Person:

Buka, Stephen

OBJECTIVE: The aim of this study was to evaluate the association between maternal smoking during pregnancy (MSP) and lifetime risk for alcohol use disorder (AUD) and to explore possible mechanisms through which MSP may be related to neurobehavioral conditions during infancy and childhood, which could, in turn, lead to increased risk for AUD. METHOD: A sample of 1,625 individuals was followed from pregnancy for more than 40 years. Capitalizing on the long follow-up time, we used survival analysis to examine lifetime risks of AUD (diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition) in relation to levels of MSP (none, <20 cigarettes/day, and ≥20 cigarettes/day). We then used structural equation modeling to test hypotheses regarding potential mechanisms, including lower birth weight, neurological abnormalities, poorer academic functioning, and behavioral dysregulation. RESULTS: Relative to unexposed offspring, offspring of mothers who smoked 20 cigarettes per day or more exhibited greater risks for AUD (hazard ratio = 1.31, 95% CI [1.08, 1.59]). However, no differences were observed among offspring exposed to fewer than 20 cigarettes per day. In structural equation models, MSP was associated with neurobehavioral problems during infancy and childhood, which, in turn, were associated with an increased risk for adult AUD. CONCLUSIONS: MSP was associated with an increased lifetime risk for AUD. Adverse consequences were evident from birth to adulthood. A two-pronged remedial intervention targeted at both the mother (to reduce smoking during pregnancy) and child (to improve academic functioning) may reduce the risk for subsequent AUD.

OBJECTIVE

Education is inversely associated with coronary heart disease (CHD) risk, however the mechanisms are poorly understood. The study objectives were to evaluate the extent to which rarely measured factors (literacy, time preference, sense of control) and more commonly measured factors (income, depressive symptomatology, body mass index) in the education-CHD literature explain the associations between education and CHD risk.

METHOD

The study sample included 346 participants, aged 38–47 years (59.5% women), of the New England Family Study birth cohort. Ten-year CHD risk was calculated using the validated Framingham risk algorithm that utilizes diabetes, smoking, blood pressure, total cholesterol, HDL cholesterol, age and gender. Multivariable regression and mediation analyses were performed.

RESULTS

Regression analyses adjusting for age, race/ethnicity and childhood confounders (e.g. parental socioeconomic status, intelligence) demonstrated that relative to those with ≥college education, men and women with <high school had 73.7% (95% confidence interval (CI): 29.5, 133.0) and 48.2% (95% CI: 17.5, 86.8) higher 10-year CHD risk, respectively. Mediation analyses demonstrated significant indirect effects for reading comprehension in women (7.2%; 95% CI: 0.7, 19.4) and men (7.2%; 95% CI: 0.8, 19.1), and depressive symptoms (11.8%; 95% CI: 2.5, 26.6) and perceived constraint (6.7%, 95% CI: 0.7, 19.1) in women.

CONCLUSIONS

Evidence suggested that reading comprehension in women and men, and depressive symptoms and perceived constraint in women, may mediate some of the association between education and CHD risk. If these mediated effects are interpreted causally, interventions targeting reading, depressive symptoms, and perceived constraint could reduce educational inequalities in CHD.