• Citer

Gender specificity in the prediction of clinically diagnosed depression. Results of a large cohort of Belgian workers.

par Godin, Isabelle ;Kornitzer, Marcel ;Clumeck, Nicolas ;Linkowski, Paul ;Valente Soares, Filomena ;Kittel, France
Référence Social psychiatry and psychiatric epidemiology, 44, 7, page (592-600)
Publication Publié, 2009-07

Article révisé par les pairs

• Détails
• Contenu

Résumé :

BACKGROUND: The recent increase in the incidence of clinical depression represents a major public health and socio-economical burden. Depression has its roots in both professional and private domains but few epidemiological studies have looked at predictors of long term clinical depression as defined by a sick-leave of 28 days or more and a diagnosis by a general practitioner in both genders. OBJECTIVES: To study baseline predictors of long term spells of clinical depression within the framework of a large prospective study, the Belstress Study, in 6,659 men and 2,737 women aged 35-59 years at baseline survey. METHODS: Kaplan-Meyer survival curves and Cox regression models were used in order to relate long term clinical depression defined by a sick-leave of 28 days or more to baseline socio-demographic and work and non-work variables. RESULTS AND CONCLUSIONS: Density incidence of long term clinical depression is 0.5 years and 1.1/1,000 persons/months for men and women respectively. In univariate analyses specific gender predictors were observed as for men predictors besides level of education, were work related: high job-strain OR 1.67 (CI 95% 1.03; 2.71) and work dissatisfaction OR 1.78 (CI 95% 1.09; 2.91) whereas for women baseline predictors are related to private life dissatisfaction OR 1.84 (CI 95% 1.16; 2.91) and to a lesser degree low social support from co-workers OR 1.50 (CI 95% 0.93; 2.40). In both genders baseline severe depression symptoms defined by a CES-D score of percentile 90 or above is a predictor of long term sick-leave for clinical depression. In multivariate analyses, in a model without baseline CES-D high job-strain and job dissatisfaction remain independent predictors for incident clinical depression in men whereas only private life dissatisfaction remains a significant predictor in women. When added to the model CES-D is the most powerful predictor of clinical depression in both genders. Together with level of education, work dissatisfaction remains borderline significant in men whereas private life dissatisfaction remains an independent predictor for clinical depression in women. In men baseline symptoms of depression alleviate the impact of high job-strain on incident clinical depression whereas in women, private life dissatisfaction remains an independent predictor of clinical depression.