This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

Letter to the Editor

Climate and Prevalence of Mood Disorders: A Cross-National Correlation Study

Jean-Yves Rotgé, MD, PhD; Philippe Fossati, MD, PhD; and Cedric Lemogne, MD, PhD

Published: April 15, 2014

Climate and Prevalence of Mood Disorders: A Cross-National Correlation Study

To the Editor: Seasonal climate variations are notably known to affect mood in both individuals with and individuals without mental disorders.1,2 Intriguingly, the prevalence of mood disorders differs greatly from one country to another. For example, lifetime prevalence has been measured at 21% in France and the United States, compared with under 5% in Nigeria.3 Since seasonal variations could affect the mood, we conducted the following study to examine whether some climate characteristics could, at least in part, explain the worldwide distribution of lifetime prevalence of mood disorders.

Method. The lifetime prevalences of mood disorders from 17 countries belonging to each continent in both the Northern and Southern hemispheres in 2001-2005 were extracted from the World Health Organization World Mental Health Survey (see eAppendix 1 at For each country, the population size, the gross domestic product (GDP) per habitant, and climate characteristics for the period 1990-2009 were obtained from the World Bank Group (

The differences between the highest and lowest monthly values of rainfall and temperature were calculated to serve as estimates of the amplitude of the seasonal variations. Four independent variables were thus obtained: the annual averages of rainfall and temperature and the amplitudes of monthly differences in rainfall and temperature. The lifetime prevalence of mood disorders was the dependent variable of the model. The selection of the best multiple regression model was based on the significance of F tests and the highest R2 values. The model was further tested with the inclusion of population size or GDP per habitant as a control variable to assess their possible influence on the relationship between climate and prevalence of mood disorders. Finally, leave-one-out sensitivity analyses were conducted by repeating analyses with the consecutive exclusion of each data point to ensure that the overall results were not influenced by outlier data.

Results. The best multiple regression model included the 4 independent variables (F = 4.78, R2 = 0.62, P = .015). The β values, indicative of the direction and the strength of the relationship between each independent variable and the lifetime prevalence of mood disorders, were −1.30 (P = .02) for the amplitude of monthly rainfall differences, 0.81 (P = .003) for the average annual rainfall, 0.23 (P = .64) for the amplitude of monthly temperature differences, and 0.53 (P = .40) for the average annual temperature. The inclusion of the population size (F = 16.84, R2 = 0.86, P < .0001) or the GDP per habitant (F = 4.33, R2 = 0.59, P = .021) as a control variable did not affect the present results. Finally, leave-one-out sensitivity analyses identified no outlier data.

Some limitations should be taken into account when interpreting the present results. First, the data did not take into account the regional variability within a given country of both climate characteristics and lifetime prevalences of mood disorders. Climate properties were measured in the capitals of the included countries, and lifetime prevalences corresponded to the national prevalences. Second, although the years of data collection for lifetime mood disorder prevalences and climate characteristics could not perfectly fit, their overlap (2001-2005 for lifetime prevalences and 1990-2009 for climate characteristics) seems reasonable and should not affect our results and conclusions, given the velocity of climate change. Finally, we were unable to consider the different subtypes of mood disorders, and we cannot specifically conclude whether climate characteristics may affect the lifetime prevalence of major depressive disorder, bipolar disorder, or both.

However, this study is the first to describe the relationship between the climate of a country and its lifetime prevalence of mood disorders (see eAppendix 1). Specifically, the highest prevalences of mood disorders are observed in countries characterized by small variations across monthly rainfall and high levels of rainfall, independently of the countries’ wealth.

The present findings extend previous results by highlighting the critical importance of the weather in mood disorders through a cross-national dimension and offer new perspectives regarding the discrepancies across national lifetime prevalences of mood disorders. It remains to be determined, for example, whether the relationship between climate and prevalence of mood disorders could be mediated by light exposure or barometric pressures and whether climate changes to come may affect the occurrence of mood disorders.


1. Schlager D, Schwartz JE, Bromet EJ. Seasonal variations of current symptoms in a healthy population. Br J Psychiatry. 1993;163(3):322-326. PubMed doi:10.1192/bjp.163.3.322

2. Magnusson A, Partonen T. The diagnosis, symptomatology, and epidemiology of seasonal affective disorder. CNS Spectr. 2005;10(8):625-634, quiz 1-14. PubMed

3. Kessler RC, Angermeyer M, Anthony JC, et al. Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative. World Psychiatry. 2007;6(3):168-176. PubMed

Jean-Yves Rotgé, MD, PhD

Philippe Fossati, MD, PhD

Cedric Lemogne, MD, PhD

Author affiliations: INSERM (French Institute of Health and Medical Research) UMR 894, Psychiatry & Neurosciences Center, Paris Descartes University, Sorbonne, Paris Cité, Faculty of Medicine (Drs Rotgé and Lemogne); Pierre and Marie Curie University; INSERM U 1127, CNRS (National Center for Scientific Research) UMR 7225; Brain and Spine Institute; and AP-HP (Paris Hospitals Public Assistance), Department of Psychiatry, Pitie-Salpetriere Hospital (Dr Fossati); and AP-HP, Department of Old Age Psychiatry, University Hospital Paris West (Dr Lemogne), Paris, France.

Author contributions: All authors were involved in the research and writing of this letter.

Potential conflicts of interest: None reported.

Funding/support: None reported.

Supplementary material: Available at PSYCHIATRIST.COM.

Related Articles

Volume: 75

Quick Links: