Fruit and vegetable consumption and smoking cessation
Smoking cessation: a confounding factor insufficiently taken into account?
It is very difficult, even impossible, to conduct randomised trials on just one food or a group of foods such as fruits and vegetables (F&V). The evidence in humans is based primarily on observational cohort studies1 and these studies have shown decreased cardiovascular risks in association with the consumption of F&V2,3,4. But these statistical associations are not proof of a causal link (do F&V really protect from cardiovascular risks?). Nutrition is linked to numerous characteristics (confounding factors) in individuals (socio-economic level, alcohol consumption, smoking, physical activity, etc). Certain confounding factors are probably ignored or insufficiently accounted for in data analysis. Smoking represents an important confounding factor since it is strongly linked to cardiovascular risk. Smoking cessation is associated with a rapid return to normal risk levels. Among the 31 studies we evaluated, 29 adjusted for smoking/non-smoking status5, but the smoking levels are often assessed only upon inclusion whereas follow-up continued over several years. Only two cohorts from the same team6,7 were reassessed for smoking levels at different times during follow-up.
F&V and coronary artery diseases: a different relationship in smokers and non-smokers?
In the PRIME cohort (étude PRospective de l'Infarctus du MyocardE - Prospective Epidemiological Study of Myocardial Infarction), we observed a decreased risk of acute coronary artery diseases in smokers but not in non smokers8. These results could be explained by the lack of adjustment for smoking status evolution. It is possible that smokers who ate more F&V upon study inclusion were also the most health conscious among the smokers and thus more willing to stop smoking to reduce their cardiovascular risk.
F&V consumption associated with a greater chance of smoking cessation: the PRIME study.
To test this hypothesis, we evaluated whether consumption of F&V in smokers measured upon study inclusion, would be associated with smoking cessation after 10 years of follow-up5. This information could be obtained only in Lille and Belfast.
The PRIME study began in 1991 in populations from four WHOMONICA centers in Belfast (Northern Ireland), Lille (Northern France), Strasbourg (Eastern France) and Toulouse (South-Western France). The goal was to recruit 2,500 men aged 50–59 years to study cardiovascular risk factors. F&V consumption was measured upon study inclusion using a food frequency questionnaire.
Smoking status was determined upon inclusion based on the responses to questions concerning present and past habits. A participant’s smoking habits were investigated by means of a questionnaire sent by post after a 10-year follow-up period (in 2001). Quitters were defined as non-smokers in the last questionnaire.
Among the PRIME subjects, we analysed data from 1,056 smokers upon inclusion (580 in Lille and 476 in Belfast). After 10 years 70.7% of the smoker had quit in Lille and 37.8% in Belfast. After adjusting for centre, consumption of F&V was associated with quitting (odds ratio (OR) for high versus low F&V intake: 1.73; 95% confidence interval (CI): (1.22–2.45); P-trend =0.002). After adjustment for socio-demographic factors, body mass index and medical diet, the association was still statistically significant (OR: 1.59; 95% CI (1.12–2.27); P-trend=0.01). In a model fully adjusted for age, smoking intensity, alcohol consumption and physical activity, the association was no longer significant (OR: 1.36; 95% CI (0.94–1.97); P=0.14).
Direct protective effect from F&V or “overall healthy behaviour”?
These studies suggest that smoking cessation may at least partially explain the observed reduction in the average size of red blood cells measured as MCV (Mean Corpuscular Volume) in F&V consumers in cohort studies, thus complicating the causal interpretation of the association.
Although the association between F&V consumption and smoking cessation is not significant after full statistical adjustment, these results illustrate the difficulties in distinguishing between the direct effect of F&V consumption and those of "healthy behaviours" - including smoking cessation – often adopted by F&V consumers.
- Dauchet L, et al. Fruits, vegetables and coronary heart disease. Nat Rev Cardiol. 2009 sept;6(9):599–608.
- Dauchet L, et al.. Fruit and vegetable consumption and risk of coronary heart disease: a meta-analysis of cohort studies. J. Nutr. 2006 oct;136(10):2588–93.
- Dauchet L, et al.. Fruit and vegetable consumption and risk of stroke: a metaanalysis of cohort studies. Neurology. 2005 oct 25;65(8):1193–7.
- He FJ, et al. Fruit and vegetable consumption and stroke: meta-analysis of cohort studies. Lancet. 2006 janv 28;367(9507):320–6.
- Poisson T, et al. Fruit and vegetable intake and smoking cessation. Eur J Clin Nutr. 2012 nov;66(11):1247–53.
- Joshipura KJ, et al. Fruit and vegetable intake in relation to risk of ischemic stroke. JAMA. 1999 oct 6;282(13):1233–9.
- Joshipura KJ, et al. The effect of fruit and vegetable intake on risk for coronary heart disease. Ann. Intern. Med. 2001 juin 19;134(12):1106–14.
- Dauchet L, et al. Association between the frequency of fruit and vegetable consumption and cardiovascular disease in male smokers and non-smokers. Eur J Clin Nutr. 2010 juin;64(6):578–86.