Hélène Vulser1, Emmanuel Wiernik, Eric Tartour, Frédérique Thomas, Bruno Pannier, Sébastien Czernichow, Olivier Hanon, Tabassome Simon, Jean-Marc Simon, Cyril Ducolombier, Silla M Consoli, Nicolas Danchin, Frédéric Limosin, Cédric Lemogne. 1. From the AP-HP, Hôpitaux Universitaires Paris Ouest, Service de Psychiatrie de l'adulte et du sujet âgé (Vulser, Consoli, Limosin, Lemogne), Paris, France; Inserm, Centre for Research in Epidemiology and Population Health (Wiernik), Villejuif, France; Université Versailles St-Quentin (Wiernik, Czernichow), Versailles, France; Université Paris Descartes, Sorbonne Paris Cité, Faculté de médecine (Tartour, Hanon, Consoli, Danchin, Limosin, Lemogne), Paris, France; AP-HP, Hôpitaux Universitaires Paris Ouest, Service d'Immunologie Biologique (Tartour), Paris, France; IPC Center, Research Department (Thomas, Pannier, Ducolombier), Paris, France; Cardiology Department, Hopital Manhes (Pannier), Fleury-Mérogis, France; Inserm, UMS 11, Population-based Epidemiologic Cohorts (Czernichow), Villejuif, France; AP-HP, Hôpital Ambroise Paré, Unité de Nutrition (Czernichow), Boulogne-Billancourt, France; AP-HP, Hôpital Broca, Service de Gériatrie (Hanon), Paris, France; Université Pierre et Marie Curie (UPMC) (T. Simon), Paris, France; AP-HP, Department of Pharmacology, Hôpital St Antoine (T. Simon), Paris, France; AP-HP, Groupe Hospitalier Pitié-Salpêtrière, Service d'Oncologie Radiothérapique (J.-M. Simon), Paris, France; Inserm U894, Centre Psychiatrie et Neurosciences (Limosin, Lemogne), Paris, France; and AP-HP, Hôpitaux Universitaires Paris Ouest, Service de Cardiologie (Danchin), Paris, France.
Abstract
OBJECTIVES: Depressive symptoms have been associated with chronic low-grade inflammation, including elevated neutrophil count. Smokers often have both high neutrophil count and depressive symptoms. Thus, smoking could explain the cross-sectional association between depressive symptoms and neutrophil count. METHODS: Total white blood cell count and subtypes, including absolute neutrophil, lymphocyte, monocyte, basophil, and eosinophil counts, were measured in 44,806 participants (28,534 men; mean [standard deviation] age = 38.9 [11.4] years), without a history of chronic disease or current medication. Depressive symptoms were assessed with the Questionnaire of Depression, Second Version, Abridged. Smoking status was self-reported and categorized in five classes. Sex, age, alcohol intake, self-rated health, body mass index, glycemia, physical activity, household composition, occupational status, and education were included as covariates. Associations were examined with general linear models and causal mediation analyses. RESULTS: After adjustment for all covariates except smoking, depressive symptoms were positively associated with neutrophil count only (β = 5.83, standard error [SE] = 2.41, p = .014). After further adjustment for a semiquantitative measure of smoking, this association was no longer significant (β = 2.40, SE = 2.36, p = .30). Causal mediation analyses revealed that smoking mediated the association (p < .001), accounting for 57% of its total variance. In contrast, depressive symptoms were negatively associated with lymphocyte count in fully adjusted model only (β = -3.21, SE = 1.11, p = .004). CONCLUSIONS: Smoking may confound or mediate the association between depressive symptoms and neutrophil count. These results advocate for including an accurate measure of smoking in future studies addressing this association. When considering the link between depression and inflammation, one should not overlook the noxious effects of smoking.
OBJECTIVES:Depressive symptoms have been associated with chronic low-grade inflammation, including elevated neutrophil count. Smokers often have both high neutrophil count and depressive symptoms. Thus, smoking could explain the cross-sectional association between depressive symptoms and neutrophil count. METHODS: Total white blood cell count and subtypes, including absolute neutrophil, lymphocyte, monocyte, basophil, and eosinophil counts, were measured in 44,806 participants (28,534 men; mean [standard deviation] age = 38.9 [11.4] years), without a history of chronic disease or current medication. Depressive symptoms were assessed with the Questionnaire of Depression, Second Version, Abridged. Smoking status was self-reported and categorized in five classes. Sex, age, alcohol intake, self-rated health, body mass index, glycemia, physical activity, household composition, occupational status, and education were included as covariates. Associations were examined with general linear models and causal mediation analyses. RESULTS: After adjustment for all covariates except smoking, depressive symptoms were positively associated with neutrophil count only (β = 5.83, standard error [SE] = 2.41, p = .014). After further adjustment for a semiquantitative measure of smoking, this association was no longer significant (β = 2.40, SE = 2.36, p = .30). Causal mediation analyses revealed that smoking mediated the association (p < .001), accounting for 57% of its total variance. In contrast, depressive symptoms were negatively associated with lymphocyte count in fully adjusted model only (β = -3.21, SE = 1.11, p = .004). CONCLUSIONS: Smoking may confound or mediate the association between depressive symptoms and neutrophil count. These results advocate for including an accurate measure of smoking in future studies addressing this association. When considering the link between depression and inflammation, one should not overlook the noxious effects of smoking.