"name":"Aayush Khadka","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Aayush","middleName":"","lastName":"Khadka","suffix":""},{"id":433440455,"identity":"a667c9d2-2141-4756-b3ef-eca5f65249e9","order_by":3,"name":"Abigail Arons","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Abigail","middleName":"","lastName":"Arons","suffix":""},{"id":433440456,"identity":"e4bdb2b4-5a7b-4e66-bd12-0f431fdd3367","order_by":4,"name":"Alicia R. Riley","email":"","orcid":"","institution":"University of California, Santa Cruz","correspondingAuthor":false,"prefix":"","firstName":"Alicia","middleName":"R.","lastName":"Riley","suffix":""},{"id":433440457,"identity":"d733729c-38a7-462a-9072-84707cb86de7","order_by":5,"name":"Elbert S. Huang","email":"","orcid":"","institution":"University of Chicago","correspondingAuthor":false,"prefix":"","firstName":"Elbert","middleName":"S.","lastName":"Huang","suffix":""},{"id":433440458,"identity":"e28ccd04-529a-4a40-aa0b-3e1a0cb7b135","order_by":6,"name":"Anusha M. Vable","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Anusha","middleName":"M.","lastName":"Vable","suffix":""}],"badges":[],"createdAt":"2024-11-15 19:53:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5462691/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5462691/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12889-025-23063-x","type":"published","date":"2025-06-03T15:57:17+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":79255007,"identity":"fd37649d-3eb8-4e51-9b86-b842de8d2367","added_by":"auto","created_at":"2025-03-26 08:52:44","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":45298,"visible":true,"origin":"","legend":"\u003cp\u003eComparing results from linear regressions estimated using OLS with unconditional quantile regressions by dichotomized education\u003c/p\u003e\n\u003cp\u003eNote: OLS stands for Ordinary Least Squares. q10 = 10th quantile, q20 = 20th quantile, and so forth. The 90th quantile in the HbA1c distribution for those with less than 12 years of education was 7.7%. The 90th quantile in the HbA1c distribution for those with 12 or more years of education was 6.9%. The point to the left of the vertical dashed gray line in each panel represents the point estimate from OLS with 95% confidence intervals. The solid line to the right of the vertical dashed gray line in each panel represents point estimates from unconditional quantile regressions and shaded areas represent the 95% confidence intervals fit at each unit quantile between the 1st-99th quantiles of the HbA1c distribution. All models were adjusted for covariates and 95% confidence intervals were estimated using bootstrapping (500 resamples). Confidence intervals at the tails of the distribution are larger due to sparse data (i.e., fewer participants falling in those quantiles).\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5462691/v1/df175d5c9d5ddb15cadc0218.jpg"},{"id":79255005,"identity":"d3a03c83-c9e8-4d22-8d4b-ac6c974f7df2","added_by":"auto","created_at":"2025-03-26 08:52:44","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":36846,"visible":true,"origin":"","legend":"\u003cp\u003ePeaks of the Counterfactual HbA1c distributions based on unconditional quantile regression estimates by dichotomized education\u003c/p\u003e\n\u003cp\u003eNote: These graphics cut off the tails of the entire distribution to focus on the peak of the HbA1c distribution. The full HbA1c range for less than 12 years of education is 3.4% to 17.0% and the full HbA1c range for 12 or more years of education is 3.0% to 18.2%. Counterfactual HbA1c distributions were created using quantile estimates from UQR models. Counterfactual distributions were created to help visualize the reshaping of the HbA1c distribution for a one-year increase in the average educational attainment in the sample.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5462691/v1/5e29e96b797c53d4eb17f0ee.jpg"},{"id":84243119,"identity":"de759a39-4266-45b0-a8ea-5eeeb3715848","added_by":"auto","created_at":"2025-06-09 16:12:28","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":939704,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5462691/v1/64525ce7-7435-4c01-8d00-bb09e8f28352.pdf"},{"id":79255016,"identity":"f0941d5a-185c-4a17-9d11-8a9a1b4d15b9","added_by":"auto","created_at":"2025-03-26 08:52:44","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":788574,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixRevisionAdditionalFile1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5462691/v1/8aef2abef70d87c700255d4b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"More schooling is associated with lower Hemoglobin A1c at the high-risk tail of the distribution: An unconditional quantile regression analysis","fulltext":[{"header":"1.\tIntroduction","content":"\u003cp\u003ePrior work, both in the US and globally, reveals a strong inverse relationship between educational attainment and average glycosylated hemoglobin (HbA1c), an index of glucose regulation over the past 2-3 months, such that those with less education are at higher risk of diabetes and diabetes-related complications.\u003csup\u003e1-6\u0026nbsp;\u003c/sup\u003eAmong people with diabetes, glycemic control is essential for reducing complications, such as kidney failure, cardiovascular events, and all-cause mortality;\u003csup\u003e7-11\u003c/sup\u003e in parallel, staying within normal ranges of HbA1c is important to prevent diabetes onset and related complications. Risk of diabetes complications increase exponentially with HbA1c, such that a percentage-point increase at a higher level of HbA1c (e.g. 6% to 7%) confers a much greater risk for diabetes-related complications compared to a percentage-point increase at lower HbA1c levels (e.g. 4% to 5%). Therefore, exposures and interventions that have larger impacts for those with higher HbA1c may be important for preventing diabetes-related complications.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEducation, a main component of socioeconomic status, is consistently associated with better health over the lifecourse.\u003csup\u003e12-13\u003c/sup\u003e Increased educational attainment may impact HbA1c through several pathways, including longer life expectancy, increased income, better access to healthcare, and more health promoting behaviors (e.g., increased physical activity).\u003csup\u003e1,4,14-16\u003c/sup\u003e While evidence suggests a strong protective association between educational attainment and mean HbA1c, no study evaluates the association across the entire HbA1c distribution. Evaluating if the relationship is constant across the HbA1c distribution is important, given the exponential relationship between HbA1c and diabetes onset and related complications. Interventions that specifically impact high levels of HbA1c are of interest given their potential to reduce diabetes onset and related complications. We hypothesize education will have larger associations among those with higher HbA1c as participants belonging to more structurally minoritized subgroups (e.g., minoritized due to race, or poverty status) will be over-represented at the high-risk end of the HbA1c distribution (i.e., higher quantiles of the HbA1c distribution),\u003csup\u003e17-19\u003c/sup\u003e and these same groups also seem to benefit more from education.\u003csup\u003e45-47\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eWe add to existing literature by evaluating the relationship between education and HbA1c across the HbA1c distribution through a novel application of quantile regression, a modeling technique that evaluates the exposure-outcome relationship across the outcome distribution. Quantile regressions can identify if educational attainment has a heterogeneous effect across the HbA1c distribution, which allows for discovery of whether, and which parts of, the HbA1c’s distribution are differentially impacted by educational attainment. In this way, our application of quantile regression to the education-HbA1c relationship has the potential to deepen our understanding of educational inequities in diabetes by uncovering details hidden by linear regression.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis paper empirically evaluates the relationship between educational attainment and late-life HbA1c using linear regression and unconditional quantile regressions (UQR) for US Health and Retirement Study participants ages 50 and older, examining whether the relationship between education and HbA1c varies across the HbA1c distribution. We stratify education at 12 years, where 12 years of schooling typically confers a high school diploma, since prior literature finds evidence for divergent health trajectories of adults with less than a high school education from those with high school or more.\u003csup\u003e20\u003c/sup\u003e\u003c/p\u003e"},{"header":"2.\tMethods","content":"\u003cp\u003e\u003cem\u003eData and Analytic Sample\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eData came from the U.S. Health and Retirement Study (HRS), a national longitudinal sample of non-institutionalized adults 50 years and older, and their spouses of any age, that began in 1992.\u003csup\u003e21\u003c/sup\u003e New cohorts of participants have been added every six years after 1998 to maintain a steady state population, and participants are surveyed biennially. A diabetes sub study collected HbA1c for a subset of HRS participants in 2003; in addition, HbA1c and other biomarker data was collected in 2006 for a randomly selected half of the sample, and in 2008 for the other half; biomarker data were subsequently collected every four years.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe eligible sample included all HRS participants with at least one HbA1c measurement between 2003 and 2006 when they were 50 years or older (N = 21,840). Individuals were excluded for missing exposure (N = 89) and covariate data (N = 18); one participant was removed due to their HbA1c measurement being recorded prior to their first HRS interview, resulting in an analytic sample of 21,732 participants (99% of eligible).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExposure\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eOur exposure, educational attainment, was created using self-reported total years of schooling. Education in HRS ranges from 0 to 17 years of schooling, where 17 years includes those with 17 or more years of education (17 or more years: N = 2,327). Due to data sparseness of participants with fewer than 5 years of education, we coded those with less than 5 years of education to 5 years to reduce the impact these outliers may have on estimates (N = 776). To assess possible heterogeneities between participants with different types of credentials, and since educational policies tend to target specific levels of education (e.g. compulsory schooling laws and child labor laws targeted K-12 education, while other policies only addressed college education), education was stratified into two levels: fewer than 12 years of education (N = 4,801; 22%) and 12 or more years of education (N = 16,931; 78%), where completing 12 years of education typically corresponds to earning a high school diploma. Education was modeled linearly within these two educational strata.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eOutcome\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eOur outcome was the participant\u0026rsquo;s first recorded HbA1c value (2003-2016) measured at or after age 50. HbA1c is glycosylated hemoglobin and reflects blood glucose over the prior 2-3 months; HbA1c values between 5.71% and 6.49% are consistent with pre-diabetes and values greater than 6.5% are consistent with diabetes.\u003csup\u003e22-24\u003c/sup\u003e HbA1c was measured using an automated ion-exchange high-performance liquid chromatography that recorded the percentage of glycosylated hemoglobin in dried blood spot samples.\u003csup\u003e21\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCovariates\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAll models were adjusted for sex (female; male), race (Non-Hispanic White; Non-Hispanic Black; Latinx/Hispanic; other), birthplace (non-Southern US; Southern US; Foreign), indicator for birth year (1905-1966), indicator for year of HbA1c measurement (2003-2016), mother\u0026rsquo;s education (5-17(+) years, linear), father\u0026rsquo;s education (5-17(+) years, linear), as well as missing indicators for mother\u0026rsquo;s education and father\u0026rsquo;s education. Sex was included as an indicator of the socially stratifying effects of gender,\u003csup\u003e25\u003c/sup\u003e race as an indicator of the socially stratifying effects of systemic racism,\u003csup\u003e26-27\u003c/sup\u003e and parent\u0026rsquo;s education as a proxy for childhood socioeconomic status. Chi-square tests were used to evaluate if there were significant differences in categorical covariates by education level (e.g., less than 12 years of education versus 12 or more years of education). See supplemental Table S1 for additional details on covariates.\u003c/p\u003e\n\u003cp\u003eRace was categorized to include an \u0026ldquo;other\u0026rdquo; category in all models for precision, but results were not reported for this group due to the heterogeneous composition and consequent lack of interpretability of estimates.\u003c/p\u003e\n\u003cp\u003eBirthplace was classified by location within the US (i.e., non-Southern vs Southern) because studies have found increased risk for adverse later-life health outcomes for those born in the Southern US.\u003csup\u003e28-31\u003c/sup\u003e A subset of participants (N = 414, 2%) were known to be born in the US, but were missing information on the region of birth; participants where the region of birth was unknown were assumed to be born in the Non-Southern US.\u003c/p\u003e\n\u003cp\u003eBirth year was modeled as an indicator variable to capture differences by individual year. Due to a small number of participants falling in the tail ends of the birth year range, values were recoded to facilitate model convergence: those born before 1917 (N = 240, 1%) were recoded as 1917; those born in 1966 (N = 52, 0.2%) were recoded as 1965.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eParental education variables (5-17(+) years) were used as a proxy for family socioeconomic status (SES) and modeled continuously. However, HRS participants in the Asset and Health Dynamics Among the Oldest Old (AHEAD) cohort (born 1900-1923) recorded parent\u0026rsquo;s education as a dichotomized measure (less than 8 years of education, 8 or more years of education) rather than continuous. Dichotomized measures of parent\u0026rsquo;s education were replaced with continuous values from a previously validated imputation method using measures of childhood socioeconomic status.\u003csup\u003e32\u003c/sup\u003e Additional missingness in parent\u0026rsquo;s education was imputed using the sample mean (mother\u0026rsquo;s education: N\u003csub\u003eMissing\u003c/sub\u003e = 2,101 (10%), sample mean = 10 (SD 3.9); father\u0026rsquo;s education: N\u003csub\u003eMissing\u003c/sub\u003e = 3,644 (17%), sample mean = 10 (SD 4.2)) and a missing indicator was added for proper model adjustment. This allowed for retention of participants with missing parental education where missingness is informative (e.g., if the parent was not in the household).\u003csup\u003e33\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eStatistical Analysis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe used linear regressions and unconditional quantile regressions (UQR) to model the relationship between education and HbA1c.\u003csup\u003e34-36\u003c/sup\u003e UQR evaluates changes in quantiles of the outcome\u0026rsquo;s unconditional distribution for a one-unit change in the mean of the exposure. We estimated parameters of the linear regression model using ordinary least squares (OLS). We fit UQR models at each unit quantile between the 1st-99th quantiles of the unconditional HbA1c distribution. We used bootstrapping (500 repetitions) to estimate 95% confidence intervals (CIs) for parameters of the linear regression and UQR models; education was modeled as a linear term within both education strata and all models were adjusted for the covariates specified in the preceding section.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo visualize the change in the sample distribution of HbA1c implied by UQR results, we created plots to show the counterfactual HbA1c distribution for a one-year increase in the sample\u0026rsquo;s mean education. First, we binned the factual, or observed, data into quantiles (1st-99th). We then added the UQR estimation of the association between education and HbA1c to the observed data by quantile, creating a potential counterfactual distribution. Finally, we plotted the observed and counterfactual distributions. Further details about constructing these datasets and plots are provided elsewhere.\u003csup\u003e36\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSensitivity Analysis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted three additional analyses to see if results were robust to different analytic decisions. First, to determine if results were sensitive to how the exposure was operationalized (i.e., coded for analysis), we re-coded education as a three category variable (\u0026lt;12 years, 12-15 years, and 16 or more years of education) rather than two. Second, given that the HRS collects data from numerous birth cohorts (from 1905-1966) and given that educational attainment has tended to increase over time, we conducted analyses stratified by HRS entry cohort (which are defined by the HRS using participant\u0026rsquo;s birth year), to test the sensitivity of the association to secular trends in education. Finally, while medication could be an important contributing factor in the education-HbA1c relationship, medication usage is downstream from education and is therefore a potential mediator of the education-HbA1c relationship. Adjusting for mediators can bias estimates,\u003csup\u003e37\u003c/sup\u003e so the main analyses did not include adjustment for medication.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSoftware and Code\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAll data cleaning and analysis was performed in R.\u003csup\u003e38\u003c/sup\u003e We used the \u003cem\u003edineq\u003c/em\u003e package for fitting UQRs. All code was reviewed by the second author as recommended practice\u003csup\u003e39\u003c/sup\u003e and can be found on GitHub.\u003c/p\u003e"},{"header":"3.\tResults","content":"\u003cp\u003eParticipants included in analysis (Table 1) had an average of 13 years of education, were predominantly women (57%), White (65%), and born in the non-Southern US (53%). Most participants self-reported having 12 or more years of education (78%) and median HbA1c (Interquartile Range - IQR) in the overall sample was 5.7% (5.3%-6.2%). Compared to participants with 12 or more years of education (henceforth 12+ years), those with less than 12 years of education (henceforth \u0026lt; 12 years) were older (mean age: 67 vs. 65), had higher proportions of Black and Latinx participants (23% vs. 17% Black; 31% vs. 8% Latinx), had a higher proportion of Southern US birthplace (42% vs. 31%), and had lower parental education (mean mother\u0026rsquo;s education: 8 vs. 11; mean father\u0026rsquo;s education: 8 vs. 10) with higher rates of missingness (20% vs. 7% missing mothers education; 30% vs. 13% missing father education). Median HbA1c (IQR) for \u0026lt; 12 years was 5.8% (5.5%-6.4%) and 5.6% (5.3%-6.1%) for 12+ years. Chi-square tests found significant differences in race and ethnicity, birthplace, proportion missing parent\u0026rsquo;s education, and proportion of the sample with HbA1c levels that are consistent with being pre-diebetic or having diabetes.\u003c/p\u003e\n\u003cp\u003eTable 1\u003c/p\u003e\n\u003cp\u003eFigure 1 displays linear regression and UQR results for both education strata (\u0026lt;12 years; 12+ years). Linear regression results indicate that each additional year of education was not associated with average HbA1c for participants with \u0026lt;12 years (Figure 1, panel a: -0.00, 95% CI: -0.02, 0.02). UQR results for the 1st-99th quantiles suggest that a one-year increase in mean education was not associated with HbA1c across most quantiles, but may have been associated with higher HbA1c at higher quantiles (91-97th quantiles), although the confidence intervals here were wide (Figure 1, panel a: b\u003csub\u003eq5\u003c/sub\u003e = -0.02%; 95% CI: -0.03, 0.03, b\u003csub\u003eq50\u003c/sub\u003e = 0.00; 95% CI: -0.01, 0.01, b\u003csub\u003eq95\u003c/sub\u003e = 0.13%; 95% CI: -0.03, 0.29).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor participants with 12+ years, linear regression results indicate that each additional year of education was associated with lower average HbA1c (Figure 1, panel b: -0.02, 95% CI: -0.03, -0.02). UQR results suggest that a one-year increase in mean education was associated with lower HbA1c across almost all quantiles with a larger magnitude at higher quantiles (90-99th quantile) (Figure 1, panel b; b\u003csub\u003eq5\u003c/sub\u003e = -0.01%; 95% CI: -0.02, -0.00, b\u003csub\u003eq50\u003c/sub\u003e = -0.02%; 95% CI: -0.02, -0.01, b\u003csub\u003eq95\u003c/sub\u003e = -0.09%; 95% CI: -0.14, -0.04).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFigure 1\u003c/p\u003e\n\u003cp\u003eFigure 2 illustrates the factual, or observed, distribution of HbA1c in the sample and the predicted reshaping (i.e., counterfactual) of the HbA1c distribution based on the UQR estimates. Figure 2 helps to visualize the potential impact of increased education on the HbA1c distribution (i.e., the counterfactual distribution). \u0026nbsp;Consistent with the magnitude of the UQR estimates in Figure 1, changes in the counterfactual HbA1c distribution are small and therefore difficult to discern; to aid in visualization, Figure 2 zooms in on the peak of the HbA1c factual and counterfactual distributions. A figure including the full HbA1c distribution is included in the appendix (eFigure 1). For those with \u0026lt;12 years, the density at the peak of the counterfactual distribution is slightly lower than the density at the peak of the factual distribution, suggesting a small rightward shift away from the densest point of the distribution, towards higher HbA1c values. For 12+ years, the density at the peak of the counterfactual distribution is slightly higher than the density at the peak of the factual distribution, suggesting a small leftward shift in the direction of the densest part of the distribution, towards lower HbA1c values.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFigure 2\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSensitivity analyses\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eResults were attenuated but substantively similar and conclusions were unchanged when stratifying into 3 education strata (\u0026lt;12 years of education, 12-15 years of education, 16+ years of education) (eFigure 2). Results for 12-15 years of education were most similar to results for the 12+ years of education in the main analysis; results for 16+ years of education were null across most quantiles but were associated with lower HbA1c at higher quantiles (90-98th quantiles).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhen evaluating differences by entry cohorts, most results included the null, but estimates were largely in the same direction (eFigures 3-9). The War Babies (birth years: 1942-1947) and Mid Baby Boomer (birth years: 1954-1959) cohort\u0026rsquo;s results were most dissimilar to the main results.\u003c/p\u003e\n\u003cp\u003eResults and conclusions were unchanged after additional adjustments for medication (eFigure 10).\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn the HRS sample of older US adults, we found that the relationship between educational attainment and HbA1c varies by education level and is heterogeneous across the HbA1c distribution. Overall, associations between education and HbA1c seem to be strongest at the high-risk tail of the HbA1c distribution (i.e., participants with the highest HbA1c levels; 90-99th quantiles). That is, participants with the highest HbA1c levels who are most at risk of developing diabetes or experiencing diabetes-related complications such as cardiovascular events and all-cause mortality had the strongest association with educational attainment. We also found that these associations varied by total years of schooling: among those with less than 12 years of education, a one-year increase in mean education was not associated with HbA1c for participants with low to average HbA1c levels, however, it was weakly associated with higher HbA1c for participants with high HbA1c levels. Conversely, among those with 12 or more years of education, a one-year increase in mean education was associated with lower HbA1c across the distribution, with larger magnitudes in the association at the high-risk tail of the distribution; more education was associated with a leftward shift in the HbA1c distribution, suggesting lower risk of diabetes and diabetes-related complications in the sample population. While the associations appear minimal, our work adds to existing literature by demonstrating a method of evaluating heterogeneous relationships and identifying that the relationship between education and HbA1c differs by education level.\u003c/p\u003e\n\u003cp\u003eThere were two analytic decisions that allowed us to identify these novel findings. First, we stratified our exposure at 12 years of education. This stratification was informed by prior research highlighting the divergent health trajectories of adults with less than a high school education from those with high school or more.\u003csup\u003e20\u003c/sup\u003e Second, we used UQRs to evaluate the relationship across the entire HbA1c distribution, as opposed to evaluating the relationship at the mean, where one assumes the change at the mean is constant across the entire HbA1c distribution. Among older adults with more than 12 years of education, estimates at the mean were associated with lower HbA1c, as were estimates for most of the quantiles in the HbA1c distribution; at the highest quantiles of the HbA1c distribution, magnitudes of the estimates were larger, suggesting that there may be a larger association for those with higher levels of HbA1c. This discovery additionally exemplifies that the mean association was not constant across the entire HbA1c distribution.\u003c/p\u003e\n\u003cp\u003eOur findings that increased educational attainment was associated with lower HbA1c among those with 12 or more years of education is consistent with prior literature showing that lower educational attainment is associated with increased risk of diabetes and more broadly that detrimental social factors strongly affect risk of diabetes.\u003csup\u003e1-17\u003c/sup\u003e Our use of quantile regressions revealed a larger magnitude of the education-HbA1c association at the high-risk tail of the HbA1c distribution, where individuals are more likely to have diabetes. We hypothesize two potential mechanisms for this: First, the impact of structural factors (race, poverty, historic systems of marginalization and exclusion) on educational attainment and HbA1c. Evidence shows that education has a larger impact among low childhood socioeconomic status or racial and ethnic minoritized groups, when compared to structurally advantaged groups (e.g., White).\u003csup\u003e41-44, 47\u003c/sup\u003e Second, is that education impacts skill-level, which, in turn, impacts the type of job one attains and their income. Therefore, individuals with diabetes who have higher levels of education likely also have greater access to resources, such as money, health care and medications, or social and behavioral resources that support lifestyle changes to better manage glucose levels\u003cem\u003e.\u003c/em\u003e\u003csup\u003e40, 45-46\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eOur findings about the relationship between HbA1c and education are relevant for both clinicians and public health policymakers. At the clinical level, this association underscores the role of social determinants of health screening as a part of diabetes risk assessment, and future research should explore the added benefit of more precise screening for education in terms of number of years rather than more commonly used broad categories of educational attainment. Such screening and subsequent linkage to social services has the potential to support clinicians in better identifying and supporting patients to lower their risk of diabetes (e.g., pursuing earlier or more intensive preventive intervention or more aggressive glycemic control once diagnosed with diabetes). Additional studies could explore the effectiveness of interventions for these patients that seek to mitigate the detrimental effect of less education on HbA1c, such as disease education programs, navigational supports, and pharmacy management tools. Finally, our results raise the question of whether increasing educational attainment, even by 1 year, for populations would be an effective intervention to reduce future glucose levels. Future secondary data analyses and intervention studies can further elucidate whether individual-level or population-level interventions that increase post-secondary educational attainment may be an effective strategy for lowering diabetes risk.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIt is important to contextualize why results for those with less than 12 years of education are different from those with 12 or more years of education, especially because the results for those with less than 12 years of education are contrary to our hypothesis and prior literature. It could be that those with less than 12 years of education are a different, more structurally minoritized group than those with more education, resulting in different relationships between education and HbA1c; we found significant differences in race and ethnicity, birthplace, proportion missing parent’s education, and proportion of the sample with HbA1c levels that are consistent with being pre-diebetic or having diabetes. Our sample of those with less than 12 years of education were more likely to be people of color, born outside of the US or in the Southern US, and their parents had less education or more missing data on education – all potential indicators of increased marginalization. Results could also be due to standard variability in the unconditional quantile regression estimates at the tails of the distribution, where the density is low, and the variance of the RIF is larger.\u003csup\u003e34\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eAn important strength of this analysis is the additional information provided by the UQR modeling technique. By looking at the exposure-outcome relationship across the entire outcome distribution, estimates can identify heterogeneities in the education and HbA1c relationship. OLS results only suggest lower HbA1c levels given an additional year of education for participants with 12 or more years of education. UQR results further suggest that the magnitude of change in HbA1c is larger for those at the high-risk tail of the distribution. Comparing the OLS and UQR results highlights the limitation of mean models in capturing an exposure’s relationship with the outcome distribution, and the necessity for evaluating the relationship across the entire outcome distribution, especially when the risk could be non-linear.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThere are limitations in these analyses that should be acknowledged. As with any observational study, residual confounding is a potential problem. We did not have information on whether participants had type 1 or type 2 diabetes; inability to differentiate if participants had type 1 or 2 diabetes was a concern given that they are two distinct diseases with unique risk factors and treatments. Additionally, we evaluated educational quantity, and assumed quality was comparable across respondents, potentially resulting in residual confounding in the relationship between educational attainment and HbA1c. We hypothesize that the education-HbA1c relationship could differ by race and ethnicity and sex, due to socially stratifying effects of gender and systemic racism; understanding if there are differential returns to education by sociodemographic subgroup is an important area for future study. Results from HRS underscores that these analyses should be replicated in other data sources to determine if these results are robust to variations in time, place, and population; evaluating if these relationships are causal is also warranted.\u003c/p\u003e\n\u003cp\u003eOur results suggest the relationship between education and HbA1c is heterogeneous, varying both by education level and across the HbA1c distribution, with the largest associations in the high-risk right tail of the HbA1c distribution where risk of diabetes-related complications is highest. We found a one-year increase in average education for those with 12 or more years of education was associated with lower HbA1c, with larger point estimates for those in the high-risk tail of the HbA1c distribution. Our results add to understanding the education-HbA1c relationship and underscore the importance of evaluating the education-HbA1c relationship across the entire outcome distribution. Our results may also suggest an avenue for intervention: policies to increase education could reduce population-level diabetes complications, such as cardiovascular events and all-cause mortality.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eHbA1c:\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eHemoglobin A1c; CI: confidence interval\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work falls within a study that has been reviewed and approved by the University of California San Francisco’s Human Research Protection Program Institutional Review Board (IRB). The study was granted exempt certification (PI Vable, IRB #23-40341, Reference #408753).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the U.S. Health and Retirement Study (HRS), but restrictions apply to the availability of these data. Both public (Cross-Wave Tracker File and RAND HRS Longitudinal File 2018) and sensitive (2003 Diabetes Study; 2006-2016 Biomarker Data) datafiles were used in analysis. The HRS is sponsored by the National Institute on Aging (grant number NIA U01AG009740) and is conducted by the University of Michigan. Access to datasets are available through the HRS website (https://hrs.isr.umich.edu). The datasets generated and analyzed during the current study are available in the “More-schooling-is-associated-with-lower-Hemoglobin-A1c-at-the-high-risk-tail-of-the-distribution” GitHub repository.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe analyses are supported by R01 AG069092 (PI Vable) from the National Institutes of Health. Elbert Huang is supported by grants from the National Institutes of Health (R01 AG060756, P30 DK092949). The study sponsors had no role in the study design, data collection, analysis, interpretation of results, writing the report, and decision to submit the report for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors’ contributions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJH cleaned and analyzed the data; they also interpreted the unconditional quantile regression results and wrote the manuscript. AI reviewed the cleaning, analysis, and graphics code and revised the manuscript. AK, ARR, and AMV provided editorial feedback and revised the manuscript. AA and ESH, experts in diabetes and clinical research, provided a thorough review of the manuscript as it related to diabetes and diabetes-related complications. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe authors thank Catherine dP Duarte for her feedback on this manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePinchevsky Y, Butkow N, Raal FJ, Chirwa T, Rothberg A. Demographic and Clinical Factors Associated with Development of Type 2 Diabetes: A Review of the Literature. International Journal of General Medicine [Internet]. 2020;13(1):121\u0026ndash;9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7127847/\u003c/li\u003e\n\u003cli\u003eMirzaei M, Rahmaninan M, Mirzaei M, Nadjarzadeh A, Dehghani tafti AA. Epidemiology of diabetes mellitus, pre-diabetes, undiagnosed and uncontrolled diabetes in Central Iran: results from Yazd health study. BMC Public Health. 2020 Feb 3;20(1).\u003c/li\u003e\n\u003cli\u003eZeru MA, Tesfa E, Mitiku AA, Seyoum A, Bokoro TA. Prevalence and risk factors of type-2 diabetes mellitus in Ethiopia: systematic review and meta-analysis. Scientific Reports. 2021 Nov 5;11(1).\u003c/li\u003e\n\u003cli\u003eNilsson PM, Johansson SE., Sundquist J. Low educational status is a risk factor for mortality among diabetic people. Diabetic Medicine. 2004 Jul 19;15(3):213\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eAamir AH, Ul-Haq Z, Mahar SA, Qureshi FM, Ahmad I, Jawa A, et al. Diabetes Prevalence Survey of Pakistan (DPS-PAK): prevalence of type 2 diabetes mellitus and prediabetes using HbA1c: a population-based survey from Pakistan. BMJ Open [Internet]. 2019 Feb;9(2):1\u0026ndash;9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6398762/\u003c/li\u003e\n\u003cli\u003e\u0026Ouml;stgren CJ, Sundstr\u0026ouml;m J, Svennblad B, Lohm L, Nilsson PM, Johansson G. Associations of HbA1c and educational level with risk of cardiovascular events in 32 871 drug‐treated patients with Type 2 diabetes: a cohort study in primary care. Diabetic Medicine. 2013 Jan;30(5):170\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eYahyavi SK, Snorgaard O, Knop FK, Schou M, Lee C, Selmer C, et al. Prediabetes Defined by First Measured HbA1c Predicts Higher Cardiovascular Risk Compared With HbA1c in the Diabetes Range: A Cohort Study of Nationwide Registries. Diabetes Care [Internet]. 2021 Oct 21;44(12):2767\u0026ndash;74. Available from: https://diabetesjournals.org/care/article/44/12/2767/138475/Prediabetes-Defined-by-First-Measured-HbA1c?searchresult=1\u003c/li\u003e\n\u003cli\u003eKuo I-Ching, Lin HYH, Niu SW, Hwang DY, Lee JJ, Tsai JC, et al. Glycated Hemoglobin and Outcomes in Patients with Advanced Diabetic Chronic Kidney Disease. Scientific Reports [Internet]. 2016 Jan 28;6(1). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4730215/\u003c/li\u003e\n\u003cli\u003eBots SH, van der Graaf Y, Nathoe HMW, de Borst GJ, Kappelle JL, Visseren FLJ, et al. The influence of baseline risk on the relation between HbA1c and risk for new cardiovascular events and mortality in patients with type 2 diabetes and symptomatic cardiovascular disease. Cardiovascular Diabetology. 2016 Jul 19;15(1).\u003c/li\u003e\n\u003cli\u003eOrozco-Beltr\u0026aacute;n D, Navarro-P\u0026eacute;rez J, Cebri\u0026aacute;n-Cuenca AM, \u0026Aacute;lvarez-Guisasola F, Caride-Miana E, Mora G, et al. The influence of hemoglobin A1c levels on cardiovascular events and all-cause mortality in people with diabetes over 70 years of age. A prospective study. Primary Care Diabetes. 2020 Jun;14(6).\u003c/li\u003e\n\u003cli\u003eJiao X, Zhang Q, Peng P, Shen Y. HbA1c is a predictive factor of severe coronary stenosis and major adverse cardiovascular events in patients with both type 2 diabetes and coronary heart disease. Diabetology \u0026amp; Metabolic Syndrome. 2023 Mar 20;15(1).\u003c/li\u003e\n\u003cli\u003eMiech R, Hauser RM. Socioeconomic Status and Health at Midlife A Comparison of Educational Attainment with Occupation-Based Indicators. Annals of Epidemiology. 2001 Feb;11(2):75\u0026ndash;84.\u003c/li\u003e\n\u003cli\u003eMa J, Pender M, Welch M. Education Pays 2016: The Benefits of Higher Education for Individuals and Society. Trends in Higher Education Series [Internet]. CollegeBoard; 2016 Dec p. 1\u0026ndash;44. Available from: https://eric.ed.gov/?id=ED572548\u003c/li\u003e\n\u003cli\u003eHales CM, Fryar CD, Carroll MD, Freedman DS, Aoki Y, Ogden CL. Differences in Obesity Prevalence by Demographic Characteristics and Urbanization Level Among Adults in the United States, 2013-2016. JAMA [Internet]. 2018 Jun 19;319(23):2419. Available from: https://jamanetwork.com/journals/jama/fullarticle/2685156\u003c/li\u003e\n\u003cli\u003eOgden CL, Fakhouri TH, Carroll MD, Hales CM, Fryar CD, Li X, et al. Prevalence of Obesity Among Adults, by Household Income and Education \u0026mdash; United States, 2011\u0026ndash;2014. MMWR Morbidity and Mortality Weekly Report [Internet]. 2017 Dec 22;66(50):1369\u0026ndash;73. Available from: https://www.cdc.gov/mmwr/volumes/66/wr/mm6650a1.htm?s_cid=mm6650a1_w#F1_down\u003c/li\u003e\n\u003cli\u003eVierboom YC. Trends in Alcohol-Related Mortality by Educational Attainment in the U.S., 2000\u0026ndash;2017. Population Research and Policy Review. 2019 Apr 5;39(1):77\u0026ndash;97.\u003c/li\u003e\n\u003cli\u003eBriggs FH, Adler NE, Berkowitz SA, Chin MH, Webb TLG, Acien AN, et al. Social determinants of health and diabetes: A scientific review. Diabetes Care [Internet]. 2020;44(1):258\u0026ndash;79. Available from: https://diabetesjournals.org/care/article/44/1/258/33180/Social-Determinants-of-Health-and-Diabetes-A\u003c/li\u003e\n\u003cli\u003eLiu C, He L, Li Y, Yang A, Zhang K, Luo B. Diabetes risk among US adults with different socioeconomic status and behavioral lifestyles: evidence from the National Health and Nutrition Examination Survey. Frontiers in Public Health [Internet]. 2023;11(11):1197947. Available from: https://pubmed.ncbi.nlm.nih.gov/37674682/\u003c/li\u003e\n\u003cli\u003eRabi DM, Edwards AL, Southern DA, Svenson LW, Sargious PM, Norton P, et al. Association of socio-economic status with diabetes prevalence and utilization of diabetes care services. BMC Health Services Research [Internet]. 2006;6(1). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1618393/\u003c/li\u003e\n\u003cli\u003eCase A, Deaton A. The Great Divide: Education, Despair, and Death. Annual Review of Economics. 2022 Aug 12;14(1):1\u0026ndash;21.\u003c/li\u003e\n\u003cli\u003eCrimmins E, Faul J, Kim JK, Guyer HM, Langa KM, Ofstedal MB, et al. Documentation of DBS Blood-Based Biomarkers in the 2016 Health and Retirement Study. Institute for Social Research, University of Michigan. 2013;\u003c/li\u003e\n\u003cli\u003eSherwani SI, Khan HA, Ekhzaimy A, Masood A, Sakharkar MK. Significance of Hba1c Test in Diagnosis and Prognosis of Diabetic Patients. Biomarker Insights [Internet]. 2016 Jul 3;11(11):95\u0026ndash;104. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4933534/\u003c/li\u003e\n\u003cli\u003eAmerican Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care [Internet]. 2010 Dec 30;33(Supplement 1):S62\u0026ndash;9. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2797383/\u003c/li\u003e\n\u003cli\u003eGillett MJ. International Expert Committee Report on the Role of the A1C Assay in the Diagnosis of Diabetes. Diabetes Care [Internet]. 2009 Jun 5;32(7):1327\u0026ndash;34. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2699715/\u003c/li\u003e\n\u003cli\u003eWalsemann KM, Gee GC, Ro A. Educational Attainment in the Context of Social Inequality. American Behavioral Scientist. 2013 May 8;57(8):1082\u0026ndash;104.\u003c/li\u003e\n\u003cli\u003eAdkins-Jackson PB, Chantarat T, Bailey ZD, Ponce NA. Measuring Structural Racism: A guide for epidemiologists and other health researchers. American Journal of Epidemiology. 2022 Mar 24;191(4).\u003c/li\u003e\n\u003cli\u003eLett E, Asabor E, Beltr\u0026aacute;n S, Cannon AM, Arah OA. Conceptualizing, Contextualizing, and Operationalizing Race in Quantitative Health Sciences Research. The Annals of Family Medicine [Internet]. 2022 Jan 19;20(2):157\u0026ndash;63. Available from: https://www.annfammed.org/content/annalsfm/early/2022/01/14/afm.2792.full.pdf\u003c/li\u003e\n\u003cli\u003eHoward G, Howard VJ, Katholi C, Oli MK, Huston S. Decline in US Stroke Mortality: An Analysis of Temporal Patterns by Sex, Race, and Geographic Region. Stroke. 2001 Oct;32(10):2213\u0026ndash;20.\u003c/li\u003e\n\u003cli\u003eLackland DT, Egan BM, Jones PJ. Impact of Nativity and Race on \u0026ldquo;Stroke Belt\u0026rdquo; Mortality. Hypertension. 1999 Jul;34(1):57\u0026ndash;62.\u003c/li\u003e\n\u003cli\u003eLanska DJ, Peterson PM. Geographic Variation in Reporting of Stroke Deaths to Underlying or Contributing Causes in the United States. Stroke. 1995 Nov;26(11):1999\u0026ndash;2003.\u003c/li\u003e\n\u003cli\u003eFang J, Madhavan S, Alderman MH. The Association between Birthplace and Mortality from Cardiovascular Causes among Black and White Residents of New York City. New England Journal of Medicine. 1996 Nov 21;335(21):1545\u0026ndash;51.\u003c/li\u003e\n\u003cli\u003eVable AM, Gilsanz P, Nguyen TT, Kawachi I, Glymour MM. Validation of a theoretically motivated approach to measuring childhood socioeconomic circumstances in the Health and Retirement Study. Fraser A, editor. PLOS ONE. 2017 Oct 13;12(10):e0185898.\u003c/li\u003e\n\u003cli\u003eGlymour MM, Avenda\u0026ntilde;o M, Haas S, Berkman LF. Lifecourse Social Conditions and Racial Disparities in Incidence of First Stroke. Annals of Epidemiology. 2008 Dec;18(12):904\u0026ndash;12.\u003c/li\u003e\n\u003cli\u003eFirpo S, Fortin NM, Lemieux T. Unconditional Quantile Regressions. Econometrica [Internet]. 2009 May;77(3):953\u0026ndash;73. Available from: https://www.jstor.org/stable/40263848\u003c/li\u003e\n\u003cli\u003eFirpo S, Pinto C. Identification and Estimation of Distributional Impacts of Interventions Using Changes in Inequality Measures. Journal of Applied Econometrics. 2015 Feb 24;31(3):457\u0026ndash;86.\u003c/li\u003e\n\u003cli\u003eKhadka A, Hebert JL, Glymour MM, Jiang F, Irish A, Duchowny KA, et al. Quantile regressions as a tool to evaluate how an exposure shifts and reshapes the outcome distribution: A primer for epidemiologists. American Journal of Epidemiology. 2024 Aug 3;\u003c/li\u003e\n\u003cli\u003eVictora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach. International Journal of Epidemiology [Internet]. 1997 Feb 1;26(1):224\u0026ndash;7. Available from: https://academic.oup.com/ije/article/26/1/224/730584\u003c/li\u003e\n\u003cli\u003eR Core Team. R: A language and environment for statistical computing. R: A language and environment for statistical computing. 2022;\u003c/li\u003e\n\u003cli\u003eVable AM, Diehl SF, Glymour MM. Code Review as a Simple Trick to Enhance Reproducibility, Accelerate Learning, and Improve the Quality of Your Team\u0026rsquo;s Research. American Journal of Epidemiology. 2021 Oct 1;190(10):2172\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eSchillinger D, Grumbach K, Piette J. Association of Health Literacy With Diabetes Outcomes. JAMA [Internet]. 2002 Jul;288(4):475\u0026ndash;82. Available from: https://jamanetwork.com/journals/jama/fullarticle/195143\u003c/li\u003e\n\u003cli\u003eVable AM, Kiang MV, Basu S, Rudolph KE, Kawachi I, Subramanian SV, et al. Military Service, Childhood Socio-Economic Status, and Late-Life Lung Function: Korean War Era Military Service Associated with Smaller Disparities. Military Medicine. 2018 Mar;183(9-10):e576\u0026ndash;82.\u003c/li\u003e\n\u003cli\u003eVable AM, Canning D, Glymour MM, Kawachi I, Jimenez MP, Subramanian SV. Can social policy influence socioeconomic disparities? Korean War GI Bill eligibility and markers of depression. Annals of Epidemiology. 2016 Feb;26(2):129-135.e3.\u003c/li\u003e\n\u003cli\u003eVable AM, Kawachi I, Canning D, Glymour MM, Jimenez MP, Subramanian SV. Are There Spillover Effects from the GI Bill? The Mental Health of Wives of Korean War Veterans. PLOS ONE. 2016 May;11(5):e0154203.\u003c/li\u003e\n\u003cli\u003eMeza E, Hebert J, Garcia ME, Torres JM, Glymour MM, Vable AM. First-generation college graduates have similar depressive symptoms in midlife as multi-generational college graduates. SSM - Population Health [Internet]. 2024 Feb;25:101633. Available from: https://www.sciencedirect.com/science/article/pii/S2352827324000338\u003c/li\u003e\n\u003cli\u003eKhadka A, Pacca L, Glymour MM, Bibbins-Domingo K, White JS, Basu S, et al. Impact of Vietnam-era G.I. Bill eligibility on later-life blood pressure distribution: evidence from the Vietnam draft lottery natural experiment. American Journal of Epidemiology. 2024 Sep;\u003c/li\u003e\n\u003cli\u003eDuarte C, Wannier R, Cohen AK, Glymour MM, Ream RK, Yen IH, et al. Lifecourse Educational Trajectories and Hypertension in Midlife: An Application of Sequence Analysis. The Journals of Gerontology: Series A [Internet]. 2021 Aug;77(2):383\u0026ndash;91. Available from: https://academic.oup.com/biomedgerontology/article/77/2/383/6359344\u003c/li\u003e\n\u003cli\u003eVable AM, Cohen AK, Leonard SA, Glymour MM, Duarte C d.P., Yen IH. Do the health benefits of education vary by sociodemographic subgroup? Differential returns to education and implications for health inequities. Annals of Epidemiology. 2018 Nov;28(11):759-766.e5.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table","content":"\u003cp\u003e\u003cstrong\u003eTABLE 1. Distribution of Covariates in the Analytic Sample\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"643\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOverall\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eN = 21,732\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLess than 12 Years of Education\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eN = 4,801; 22%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e12 or More Years of Education\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eN = 16,931; 78%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChi-Squared p-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEducational Attainment (Years)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e12.7 (3.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e8.5 (2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e13.9 (1.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge (Years)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e65.1 (10.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e67.2 (11.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e64.5 (10.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAverage Birth Year\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e1944 (12.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e1942 (12.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e1944 (12.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion Female\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e57%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e57%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e57%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e0.7295\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion in each Race and Ethnicity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-Hispanic White\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e65%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e43%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e71%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-Hispanic Black\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e18%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e23%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e17%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLatinx/Hispanic\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e13%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e31%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e8%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOther\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e4%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e3%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e4%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion in each Birthplace\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNon-Southern US\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e53%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e32%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e59%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSouthern US\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e34%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e42%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e31%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eForeign\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e13%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e26%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e10%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAverage Mother\u0026rsquo;s Education\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e10.1 (3.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e8.1 (2.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e10.7 (2.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion Missing Mother\u0026rsquo;s Education\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e10%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e20%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e7%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAverage Father\u0026rsquo;s Education\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e9.9 (3.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e8.1 (2.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e10.4 (3.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion Missing Father\u0026rsquo;s Education\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e17%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e30%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e13%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHbA1c Measurement Year\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2003\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e6%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e8%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e5%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2006\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e27%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e26%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e27%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2008\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e27%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e28%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e26%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2010\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e14%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e14%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e14%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2012\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e13%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e12%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e14%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2014\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e5%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e5%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e5%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2016\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e8%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e7%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e9%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMedian HbA1c (IQR)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e5.7%\u003c/p\u003e\n \u003cp\u003e(5.3%-6.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e5.8%\u003c/p\u003e\n \u003cp\u003e(5.5%-6.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e5.6%\u003c/p\u003e\n \u003cp\u003e(5.3%-6.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion Pre-Diabetic\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(HbA1c between 5.71%-6.49%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e30%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e34%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e29%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProportion Diabetic (HbA1c \u0026ge; 6.5%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e17%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e23%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e15%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMedication (Yes)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.6407%;\"\u003e\n \u003cp\u003e19%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 19.5956%;\"\u003e\n \u003cp\u003e28%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 18.9736%;\"\u003e\n \u003cp\u003e17%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.3297%;\"\u003e\n \u003cp\u003e\u0026lt;0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eNote: Reported average mother and father\u0026rsquo;s education includes participants with missing values that were replaced with the sample mean (mean of 10 years for both mother\u0026rsquo;s and father\u0026rsquo;s education). Chi-square tests are used to compare significant differences in categorical variables by education level.\u003c/p\u003e\n\u003cp\u003eAbbreviations: IQR = interquartile range\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Unconditional quantile regression, Distributional effects, Effect heterogeneity, Diabetes, US Health and Retirement Study (HRS)","lastPublishedDoi":"10.21203/rs.3.rs-5462691/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5462691/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRisk of diabetes increases exponentially with higher levels of glycosylated hemoglobin (HbA1c). Education is inversely associated with average HbA1c, however, differential associations between education and HbA1c across the HbA1c distribution have not been evaluated.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHealth and Retirement Study data (N=21,732) was used to evaluate the association between education (linear terms among those with \u0026lt;12 years and ≥ 12 years of education) and first recorded HbA1c (2003-2016) at the mean using linear regression, and at the 1st-99th quantiles of the marginal outcome distribution using unconditional quantile regressions, controlling for birth year, race and ethnicity, gender, birthplace, parent’s education, and year of HbA1c measurement.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMean HbA1c was 5.9%; 16.6% of participants had HbA1c above the diabetes diagnostic threshold of 6.5%. For those with less than 12 years of schooling, there was no association between education and HbA1c at the mean or across the quantiles. For those with 12 or more years of schooling, an additional year of education was negatively associated with mean HbA1c (b\u003csub\u003eOLS\u003c/sub\u003e=-0.02, 95% confidence interval (CI) -0.03,-0.02); a one-year increase in mean education was associated with lower HbA1c across the distribution, but the magnitude was larger at higher quantiles (b\u003csub\u003eq50\u003c/sub\u003e=-0.02, 95%CI -0.02,-0.01; b\u003csub\u003eq90\u003c/sub\u003e=-0.06, 95%CI -0.09,-0.04).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEducational attainment is inversely associated with HbA1camong those with 12 or more years of schooling, with larger point estimates for those in the high-risk tail of the HbA1c distribution.\u003c/p\u003e","manuscriptTitle":"More schooling is associated with lower Hemoglobin A1c at the high-risk tail of the distribution: An unconditional quantile regression analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-26 08:52:40","doi":"10.21203/rs.3.rs-5462691/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Accepted","date":"2025-05-06T07:59:15+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-07T23:56:17+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-07T09:13:32+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"52301228294142820478191748506406619486","date":"2025-04-02T22:21:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"139829572803898020508317235205363256654","date":"2025-03-28T15:11:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"44011307135102473187143876296063971495","date":"2025-03-25T00:18:03+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-03-24T17:17:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-03-24T09:40:57+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Public Health","date":"2025-03-21T19:04:13+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ef463e24-7b53-4054-872e-e3398880765f","owner":[],"postedDate":"March 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-06-09T16:08:50+00:00","versionOfRecord":{"articleIdentity":"rs-5462691","link":"https://doi.org/10.1186/s12889-025-23063-x","journal":{"identity":"bmc-public-health","isVorOnly":false,"title":"BMC Public Health"},"publishedOn":"2025-06-03 15:57:17","publishedOnDateReadable":"June 3rd, 2025"},"versionCreatedAt":"2025-03-26 08:52:40","video":"","vorDoi":"10.1186/s12889-025-23063-x","vorDoiUrl":"https://doi.org/10.1186/s12889-025-23063-x","workflowStages":[]},"version":"v1","identity":"rs-5462691","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5462691","identity":"rs-5462691","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}