The association between smoking and psychopathology adjusted for body mass index and gender.

OBJECTIVE: This study examined the correlation between smoking habits and psychopathology status, as well as the impact of confounders such as body mass index and gender. METHOD: A total of 134 non-smokers and 152 smokers were enrolled in this study. We measured psychopathology features using Symptom Checklist 90-Revised. We ran logistic regression models testing the smoking-psychopathology association, controlling for body mass index and gender. RESULTS: Smoking was positively correlated with depression, interpersonal sensitivity, hostility, somatization, paranoid ideation and psychoticism (P<0.05). Adjusting for body mass index and gender, the results remained largely unchanged, with a slight independent effect of body mass index. CONCLUSIONS: Our data suggest that smoking is a stronger predictor of psychopathology than body mass index and gender.

The Association Between Obesity and Hostility: The Mediating Role of Plasma Lipids.

OBJECTIVE: Recent research indicates an association between obesity and psychopathology status, the nature of which remains unclear. We evaluated the mediating role of biochemical disturbances in this association among a treatment-seeking sample of obese individuals. METHOD: The study enrolled 143 consecutive overweight and obese individuals (mean age 35±9 y) and 143 normal-weight controls (mean age 34±9 y), matched by age and sex. We measured psychopathology features using the Symptom Checklist 90-Revised (SCL-90-R), a standardized self-evaluation rating scale, and biochemical parameters (plasma cholesterol, triglyceride, and fasting glucose levels) of all participants. Nonlinear regression models were used to estimate the associations among obesity, psychopathology, and biochemical factors. RESULTS: Obesity was associated positively and significantly (P<0.05) with all of the SCL-90-R subscales, with the exception of anxiety and phobic anxiety, as well as with levels of plasma glucose, cholesterol (P<0.01), and triglycerides (P<0.001). Tests for mediation showed that obesity was significantly associated, for the mediators of plasma cholesterol [parameter estimate=-0.033, P<0.05] and triglycerides (parameter estimate=-0.059, P<0.05), only with hostility (parameter estimate=-0.024, P<0.05 and parameter estimate=-0.041, P<0.05, respectively). CONCLUSIONS: Our data suggest that biological substrates that are critically related to obesity, such as dyslipidemia, may mediate, at least in part, the association between obesity and hostility.