It was with great interest that I read 'Cooking frequency may enhance survival in Taiwanese elderly' by Chen et al. (1) The authors found that cooking behaviour favourably predicted survivorship, possibly representing a surrogate for health status or socioeconomic circumstance. The association remained statistically significant after adjustment for many covariates, including common and classical risk factors such as smoking, alcohol, physical inactivity and cognitive function. However, hypertension, an important predictor of mortality in adulthood, was not taken into account (2). In recent research, higher blood pressure in early adulthood was associated with elevated risk of all-cause mortality and other chronic diseases (3). In this context, therefore, not adjusting for hypertension could bias the observed effect. In addition, the prevalence of hypertension may increase more among women than men in adulthood (4). Yet in Chen et al.'s study, women did much more cooking than men. A stratified analysis by sex might additionally be considered. Moreover, in the adjusted regression models, it was not clearly stated whether shopping frequency was controlled for. Since shopping frequency is highly correlated with cooking frequency, which of them is more predictive of survival needs to be more clearly investigated in this population.