Authors reply: the effect of physical activity, body mass index and waist circumference on incident diabetes

Dear editor,

Thank you for the opportunity to submit a response letter. In reference to the letter written about our article [1,2], the main points raised are noted below. Our responses to each of these comments are in italicized text.

The questionnaire used in the study should have been highlighted with regard to its validation.

As noted in Section 2.4 of our manuscript, we included text and a reference indicating the reliability and validity of the questionnaire we used.

It was not mentioned if the questions were asked in English or another language.

Given that the evaluated sample is an African American sample, residing in an American city, coupled with the manuscript text written in English, it is implied that the survey questions were in the English language.

A proper reference has to be cited in order to label the case as incident diabetes while considering the value of A1C or plasma glucose which was found to be missing.

As noted in Section 2.6, the thresholds that we utilized for incident diabetes were ≥6.5% and ≥126 mg/dL, respectively, for A1C and fasting plasma glucose. These are standard and well established cut-points for diabetes.

The authors mention that there were no three-way interaction effects of physical activity, body mass index, and waist circumference on incident diabetes, which is not well understood and debatable.

We agree that this three-way interaction may be observable in other cohorts; it is not debatable as to whether it is present in our database.

The duration or intensity of physical activity could have been measured in more detail.

With large-scale epidemiological studies, there is always a trade-off between the comprehensiveness and feasibility of the assessments. As noted in Section 4 of our manuscript, we have addressed the limitations of the self-reported physical activity measurement used in our study.

Many factors such as diet, smoking or any other form of addiction, level of cholesterol and lipid profile, and presence of hypertension should be ascertained in all the studied subjects. It is also important to measure the lifestyle and history of the mental state. Genetic analysis may prove to be more beneficial in order to estimate the incident diabetes.

As noted in Section 4 of our manuscript, we have indicated that a common limitation for large-scale epidemiological studies is the inability to fully discount residual confounding bias. Our analyses did, however, control for various confounding parameters (including, for example, level of cholesterol and lipid profile, among many other potential confounders).

Overall, an interesting article. We congratulate the authors and thank the editor for publishing such an important study.

We appreciate your interest in our article.

Sincerely,

Authors

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This paper was not funded.