We were interested to read the article by Okuda et al.1 that was published in Hypertension Research in July 2016. The authors aimed to evaluate the effects of height, weight, age, spot urine mineral concentration, maturation and physical activity on 24-h sodium and potassium excretion in Japanese secondary school students. The authors noted in their conclusion that physical activity is an important determinant of 24-h sodium and potassium excretion in adolescents.1 Such a conclusion can be considered optimistic, unless the result can be validated internally or externally using bootstrap reputation, split validation and cross-validation.2

Regression models of 24-h sodium and potassium excretion showed small regression co-efficients for the explanatory variables and large intercepts. The intercept will be large if the number of observations and the number of explanatory variables are low. The sample size in the regression analysis by Okuda et al.1 was sufficient, as 320 secondary school adolescents were recruited for the study. It seems that the large intercept was a result of neglecting other relevant determinants of 24-h sodium and potassium excretion. Given that the authors reported important determinants of 24-h sodium and potassium excretion in adolescents, such a conclusion should be interpreted with caution.

Moreover, the authors should have reported the confidence intervals for the estimated associations. Reporting only p-values is a common mistake.3