A recent study using data from the Framingham Heart Study noted associations between frequent consumption of artificially sweetened soft drinks (≥ 1 per day versus none) with incident stroke and dementia over a 10-year period. 1 The researchers state in their discussion that their study provides further evidence linking consumption of artificially sweetened soft drinks (≥ 1 per day versus none) with risk of ischemic stroke. There are 2 reasons why this study should be interpreted with caution:

–The researchers emphasized the less controlled analyses for dementia rather than the fully controlled models, which are not the main findings for the study. They reported hazard ratios of 2.96 (95% confidence interval, 1.26–6.97) for ischemic stroke and 2.89 (95% confidence interval, 1.18–7.07) for Alzheimer’s disease in their abstract, which were not their final fully controlled models. Undertaking mediation analyses is commendable, but the analysis after the mediation analysis should have controlled for the potential confounders of diet quality, physical activity, smoking, as well as the cardiometabolic factors that were not tested as mediators. 2 Instead, the researchers controlled only for age, sex, education, and energy intake (model 1 variables).–An excessive number of analyses without adjustment of