Person:

Subramanian, Sankaran

Objectives: India bears a significant portion of the global tobacco burden with high prevalence of tobacco use. This study examines the socioeconomic patterning of tobacco use and identifies the changing gender and socioeconomic dynamics in light of the Cigarette Epidemic Model. Design: Secondary analyses of second and third National Family Health Survey (NFHS) data. Setting and participants: Data were analysed from 201 219 men and 255 028 women over two survey rounds. Outcomes and methods: Outcomes included smoking (cigarettes, bidis and pipes/cigar), chewed tobacco (paan masala, gutkha and others) and dual use, examined by education, wealth, living environment and caste. Standardised prevalence and percentage change were estimated. Pooled multilevel models estimated the effect of socioeconomic covariates on the log odds of tobacco use by gender, along with fixed and random parameters. Findings: Among men (2005−2006), gradients in smoking by education (illiterates: 44% vs postgraduates: 15%) and chewing (illiterates: 47% vs postgraduates: 19%) were observed. Inverse gradients were also observed by wealth, living environment and caste. Chewed tobacco use by women showed inverse socioeconomic status (SES) gradients comparing the illiterates (7.4%) versus postgraduates (0.33%), and poorest (17%) versus richest (2%) quintiles. However, proportional increases in smoking were higher among more educated (postgraduates (98%) vs high schooling only (17%)) and chewing among richer (richest quintile (49%) vs poorest quintile (35%)). Among women, higher educated showed larger declines for smoking—90% (postgraduates) versus 12% (illiterates). Younger men (15–24 years) showed increasing tobacco use (smoking: 123% and chewing: 112%). Older women (35–49 years) show higher prevalence of smoking (3.2%) compared to younger women (0.3%). Conclusions: Indian tobacco use patterns show significant diversions from the Cigarette Epidemic Model—from gender and socioeconomic perspectives. Separate analysis by type is needed to further understand social determinants of tobacco use in India.

Background: Poor indoor air quality (IAQ) and psychosocial problems are common in schools worldwide, yet longitudinal research on the issue is scarce. We examined whether the level of or a change in pupil-reported school environment (IAQ, school satisfaction, and bullying) predicts recorded sick leaves among teachers. Methods: Changes in the school environment were assessed using pupil surveys at two time points (2001/02 and 2004/05) in 92 secondary schools in Finland. Variables indicating change were based on median values at baseline. We linked these data to individual-level records of teachers’ (n = 1678) sick leaves in 2001–02 and in 2004–05. Results: Multilevel multinomial logistic regression models adjusted for baseline sick leave and covariates showed a decreased risk for short-term (one to three days) sick leaves among teachers working in schools with good perceived IAQ at both times (OR = 0.6, 95% CI: 0.5-0.9), and for those with a positive change in IAQ (OR = 0.6, 95% CI: 0.4-0.9), compared to teachers in schools where IAQ was constantly poor. Negative changes in pupil school satisfaction (OR = 1.8, 95% CI: 1.1-2.8) and bullying (OR = 1.5, 95% CI: 1.0-2.3) increased the risk for short-term leaves among teachers when compared to teachers in schools where the level of satisfaction and bullying had remained stable. School environment factors were not associated with long-term sick leaves. Conclusions: Good and improved IAQ are associated with decreased teacher absenteeism. While pupil-related psychosocial factors also contribute to sick leaves, no effect modification or mediation of psychosocial factors on the association between IAQ and sick leave was observed.

Background: The unmet needs for health care have been used as an alternative measurement to monitor equity in health services. We sought to examine contextual influences on unmet needs for health care whereas precedent studies have been focused on individual characteristics on them. Methods and Findings: The current study conducted multilevel logistic regression analysis to assess the effects of individual- and contextual-level predictors in meeting individual health care needs in South Korea. We sampled 7,200 individuals over the age of 19 in the Fourth Korea National Health and Nutrition Examination Survey in 2009. Included in the regression model were individual predictors such as demographic variables, socio-economic status, and self-rated health; the density of beds and physicians in public and private sectors within different regions were used as contextual-level predictors. This study showed the inverse association between unmet needs and regional resources in private sectors after controlling for the effects of individual-level predictors. Conclusion: Our findings suggest that increasing regional resources in private sectors might produce inefficiency in the health care system and inequity in access to health services, particularly where the competition in private health care sectors was highly stimulated under the fee-for-service reimbursement scheme. Policies for the reallocation of health care resources and for reduction of individual health care costs are needed in Korea.

Objective: We examined body mass index (BMI) across place and time to determine the pattern of BMI mean and standard deviation trajectories. Methods: We included participants in the Framingham Heart Study (FHS) Offspring Cohort over eight waves of follow-up, from 1971 to 2008. After exclusions, the final sample size was 4569 subjects with 28,625 observations. We used multi-level models to examine population means and variation at the individual and neighborhood (census tracts) levels across time with measured BMI as the outcome, controlling for individual demographics and behaviors and neighborhood poverty. Because neighborhoods accounted for limited BMI variance, we removed this level as a source of variation in final models. We examined sex-stratified models with all subjects and models stratified by sex and baseline weight classification. Results: Mean BMI increased from 24.0 kg/m2 at Wave 1 to 27.7 at Wave 8 for women and from 26.6 kg/m2 to 29.0 for men. In final models, BMI variation also increased from Waves 1 to 8, with the standard deviation increasing from 4.18 kg/m2 to 6.15 for women and 3.31 kg/m2 to 4.73 for men. BMI means increased in parallel across most baseline BMI weight classifications, except for more rapid increases through middle-age for obese women followed by declines in the last wave. BMI standard deviations also increased in parallel across baseline BMI classifications for women, with greater divergence of BMI variance for obese men compared to other weight classifications. Conclusion: Over nearly 40 years, BMI mean and variation increased in parallel across most baseline weight classifications in our sample. Individual-level characteristics, especially baseline BMI, were the primary factors in rising BMI. These findings have important implications not only for understanding the sources of the obesity epidemic in the United States but also for the targeting of interventions to address the epidemic.

Background: In recent years, South Korea has witnessed a sustained rise in the prevalence of adolescent depression. In the present study, we sought to investigate family and school environmental influences on adolescent depression. Methods and Findings Middle and high school students (N = 75,066) were randomly selected respondents to a web-based survey and answered questions on their academic and socioeconomic backgrounds, parental support, parental education level, physical activities, lifestyle habits and their experience of depression in the past one year. Two-level multilevel analysis was used to investigate the relationship between depression and individual (level 1) and school (level 2) factors. Girls reported having experienced depression in greater numbers than boys (43.96% vs. 32.03%). A significant association was found between adolescent depression experience and gender, grade, self-rated academic achievement, family affluence scale, parental support, parental education level, lifestyle habits, physical activity and sleep dissatisfaction. The students living with rich parents were more likely to be depressive, and maternal higher education was significantly associated with higher probability of boys’ depression experience. Low academic achievement was highly associated with the experience of depression. In school level contexts, girls were found to be less likely to be depressive in girls-only schools. Conclusion: The adolescent depression experience is not only an individual phenomenon but is highly associated with other factors such as parents, peers, academic achievement, and even gender mix in the school. Thus, prevention measures on youth depression need to focus on emphasizing less pressure from parents on academic performance, and establishing healthy inter-gender relationships within co-education schools.

Objective: We examined the extent to which differences in hospital-level cesarean delivery rates in Massachusetts were attributable to hospital-level, rather than maternal, characteristics. Methods: Birth certificate and maternal in-patient hospital discharge records for 2004–06 in Massachusetts were linked. The study population was nulliparous, term, singleton, and vertex births (NTSV) (n = 80,371) in 49 hospitals. Covariates included mother's age, race/ethnicity, education, infant birth weight, gestational age, labor induction (yes/no), hospital shift at time of birth, and preexisting health conditions. We estimated multilevel logistic regression models to assess the likelihood of a cesarean delivery. Results: Overall, among women with NTSV births, 26.5% births were cesarean, with a range of 14% to 38.3% across hospitals. In unadjusted models, the between-hospital variance was 0.103 (SE 0.022); adjusting for demographic, socioeconomic and preexisting medical conditions did not reduce any hospital-level variation 0.108 (SE 0.023). Conclusion: Even after adjusting for both socio-demographic and clinical factors, the chance of a cesarean delivery for NTSV pregnancies varied according to hospital, suggesting the importance of hospital practices and culture in determining a hospital's cesarean rate.

Background: Multilevel analyses are ideally suited to assess the effects of ecological (higher level) and individual (lower level) exposure variables simultaneously. In applying such analyses to measures of ecologies in epidemiological studies, individual variables are usually aggregated into the higher level unit. Typically, the aggregated measure includes responses of every individual belonging to that group (i.e. it constitutes a self-included measure). More recently, researchers have developed an aggregate measure which excludes the response of the individual to whom the aggregate measure is linked (i.e. a self-excluded measure). In this study, we clarify the substantive and technical properties of these two measures when they are used as exposures in multilevel models. Methods: Although the differences between the two aggregated measures are mathematically subtle, distinguishing between them is important in terms of the specific scientific questions to be addressed. We then show how these measures can be used in two distinct types of multilevel models—self-included model and self-excluded model—and interpret the parameters in each model by imposing hypothetical interventions. The concept is tested on empirical data of workplace social capital and employees' systolic blood pressure. Results: Researchers assume group-level interventions when using a self-included model, and individual-level interventions when using a self-excluded model. Analytical re-parameterizations of these two models highlight their differences in parameter interpretation. Cluster-mean centered self-included models enable researchers to decompose the collective effect into its within- and between-group components. The benefit of cluster-mean centering procedure is further discussed in terms of hypothetical interventions. Conclusions: When investigating the potential roles of aggregated variables, researchers should carefully explore which type of model—self-included or self-excluded—is suitable for a given situation, particularly when group sizes are relatively small.

Objective: To describe the socioeconomic and geographic distribution of smoking behaviour in Canada among 19,383 individuals (51% women) aged 15–85 years. Methods: Current smoking and quitting were modeled using standard and multilevel logistic regression. Markers of socioeconomic status (SES) were education and occupation. Geography was defined by Canadian Provinces. Results: The adjusted prevalence of current smoking was 20.2% (95% confidence interval [CI]: 18.8–21.7) and 63.7% (95% CI: 61.1–66.3) of ever smokers had quit. Current smoking decreased and quitting increased with increasing SES. The adjusted prevalence of current smoking was 32.8% (95% CI: 28.4–37.5) among the least educated compared to 11.0% (95% CI: 8.9–13.4) for the highest educated. Among the least educated, 53.0% (95% CI: 46.8–59.2) had quit, rising to 68.7% (95% CI: 62.7–74.1) for the most educated. There was substantial variation in current smoking and quitting at the provincial level; current smoking varied from 17.9% in British Columbia to 26.1% in Nova Scotia, and quitting varied from 57.4% in Nova Scotia to 67.8% in Prince Edward Island. Nationally, increasing education and occupation level were inversely associated with current smoking (odds ratio [OR] 0.64, 95% CI: 0.60–0.68 for education; OR 0.82, 95% CI: 0.77–0.87 for occupation) and positively associated with quitting (OR 1.27, 95% CI: 1.16–1.40 for education; OR 1.20, 95% CI: 1.12–1.27 for occupation). These associations were consistent in direction across provinces although with some variability in magnitude. Conclusion: Our findings indicate that socioeconomic inequalities in smoking have persisted in Canada; current smoking was less likely and quitting was more likely among the better off groups and in certain provinces. Current prevention and cessation policies have not been successful in improving the situation for all areas and groups. Future efforts to reduce smoking uptake and increase cessation in Canada will need consideration of socioeconomic and geographic factors to be successful.

Background: Public health research has turned towards examining upstream, community-level determinants of cardiovascular disease risk factors. Objective measures of the environment, such as those derived from direct observation, and perception-based measures by residents have both been associated with health behaviours. However, current methods are generally limited to objective measures, often derived from administrative data, and few instruments have been evaluated for use in rural areas or in low-income countries. We evaluate the reliability of a quantitative tool designed to capture perceptions of community tobacco, nutrition, and social environments obtained from interviews with residents in communities in 5 countries. Methodology/Principal Findings: Thirteen measures of the community environment were developed from responses to questionnaire items from 2,360 individuals residing in 84 urban and rural communities in 5 countries (China, India, Brazil, Colombia, and Canada) in the Environmental Profile of a Community’s Health (EPOCH) study. Reliability and other properties of the community-level measures were assessed using multilevel models. High reliability (>0.80) was demonstrated for all community-level measures at the mean number of survey respondents per community (n = 28 respondents). Questionnaire items included in each scale were found to represent a common latent factor at the community level in multilevel factor analysis models. Conclusions/ Significance: Reliable measures which represent aspects of communities potentially related to cardiovascular disease (CVD)/risk factors can be obtained using feasible sample sizes. The EPOCH instrument is suitable for use in different settings to explore upstream determinants of CVD/risk factors.

Objectives: To quantify the association between socioeconomic status (SES) and type 2 diabetes in India. Design: Nationally representative cross-sectional household survey. Setting: Urban and rural areas across 29 states in India. Participants: 168 135 survey respondents aged 18–49 years (women) and 18–54 years (men). Primary outcome measure: Self-reported diabetes status. Results: Markers of SES were social caste, household wealth and education. The overall prevalence of self-reported diabetes was 1.5%; this increased to 1.9% and 2.5% for those with the highest levels of education and household wealth, respectively. In multilevel logistic regression models (adjusted for age, gender, religion, marital status and place of residence), education (OR 1.87 for higher education vs no education) and household wealth (OR 4.04 for richest quintile vs poorest) were positively related to self-reported diabetes (p<0.0001). In a fully adjusted model including all socioeconomic variables and body mass index, household wealth emerged as positive and statistically significant with an OR for self-reported diabetes of 2.58 (95% credible interval (CrI): 1.99 to 3.40) for the richest quintile of household wealth versus the poorest. Nationally in India, a one-quintile increase in household wealth was associated with an OR of 1.31 (95% CrI 1.20 to 1.42) for self-reported diabetes. This association was consistent across states with the relationship found to be positive in 97% of states (28 of 29) and statistically significant in 69% (20 of 29 states). Conclusions: The authors found that the highest SES groups in India appear to be at greatest risk for type 2 diabetes. This raises important policy implications for addressing the disease burdens among the poor versus those among the non-poor in the context of India, where >40% of the population is living in poverty.