Reader Comments
Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Thank You!
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.
closeReverse causation not ruled out
Posted by jfranklin on 01 Oct 2015 at 09:32 GMT
The authors have in general taken care to report their results as an association between change in intake and change in weight, without implying that the relationship is causal. However, through use of phrases such as 'The benefits were greater for fruits compared to vegetables...' (Discussion line 2), causality still creeps in.
My concern is with possible reverse causation, that is, a weight gain could cause the participant to eat more vegetables or fruit because he/she perceives these to be healthy or undertakes a diet which prescribes such food choices. Such behaviour would lead to a negative association, as observed, in this study because the corresponding intake changes and weight changes were oberved over the same 4-year period.
A sensitivity analysis was performed 'examining change in diet over 4 y and change in weight over the following 4 y interval'. The results are shown in supplementary table 9 (last column): the associations seen in the main analysis have disappeared. Unfortunately, these sensitivity results are not reported, discussed or even referenced in the main article.
In my view, it is important to draw attention to these sensitivity results in order to avoid overinterpretation of the results of this important study. Further, a more detailed investigation of various 'time lags' between intake data and weight-change data might be valuable.
RE: Reverse causation not ruled out
mbertoia replied to jfranklin on 02 Oct 2015 at 20:59 GMT
We agree that these results cannot be interpreted as causal and that the potential for reverse causality cannot be ruled out. How reverse causality would affect our estimates is unclear; it's possible that the positive associations between starchy vegetables and weight could be overestimated if reverse causality is present, for example if individuals who gained weight in the beginning of a 4-year period began eating more starchy vegetables in response during the end of that 4-year period. On the other hand, inverse associations would be underestimated with this particular scenario.
Sensitivity results comparing various time lags between changes in intake and changes in weight in these cohorts have been analyzed and discussed in detail in the following publication, "A Comparison of Different Methods for Evaluating Diet, Physical Activity, and Long-Term Weight Gain in 3 Prospective Cohort Studies" by Smith et al. in the Journal of Nutrition, published on September 16 2015. This paper concludes that the analysis of concurrent change in diet and weight is best, although it still has important limitations.