Estimating the direct and indirect pathways between education and diabetes incidence among Canadian men and women: a mediation analysis

Exploring Mediators of Mental Health Service Use Among Transgender Individuals in Ontario, Canada

OBJECTIVES

To determine if and to what degree neighbourhood-level marginalization mediates mental health service use among transgender individuals.

METHODS

This retrospective cohort study identified 2,085 transgender individuals through data obtained from 4 outpatient community and hospital clinics in 3 large cities in Ontario, which were linked with administrative health data between January 2015 and December 2019. An age-matched 1:5 comparison cohort was created from the general population of Ontario. Outcome measures were analysed from March 2020 to May 2022. The primary outcome was mental health service utilization, which included mental health-related visits to primary care providers, psychiatrists, mental health- and self-harm-related emergency department visits, and mental health hospitalizations. Mediation variables included ethnic concentration, residential instability, dependency, and material deprivation at the neighbourhood level and were derived from the Ontario Marginalization Index.

RESULTS

This study identified 2,085 transgender individuals from participating outpatient community and hospital clinics, who were matched to the general population (n = 10,425). Overall, neighbourhood-level marginalization did not clinically mediate mental health service use. However, transgender individuals were more likely to be exposed to all forms of neighbourhood-level marginalization, as well as having higher rates of health service use across all outcome measures.

CONCLUSIONS

In this study, mental health service use among transgender individuals was not clinically mediated by marginalization at the neighbourhood level. This study highlights the need to explore marginalization and mental health service use at the individual level to better understand the mental health disparities experienced by transgender individuals and to ensure that health-care services are inclusive and affirming.

Triglycerides, Obesity and Education Status Are Associated with the Risk of Developing Type 2 Diabetes in Young Adults, Cohort Study

BACKGROUND

It is important to determine the influence of traditional risk factors on the development of type 2 diabetes mellitus (T2DM) in young adults. Goal of the research: To study the incidence of T2DM and factors that increase the risk of its occurrence during the observation of a cohort of young adults.

MATERIALS AND METHODS

1341 people aged 25-44 were included in the study from 2013 to 2017, of whom 622 were men (46.4%). The examination included anamnesis, anthropometric data, and a blood test. Cases of developed T2DM were identified by comparing the Diabetes Mellitus Register, medical records of patients, and the database of examined individuals from 2019 to 2023. T2DM Results: In the examined population, 11 participants (0.82%) developed T2DM. The prevalence of T2DM was 0.96% in men and 0.69% in women. Patients with T2DM had a higher waist circumference, BMI, SBP, TG, and lower HDL than patients without T2DM, and were also less likely to have a higher education. The risk of developing T2DM increases 6.5 times at a BMI of ≥30 kg/m2, and 5.2 times at a TG level of ≥1.7 mmol/L, regardless of other risk factors. In the absence of a higher education, the risk of developing T2DM is increased by 5.6 times.

CONCLUSION

In young people, high triglyceride levels, obesity, and a low level of education are associated with the risk of developing type 2 diabetes, regardless of other factors.

Obesity mediates the opposite association of education and diabetes in Chinese men and women: Results from the REACTION study

BACKGROUND

Evidence regarding the impact of education on diabetes risk is scarce in developing countries. We aimed to explore the association between education and diabetes within a large population in China and to identify the possible mediators between them.

METHODS

Information on educational level and lifestyle factors was collected through questionnaires. Diabetes was diagnosed from self-report and biochemical measurements. A structural equation model was constructed to quantify the mediation effect of each mediator.

RESULTS

Compared with their least educated counterparts, men with college education had a higher risk of diabetes (odds ratio [OR] 1.19; 95% confidence interval [CI], 1.12-1.27), while college-educated women were less likely to have diabetes (OR 0.77; 95% CI, 0.73-0.82). Obesity was the strongest mediator in both genders (proportion of mediation: 11.6% in men and 23.9% in women), and its association with education was positive in men (β[SE] 0.0387 [0.0037]) and negative in women (β[SE] -0.0824 [0.0030]). Taken together, all behavioral factors explained 12.4% of the excess risk of diabetes in men and 33.3% in women.

CONCLUSIONS

In a general Chinese population, the association between education level and diabetes was positive in men but negative in women. Obesity was the major mediator underlying the education disparities of diabetes risk, with a stronger mediation effect among women.

Understanding socioeconomic differences in incident metabolic syndrome among adults: What is the mediating role of health behaviours?

BACKGROUND

The incidence of metabolic syndrome (MetS) strongly varies by socioeconomic position (SEP), but little is known about the mediating role of health behaviours in this association. This study examines the associations between the SEP measures, education, income and occupational prestige, and incident MetS and whether the associations are mediated by health behaviours, including physical activity, smoking, alcohol intake and diet quality.

METHODS

A subsample (n = 85,910) of the adult Lifelines Cohort Study without MetS at baseline was used. MetS was measured at the second assessment (median follow-up time 3.8 years) defined according to the NCEP-ATPIII criteria. Direct associations between SEP, health behaviours and incident MetS were estimated using multivariable logistic regression analyses. The mediating percentages of health behaviours explaining the associations between SEP and incident MetS were estimated using the Karlson-Holm-Breen method. Analyses were independent of age, sex, the other SEP measures and follow-up time.

RESULTS

Education and occupational prestige were inversely associated with MetS. Income was not associated with MetS. Health behaviours explained only partly (13.8%) the association between education and MetS, with smoking as the strongest mediating factor (8.8%). Health behaviours played also a minor role (2.7%) in explaining occupational MetS differences, with physical activity as the strongest suppressing factor (-9.4%).

CONCLUSION

Individuals with more years of education or a higher occupational prestige had a lower risk to develop MetS. This was mainly because of non-smoking, less excessive alcohol intake and a higher diet quality. However, individuals with a higher SEP were more often physically inactive.

Prevalence, incidence and risk factors of diabetes in Australian adults aged ≥45 years: A cohort study using linked routinely-collected data

Aims

To use linked routinely-collected health data to estimate diabetes prevalence and incidence in an Australian cohort of adults aged ≥45 years, and examine risk factors associated with incident disease.

Research design and methods

The EXamining ouTcomEs in chroNic Disease in the 45 and Up Study (EXTEND45) Study is a linked data study that combines baseline questionnaire responses from the population-based 45 and Up Study (2006–2009, n = 267,153) with multiple routinely-collected health databases up to December 2014. Among participants with ≥1 linked result for any laboratory test, diabetes status was determined from multiple data sources according to standard biochemical criteria, use of glucose-lowering medication or self-report, and the prevalence and incidence rate calculated. Independent risk factors of incident diabetes were examined using multivariable Cox regression.

Results

Among 152,169 45 and Up Study participants with ≥1 linked laboratory result in the EXTEND45 database (mean age 63.0 years; 54.9% female), diabetes prevalence was 10.8% (95% confidence interval [CI] 10.6%–10.9%). Incident disease in those without diabetes at baseline (n = 135,810; mean age 62.5 years; 56.1% female) was 10.0 per 1,000 person-years (95% CI 9.8–10.2). In all age groups, diabetes incidence was lower in women compared to men, an association that persisted in the fully adjusted analyses. Other independent risk factors of diabetes were older age, being born outside of Australia (with the highest rate of 19.2 per 1,000 person-years observed in people born in South and Central Asia), lower education status, lower annual household income, residence in a major city, family history of diabetes, personal history of cardiovascular disease or hypertension, higher body mass index, smoking and long sleeping hours.

Conclusions

Our study represents an efficient approach to assessing diabetes frequency and its risk factors in the community. The infrastructure provided by the EXTEND45 Study will be useful for diabetes surveillance and examining other important clinical and epidemiological questions.

Education and incident type 2 diabetes: quantifying the impact of differential exposure and susceptibility to being overweight or obese

AIMS/HYPOTHESIS

Educational inequality in type 2 diabetes incidence is evident in many high-income countries. Previous studies have shown that differential exposure to being overweight/obese across educational groups may partly explain this inequality. Whether differential susceptibility to being overweight/obese across educational groups contributes to this inequality has been investigated less frequently, even though it is a plausible mechanism. The two mechanisms may even be highly intertwined. In this longitudinal cohort study, we investigated the simultaneous contribution of differential exposure and differential susceptibility to being overweight/obese to educational inequality in type 2 diabetes incidence.

METHODS

The study population comprised 53,159 Danish men and women aged 50-64 years at baseline who were followed for a mean of 14.7 years. We estimated rate differences of type 2 diabetes by education level per 100,000 person-years. Using counterfactual mediation analysis, these rate differences were decomposed into proportions attributable to differential exposure, differential susceptibility and all other pathways, respectively. We compared this approach with conventional approaches to mediation and interaction analysis.

RESULTS

Compared with a high level of education, a low education level was associated with 454 (95% CI 398, 510) additional cases of type 2 diabetes, and a medium education level with 316 (CI 268, 363) additional cases. Differential exposure to being overweight/obese accounted for 37% (CI 31%, 45%) of the additional cases among those with a low education level and 29% (CI 24%, 36%) of the additional cases among those with a medium education level. Differential susceptibility accounted for 9% (CI 4%, 14%) and 6% (CI 3%, 10%) of the additional cases among those with a low and medium education level, respectively. Compared with the counterfactual approach, the conventional approaches suggested stronger effects of both mechanisms.

CONCLUSIONS/INTERPRETATION

Differential exposure and susceptibility to being overweight/obese are both important mechanisms in the association between education and type 2 diabetes incidence.

How does socioeconomic development in Brazil shape social inequalities in diabetes?

Many countries, including Brazil, are facing growing social inequalities in diabetes prevalence. The different states in Brazil represent different levels of development and by comparing diabetes inequalities across states we aim to get a better understanding of how educational inequalities in diabetes are linked to social development. We use the latest cross-sectional national health survey of Brazil - PNS-2013 (N = 60,202) and analyse the disparities in diabetes as well as the differential exposure and susceptibility to the effect of obesity across states for men and women. Among women in high-HDI states the prevalence of diabetes is 11.7 percentage units (CI: 9.3; 14.0) higher among the lowest compared to the highest educated. In less-developed states the disparity is smaller. Among men, there is no social gradient found for diabetes, but obesity is positively associated with education. The association between obesity and diabetes is stronger among the low educated particularly for men in high-HDI states. Here the interaction effect between low education and obesity is 11.7 (CI 8.1; 15.4) percentage units. The fact that economic development is associated with increasingly unequal levels of diabetes and with unequal levels of exposure and susceptibility to obesity indicates that other interacting determinants are important for the development of the diabetes epidemic in Brazil.

Social inequality in lifestyle, motivation to change lifestyle and received health advice in individuals with diabetes: A nationwide study

Aims: The aim was to investigate the association between socioeconomic position (SEP) and physical activity, alcohol consumption and smoking, motivation to change lifestyle and health advices from general practitioners (GPs) in individuals with diabetes. Methods: Data were provided by the Danish National Health Survey 2013 and 7504 adults (⩾ 40 years) with diabetes were included. Educational level was used as SEP indicator and categorized into low, middle and high SEP. Dependent variables included physical activity, alcohol consumption, smoking, motivation to change lifestyle and GP lifestyle advices. Multiple logistic regression analyses adjusted for age, body mass index and ethnic background were performed. Results: Higher SEP were associated with reduced odds of being physically inactive (middle SEP odds ratio (OR) men 0.58 (95% confidence intervals 0.47–0.72) and women 0.59 (0.47–0.75)) and non-smoking (middle SEP OR men 0.74 (0.59–0.93) and high SEP OR women 0.54 (0.38–0.77)) compared to participants with a low SEP. Alcohol consumption above the recommended maximum was associated with high SEP in men, OR 1.83 (1.30–2.61). Elevated SEP was associated with a motivation to increase physical activity levels (middle SEP OR men 1.45 (1.19–1.76) and women 1.35 (1.09–1.67)), high SEP was associated with none advice from GPs regarding smoking cessation among women, OR 0.47 (0.25–0.89). Conclusions: Socioeconomic position was strongly associated with lifestyle in individuals with diabetes. The most pronounced inequalities were found in physical activity levels, smoking status and the motivation to become more physically active. Municipalities and GPs may need a greater focus on SEP in interventions to change lifestyle in individuals with diabetes.

Sex Differences in Diabetes Mellitus Mortality Trends in Brazil, 1980-2012

Aims

To investigate the hypothesis that the change from the female predominance of diabetes mellitus to a standard of equality or even male preponderance can already be observed in Brazilian mortality statistics.

Methods

Data on deaths for which diabetes mellitus was listed as the underlying cause were obtained from the Brazilian Mortality Information System for the years 1980 to 2012. The mortality data were also analyzed according to the multiple causes of death approach from 2001 to 2012. The population data came from the Brazilian Institute of Geography and Statistics. The mortality rates were standardized to the world population. We used a log-linear joinpoint regression to evaluate trends in age-standardized mortality rates (ASMR).

Results

From 1980 to 2012, we found a marked increment in the diabetes ASMR among Brazilian men and a less sharp increase in the rate among women, with the latter period (2003–2012) showing a slight decrease among women, though it was not statistically significant.

Conclusions

The results of this study suggest that diabetes mellitus in Brazil has changed from a pattern of higher mortality among women compared to men to equality or even male predominance.

Causal mediation analysis with a binary outcome and multiple continuous or ordinal mediators: Simulations and application to an alcohol intervention

We investigate a method to estimate the combined effect of multiple continuous/ordinal mediators on a binary outcome: 1) fit a structural equation model with probit link for the outcome and identity/probit link for continuous/ordinal mediators, 2) predict potential outcome probabilities, and 3) compute natural direct and indirect effects. Step 2 involves rescaling the latent continuous variable underlying the outcome to address residual mediator variance/covariance. We evaluate the estimation of risk-difference- and risk-ratio-based effects (RDs, RRs) using the ML, WLSMV and Bayes estimators in Mplus. Across most variations in path-coefficient and mediator-residual-correlation signs and strengths, and confounding situations investigated, the method performs well with all estimators, but favors ML/WLSMV for RDs with continuous mediators, and Bayes for RRs with ordinal mediators. Bayes outperforms WLSMV/ML regardless of mediator type when estimating RRs with small potential outcome probabilities and in two other special cases. An adolescent alcohol prevention study is used for illustration.

Risk factors of incident type 2-diabetes mellitus over a 3-year follow-up: Results from a large Australian sample

AIMS

To describe the incidence of type 2 diabetes mellitus (T2DM) among middle-aged and older Australian adults and to examine a broad range of risk factors of T2DM.

METHODS

A large cohort of Australian adults aged 45 and up was sampled from the general population and was followed up for approximately 3 years (n=60,404). Physician-diagnosed T2DM was self-reported at baseline (2006-2008) and follow-up (2010). Incident T2DM was determined as not reporting T2DM at baseline, but reporting T2DM at follow-up. A broad range of risk factors, including socio-demographic characteristics, health status, family history, and lifestyle behaviors were examined at baseline. Multiple logistic regression was used for selecting potential predictors of incident T2DM, and age and reported family history of T2DM were tested as potential effect modifiers.

RESULTS

Of the 54,997 without T2DM at baseline, 888 reported T2DM at follow-up (cumulative incidence 1.6% over 3.4 years, annual incidence rate 0.44%). Adjusted for other risk factors, being male, older age, higher relative socio-economic disadvantage, being born in Asia, lower educational attainment, medical history of hypertension and dyslipidemia, family history of T2DM, overweight/obese, smoking, long sleeping hours, and psychological distress were significantly associated with higher odds of developing T2DM. Particularly, hypertension, dyslipidemia, and overweight/obesity were stronger predictors of T2DM among middle-aged than older adults (≥60 years).

CONCLUSIONS

Understanding risk factors for incident T2DM could help identify at-risk populations and develop upstream preventive strategies to combat the epidemic of diabetes.

Racial residential segregation and preterm birth: built environment as a mediator

BACKGROUND

Racial residential segregation has been associated with preterm birth. Few studies have examined mediating pathways, in part because, with binary outcomes, indirect effects estimated from multiplicative models generally lack causal interpretation. We develop a method to estimate additive-scale natural direct and indirect effects from logistic regression. We then evaluate whether segregation operates through poor-quality built environment to affect preterm birth.

METHODS

To estimate natural direct and indirect effects, we derive risk differences from logistic regression coefficients. Birth records (2000-2008) for Durham, North Carolina, were linked to neighborhood-level measures of racial isolation and a composite construct of poor-quality built environment. We decomposed the total effect of racial isolation on preterm birth into direct and indirect effects.

RESULTS

The adjusted total effect of an interquartile increase in racial isolation on preterm birth was an extra 27 preterm events per 1000 births (risk difference = 0.027 [95% confidence interval = 0.007 to 0.047]). With poor-quality built environment held at the level it would take under isolation at the 25th percentile, the direct effect of an interquartile increase in isolation was 0.022 (-0.001 to 0.042). Poor-quality built environment accounted for 35% (11% to 65%) of the total effect.

CONCLUSION

Our methodology facilitates the estimation of additive-scale natural effects with binary outcomes. In this study, the total effect of racial segregation on preterm birth was partially mediated by poor-quality built environment.

Financial strain, inflammatory factors, and haemoglobin A1c levels in African American women

OBJECTIVE

Type 2 diabetes disproportionately affects African American women, a population exposed to high levels of stress, including financial strain (Centers for Disease Control & Prevention, 2011, http://www.cdc.gov/diabetes/pubs/pdf/ndfs_2011.pdf). We tested a mediational model in which chronic financial strain among African American women contributes to elevated serum inflammation markers, which, in turn, lead to increased haemoglobin A1C (HbA1c) levels and risk for type 2 diabetes.

METHODS

We assessed level of financial strain four times over a 10-year period and tested its effect on two serum inflammation markers, C-reactive protein (CRP) and soluble interleukin-6 receptor (sIL-6R) in year 11 of the study. We tested the inflammation markers as mediators in the association between chronic financial strain and HbA1c, an index of average blood glucose level over several months.

DESIGN

Data were from 312 non-diabetic African American women from the Family and Community Health Study (FACHS; Cutrona et al., 2000, J. Pers. Soc. Psychol., 79, 1088).

RESULTS

Chronic financial strain predicted circulating sIL-6R after controlling for age, BMI, health behaviours, and physical health measures. In turn, sIL-6R significantly predicted HbA1c levels. The path between chronic financial strain and HbA1c was significantly mediated by sIL-6R. Contrary to prediction, CRP was not predicted by chronic financial strain.

CONCLUSIONS

Results support the role of inflammatory factors in mediating the effects of psychosocial stressors on risk for type 2 diabetes. Findings have implications for interventions that boost economic security and foster effective coping as well as medical interventions that reduce serum inflammation to prevent the onset of type 2 diabetes.

Mediation misgivings: ambiguous clinical and public health interpretations of natural direct and indirect effects

Recent methodological innovation is giving rise to an increasing number of applied papers in medical and epidemiological journals in which natural direct and indirect effects are estimated. However, there is a longstanding debate on whether such effects are relevant targets of inference in population health. In light of the repeated calls for a more pragmatic and consequential epidemiology, we review three issues often raised in this debate: (i) the use of composite cross-world counterfactuals and the need for cross-world independence assumptions; (ii) interventional vs non-interventional identifiability; and (iii) the interpretational ambiguity of natural direct and indirect effect estimates. We use potential outcomes notation and directed acyclic graphs to explain 'cross-world' assumptions, illustrate implications of this assumption via regression models and discuss ensuing issues of interpretation. We argue that the debate on the relevance of natural direct and indirect effects rests on whether one takes as a target of inference the mathematical object per se, or the change in the world that the mathematical object represents. We further note that public health questions may be better served by estimating controlled direct effects.