A recent study by Boden et al ^{Reference Boden, Fergusson and Horwood1} concluded that there is a cause–effect relationship between cigarette smoking and depression in which tobacco use increases the risk of symptoms of depression. In a large longitudinal study, Kang & Lee ^{Reference Kang and Lee2} showed that smoking caused depression. Shahab & West ^{Reference Shahab and West3} reported evidence from a cross-sectional survey that ex-smokers feel happier following cessation.

These results may have very important clinical consequences – if smokers can be reassured that their mood can be improved after smoking cessation, it could motivate patients in their attempts to quit. Our own data are consistent with such findings and with the current literature regarding the relationship between depression and smoking status as well as gender. We performed an investigation focusing on depression symptoms among 1021 unrelated blood donors categorised as former smokers, current smokers and never smokers. The sample distribution was: former smokers, n = 131; current smokers, n = 254; and never smokers, n = 636. Former smokers were individuals who had reached 6 months of tobacco abstinence. Using a cross-sectional design, the participants were selected during the period from October 2004 to August 2008. Inclusion criteria were: to be Brazilian of European descent, ≥18 to ≤65 years old, male or female and eligible for blood donation. Exclusion criteria included other addictions, current use of any psychopharmacological medication and major psychopathologies, except major depressive disorder. All participants completed a standardised self-report questionnaire that included demographic characteristics and a smoking history. Depression symptoms were evaluated by the Portuguese version of the Beck Depression Inventory (BDI). ^{Reference Beck, Steer and Garbin4,Reference Gorenstein and Andrade5} The BDI scores were analysed as a continuous measure or as a cut-off of ≥15 indicating depressive symptoms.

Level of education was higher among never smokers (n = 164, 25.8%) compared with current smokers (n = 40, 15.7%) and former smokers (n = 24, 18.3%) (χ² ₄ = 21.56, P<0.001). This suggests that current and former smokers might share a premorbid behavioural profile different from never smokers. More current smokers had a BDI score ≥15 (current smokers, n = 38, 15.0%; never smokers, n = 47, 7.4%; former smokers, n =9, 6.9%; χ² ₂ = 13.43, P = 0.001). Average BDI scores were also higher among current smokers (mean 7.4, s.d. = 7.8) compared with never smokers (mean 5.2, s.d. = 6.5) and former smokers (mean 5.0, s.d. = 5.6) even after adjustment for gender, age and years of schooling (F = 10.93, P<0.001). There were no significant differences between former and never smokers on depression indices. There was no significant interaction between smoking status and gender – that is, females had higher depression scores than males, regardless of smoking status, pointing to the cross-gender validity of the association. Beck Depression Inventory scores were significantly correlated with Fagerström Test for Nicotine Dependence ^{Reference Fagerström and Furberg6} scores (r = 0.16, P = 0.01) and average daily number of cigarettes smoked (r = 0.16, P = 0.01). The results of our relatively large sample suggest that depression scores are lower among former smokers, despite the similar profiles in other characteristics such as education and gender across all three groups.

This issue has been raised by other authors. Wu & Anthony ^{Reference Wu and Anthony7} verified in a longitudinal study that although smoking increased the risk for depression, antecedent depressed mood was not associated with later cigarette smoking. A review by the National Institute of Mental Health ^{Reference Ziedonis, Hitsman, Beckham, Zvolensky, Adler and Audrain-McGovern8} pointed out the danger posed by over-reliance on the self-medication hypothesis. According to the authors, this misconception may have led to a grossly inadequate attention to tobacco-smoking in mental health settings. Munafò et al ^{Reference Munafò, Hitsman, Rende, Metcalfe and Niaura9} have suggested a causal relationship between cigarette smoking and depression.

The interpretation of our results should be cautious, since cause–effect relationships cannot be explained in cross-sectional studies, where recall bias is always a possibility. Former smokers may differ from current smokers both in terms of their primary depression and nicotine dependence severity. As Fagerström& Furberg ^{Reference Fagerström and Furberg6} pointed out, less dependent smokers may quit more easily and remaining dependent smokers may need more intensive treatment. Another scenario is that previous depressive symptoms might have predisposed some individuals to smoke, and when symptoms faded, they stopped smoking.

Our preliminary results are consistent with these findings, suggesting that former smokers have a better mood than current smokers. If confirmed in future follow-up studies, this evidence will certainly stimulate new approaches for smoking prevention in adolescence and smoking cessation techniques for adults. If smokers can be reassured that their mood may actually improve after smoking cessation, once the withdrawal syndrome has ended, ^{Reference Munafò and Araya10} this knowledge could motivate patients in their attempts to quit. We agree with this position and suggest that it is equally valid for both genders.