Goals of work: Physical activity is claimed to have a beneficial influence on a person's mental health, though its impact may be different in cancer patients and individuals in the general population. The objective of this cross-sectional study was (1) to estimate the associations between the level of self-reported physical activity as a lifestyle factor (LPA) and the prevalence of depression and anxiety disorder in testicular cancer survivors (TCSs) and in men of similar age from the general population (GenPop), and (2) to examine whether these associations differ in the two groups.
Patients and method: A total of 1260 TCSs and 20,207 men from the GenPop completed a questionnaire that assessed LPA, and depression and anxiety disorder on the Hospital Anxiety and Depression Scale (HADS).
Main results: The prevalence of HADS-defined depression was lower among those who were physically active than in those who were physically inactive (TCSs 9% vs 17%, P<0.001; GenPop 8% vs 15%, P<0.001). Among physically inactive TCSs there was a trend towards higher a prevalence of HADS-defined anxiety disorder compared to physically active TCSs (P=0.07). In the GenPop this difference was statistically significant (P<0.001). Multivariate analysis confirmed the association between LPA and HADS-defined depression in both the TCSs [adjusted odds ratio (aOR)=0.56; 95% confidence interval (CI) (0.31, 1.02)] and the GenPop [aOR=0.58; 95% CI (0.51, 0.65)], but not the association between LPA and HADS-defined anxiety disorder. The associations between LPA and HADS-defined depression and HADS-defined anxiety disorder did not differ between the TCSs and the GenPop.
Conclusion: In both the TCSs and the GenPop, the prevalence of HADS-defined depression was higher among those who were physically inactive than among those who were physically active, with no intergroup difference. There was no association between LPA and prevalence of HADS-defined anxiety disorder in either of the groups after adjusting for background variables. Prospective studies are needed to allow causal inferences to be drawn.