Trivellore E. Raghunathan

INcome inequality, the psychosocial environment, and health: comparisons of wealthy nations.

BACKGROUND The theory that income inequality and characteristics of the psychosocial environment (indexed by such things as social capital and sense of control over lifes circumstances) are key determinants of health and could account for health differences between countries has become influential in health inequalities research and for population health policy. METHODS We examined cross-sectional associations between income inequality and low birthweight, life expectancy, self-rated health, and age-specific and cause-specific mortality among countries providing data in wave III (around 1989-92) of the Luxembourg Income Study. We also used data from the 1990-91 wave of the World Values Survey (WVS). We obtained life expectancy, mortality, and low birthweight data from the WHO Statistical Information System. FINDINGS Among the countries studied, higher income inequality was strongly associated with greater infant mortality (r=0.69, p=0.004 for women; r=0.74, p=0.002 for men). Associations between income inequality and mortality declined with age at death, and then reversed among those aged 65 years and older. Income inequality was inconsistently associated with specific causes of death and was not associated with coronary heart disease (CHD), breast or prostate cancer, cirrhosis, or diabetes mortality. Countries that had greater trade union membership and political representation by women had better child mortality profiles. Differences between countries in levels of social capital showed generally weak and somewhat inconsistent associations with cause-specific and age-specific mortality. INTERPRETATION Income inequality and characteristics of the psychosocial environment like trust, control, and organisational membership do not seem to be key factors in understanding health differences between these wealthy countries. The associations that do exist are largely limited to child health outcomes and cirrhosis. Explanations for between-country differences in health will require an appreciation of the complex interactions of history, culture, politics, economics, and the status of women and ethnic minorities.

Family History as a Risk Factor for Primary Cardiac Arrest

BACKGROUND The hypothesis that a family history of myocardial infarction (MI) or primary cardiac arrest (PCA) is an independent risk factor for primary cardiac arrest was examined in a population-based case-control study. In addition, we investigated whether recognized risk factors account for the familial aggregation of these cardiovascular events. METHODS AND RESULTS PCA cases, 25 to 74 years old, attended by paramedics during the period 1988 to 1994 and population-based control subjects matched for age and sex were identified from the community by random digit dialing. All subjects were free of recognized clinical heart disease and major comorbidity. A detailed history of MI and PCA in first-degree relatives was collected in interviews with the spouses of case and control subjects by trained interviewers using a standardized questionnaire. For each familial relationship, there was a higher rate of MI or primary cardiac arrest (MI/PCA) in relatives of case compared with relatives of control subjects. Overall, the rate of MI/PCA among first-degree relatives of cardiac arrest patients was almost 50% higher than that in first-degree relatives of control subjects (rate ratio [RR]=1.46; 95% CI=1.23 to 1.72). In a multivariate logistic model, family history of MI/PCA was associated with PCA (RR=1.57; 95% CI=1.27 to 1.95) even after adjustment for other common risk factors. CONCLUSIONS Family history of MI or PCA is positively associated with the risk of primary cardiac arrest. This association is mostly independent of familial aggregation of other common risk factors.

Myocardial Infarction in Young Women in Relation to Plasma Total Homocysteine, Folate, and a Common Variant in the Methylenetetrahydrofolate Reductase Gene

BACKGROUND In a population-based study, we examined the relationship between the risk of myocardial infarction (MI) among young women and plasma total homocysteine (tHCY), folate, vitamin B12, and a common cytosine (C) to thymine (T) polymorphism in the gene for 5,10-methylenetetrahydrofolate reductase (MTHFR). METHODS AND RESULTS In-person interviews and nonfasting blood samples were obtained from 79 women < 45 years old diagnosed with MI and 386 demographically similar control subjects living in western Washington state between 1991 and 1995. Compared with control subjects, case patients had higher mean tHCY concentrations (13.4+/-5.2 versus 11.1+/-4.4 micromol/L, P=.0004) and lower mean folate concentrations (12.4+/-13.4 versus 16.1+/-12.2 nmol/L, P=.018). There was no difference in vitamin B12 concentrations between case patients and control subjects (346.8+/-188.4 versus 349.7+/-132.4 pmol/L, P=.90). After adjusting for cardiovascular risk factors, we found that women with tHCY > or = 15.6 micromol/L were at approximately twice the risk of MI as women with tHCY < 10.0 micromol/L (OR, 2.3; 95% CI, 0.94 to 5.64). Women with folate > or = 8.39 nmol/L had an approximately 50% lower risk of MI than women with folate < 5.27 nmol/L (OR, 0.54; 95% CI, 0.23 to 1.28). There was no association with vitamin B12 concentration. Among control subjects, 12.7% were homozygous for the MTHFR T677 allele, and these women had higher plasma tHCY and lower plasma folate than women with other genotypes. Ten percent of case patients were homozygous for the T677 allele, and there was no association of homozygosity for T677 with MI risk (OR, 0.90; 95% CI, 0.31 to 2.29). CONCLUSIONS These data support the hypothesis that elevated plasma tHCY and low plasma folate are risk factors for MI among young women. Although homozygosity for MTHFR T677 is related to increased plasma tHCY and low plasma folate, this genetic characteristic is not a risk factor for MI in this population.

Large-Sample Significance Levels from Multiply Imputed Data Using Moment-Based Statistics and an F Reference Distribution

Abstract We present a procedure for computing significance levels from data sets whose missing values have been multiply imputed data. This procedure uses moment-based statistics, m ≤ 3 repeated imputations, and an F reference distribution. When m = ∞, we show first that our procedure is essentially the same as the ideal procedure in cases of practical importance and, second, that its deviations from the ideal are basically a function of the coefficient of variation of the canonical ratios of complete to observed information. For small m our procedures performance is largely governed by this coefficient of variation and the mean of these ratios. Using simulation techniques with small m, we compare our procedures actual and nominal large-sample significance levels and conclude that it is essentially calibrated and thus represents a definite improvement over previously available procedures. Furthermore, we compare the large-sample power of the procedure as a function of m and other factors, such as the di...

Neighborhood characteristics and hypertension.

Background: The goal of this study was to investigate cross-sectional associations between features of neighborhoods and hypertension and to examine the sensitivity of results to various methods of estimating neighborhood conditions. Methods: We used data from the Multi-Ethnic Study of Atherosclerosis on 2612 individuals 45–85 years of age. Hypertension was defined as systolic blood pressure above 140 mm Hg, diastolic pressure above 90 mm Hg, or use of antihypertensive medications. Neighborhood (census tract) conditions potentially related to hypertension (walking environment, availability of healthy foods, safety, social cohesion) were measured using information from a separate phone survey conducted in the study neighborhoods. For each neighborhood we estimated scale scores by aggregating residents’ responses using simple aggregation (crude means) and empirical Bayes estimation (unconditional, conditional, and spatial). These estimates of neighborhood conditions were linked to each study participant based on the census tract of residence. Two-level binomial regression methods were used to estimate adjusted associations between neighborhood conditions and hypertension. Results: Residents of neighborhoods with better walkability, availability of healthy foods, greater safety, and more social cohesion were less likely to be hypertensive (relative prevalence [95% confidence interval] for 90th vs. 10th percentile of conditional empirical Bayes estimate = 0.75 [0.64–0.88], 0.72 [0.61–0.85], 0.74 [0.63–0.86], and 0.69 [0.57–0.83]), respectively, after adjusting for site, age, sex, income, and education. Associations were attenuated and often disappeared after additional adjustments for race/ethnicity. Conclusion: Neighborhood walkability, food availability, safety, and social cohesion may be mechanisms that link neighborhoods to hypertension.

Stroke and Use of Low-Dose Oral Contraceptives in Young Women A Pooled Analysis of Two US Studies

BACKGROUND AND PURPOSE The available data on low-dose oral contraceptive pill (OCP) use and stroke risk in US women are limited by small numbers. We sought more precise estimates by conducting a pooled analysis of data from 2 US population-based case-control studies. METHODS We analyzed interview data from 175 ischemic stroke cases, 198 hemorrhagic stroke cases, and 1191 control subjects 18 to 44 years of age. RESULTS For ischemic stroke, the pooled odds ratio (pOR) adjusted for stroke risk factors for current use of low-dose OCPs compared with women who had never used OCP (never users) was 0.66 (95% confidence interval [CI], 0.29 to 1.47) and compared with women not currently using OCPs (nonusers) the pOR was 1.09 (95% CI, 0.54 to 2.21). For hemorrhagic stroke, the pOR for current use of low-dose OCPs compared with never users was 0.95 (95% CI, 0.46 to 1.93) and compared with nonusers the pOR was 1.11 (95% CI, 0.61 to 2.01). The pORs for current low-dose OCP use and either stroke type were not elevated among women who were >/=35 years, cigarette smokers, obese, or not receiving medical therapy for hypertension. pORs for current low-dose OCP use were 2.08 (95% CI, 1. 19 to 3.65) for ischemic stroke and 2.15 (95% CI, 0.85 to 5.45) for hemorrhagic stroke among women reporting a history of migraine but were not elevated among women without such a history. Past OCP use (irrespective of formulation) was inversely related to ischemic stroke but unrelated to hemorrhagic stroke. CONCLUSIONS Women who use low-dose OCPs are, in the aggregate, not at increased risk of stroke. Studies are needed to clarify the risk of stroke among users who may be susceptible on the basis of age, smoking, obesity, hypertension, or migraine history.

The Multiple Adaptations of Multiple Imputation

Multiple imputation was first conceived as a tool that statistical agencies could use to handle nonresponse in large-sample public use surveys. In the last two decades, the multiple-imputation framework has been adapted for other statistical contexts. For example, individual researchers use multiple imputation to handle missing data in small samples, statistical agencies disseminate multiply-imputed data sets for purposes of protecting data confidentiality, and survey methodologists and epidemiologists use multiple imputation to correct for measurement errors. In some of these settings, Rubins original rules for combining the point and variance estimates from the multiply-imputed data sets are not appropriate, because what is known—and thus the conditional expectations and variances used to derive inferential methods—differs from that in the missing-data context. These applications require new combining rules and methods of inference. In fact, more than 10 combining rules exist in the published literature. This article describes some of the main adaptations of the multiple-imputation framework, including missing data in large and small samples, data confidentiality, and measurement error. It reviews the combining rules for each setting and explains why they differ. Finally, it highlights research topics in extending the multiple-imputation framework.

Multiple Imputation of Missing Income Data in the National Health Interview Survey

The National Health Interview Survey (NHIS) provides a rich source of data for studying relationships between income and health and for monitoring health and health care for persons at different income levels. However, the nonresponse rates are high for two key items, total family income in the previous calendar year and personal earnings from employment in the previous calendar year. To handle the missing data on family income and personal earnings in the NHIS, multiple imputation of these items, along with employment status and ratio of family income to the federal poverty threshold (derived from the imputed values of family income), has been performed for the survey years 1997–2004. (There are plans to continue this work for years beyond 2004 as well.) Files of the imputed values, as well as documentation, are available at the NHIS website (http://www.cdc.gov/nchs/nhis.htm). This article describes the approach used in the multiple-imputation project and evaluates the methods through analyses of the multiply imputed data. The analyses suggest that imputation corrects for biases that occur in estimates based on the data without imputation, and that multiple imputation results in gains in efficiency as well.

Long-term Exposure to Ambient Particulate Matter and Prevalence of Subclinical Atherosclerosis in the Multi-Ethnic Study of Atherosclerosis

Exposure to airborne particulate matter has been linked to cardiovascular events. Whether this finding reflects an effect of particulate matter exposure on the triggering of events or development of atherosclerosis remains unknown. Using data from the Multi-Ethnic Study of Atherosclerosis collected at baseline (2000-2002), the authors investigated associations of 20-year exposures to particulate matter with measures of subclinical disease (coronary calcium, common carotid intimal-medial thickness, and ankle-brachial index) in 5,172 US adults without clinical cardiovascular disease. Particulate matter exposures for the 20 years prior to assessment of subclinical disease were obtained from a space-time model of Environmental Protection Agency monitor data linked to residential history data for each participant. Intimal-medial thickness was weakly, positively associated with exposures to particulate matter <10 microm in aerodynamic diameter and <2.5 microm in aerodynamic diameter after controlling for age, sex, race/ethnicity, socioeconomic factors, diet, smoking, physical activity, blood lipids, diabetes, hypertension, and body mass index (1-4% increase per 21-microg/m(3) increase in particulate matter <10 microm in aerodynamic diameter or a 12.5-microg/m(3) increase in particulate matter <2.5 microm in aerodynamic diameter). No consistent associations with other measures of atherosclerosis were observed. There was no evidence of effect modification by sociodemographic factors, lipid status, smoking, diabetes, body mass index, or site. Results are compatible with some effect of particulate matter exposures on development of carotid atherosclerosis.

Race/ethnicity, life-course socioeconomic position, and body weight trajectories over 34 years : The alameda county study

OBJECTIVES We investigated whether race differences in weight gain over 34 years were because of socioeconomic position (SEP) and psychosocial and behavioral factors (physical activity, cigarette smoking, alcohol consumption, depression, marital status, number of children). We used a life-course approach to SEP with 4 measures of SEP (childhood SEP, education, occupation, income) and a cumulative measure of SEP. METHODS We used mixed models and data collected from the Alameda County Study to examine the association between race and weight change slopes and baseline weight in men (n=1186) and women (n=1375) aged 17 to 40 years at baseline (in 1965). RESULTS All subjects gained weight over time. African American women weighed 4.96 kg (P < .001) more at baseline and gained 0.10 kg/year (P = .043) more weight than White women. Black men weighed 2.41 kg (P = .006) more at baseline but did not gain more weight than White men. The association of race with weight gain in women was largely because of cumulative SEP score. CONCLUSIONS Interventions to prevent overweight and obesity should begin early in life and target the socioeconomically disadvantaged.