Letter by Archer Regarding Article, "Sugar- and Artificially Sweetened Beverages and the Risks of Incident Stroke and Dementia: A Prospective Cohort Study".

Abstract

To the Editor: I read the recent article by Pase et al with great interest. However, my enthusiasm was short lived because these investigators failed to incorporate findings that directly impacted the validity of their report. In their investigation, Pase et al used memorybased dietary assessment methods (M-BMs; eg, Food Frequency Questionnaires) without caveat, despite the fact that the validity of these methods has been seriously questioned. My colleagues and I showed that M-BM data were physiologically implausible, incompatible with survival, and inadmissible as scientific evidence. Our conclusions were supported by ≥60 years of highly replicated, rigorous evidence and suggest that the data by Pase et al were merely meaningless numbers derived from highly edited, pseudo-quantified anecdotes of dietary intake. It is unequivocal that there are clinically significant qualitative and quantitative differences between what people remember eating and drinking, what they are willing and able to report eating and drinking, and the actual foods and beverages they consumed. Thus, our work demonstrated that Pase et al did not measure beverage consumption but instead collected data on what their respondents were willing and able to report consuming. We contend that these nominal psychological data (ie, reported memories) are irrelevant to examining the physiological effects of actual dietary intake on health. Furthermore, the measurement errors associated with the M-BM data by Pase et al were nonquantifiable because of false memories, confabulations, forgetting, and intentional misreporting (ie, lying). As we wrote previously, “without objective corroboration it is impossible to quantify what percentage of the recalled foods and beverages are completely false, grossly inaccurate, or somewhat congruent with actual consumption” (p. 919) and “neither the researchers nor the participants know the validity or reliability of the reported food and beverage consumption ...”. (p. 918) Thus, the true beverage consumption of the participants included in the study by Pase et al was unknown, and the measurement error was nonquantifiable. Given that the demarcation between scientific and pseudoscientific (ie, unfalsifiable) data is contingent on the quantifiability of measurement error, M-BM data are clearly inadmissible as scientific evidence. In response to our reports, one group of eminent epidemiologists stated, “...do not use self-reported energy intake (M-BMdata) as a measure of true energy intake ...”. Nevertheless, Pase et al presented their clearly implausible M-BM data as estimates of true energy intake (eg, Table 1, p. 3) without caveat or reference to contrary evidence. Most epidemiologists now agree that M-BMs energy-intake data are not plausible and should not be used as estimates of true energy intake. Thus, it is misleading to the larger research and medical communities when these data are published with no acknowledgment of the research questioning the validity of their methods and conclusions. In summary, Pase et al failed to cite rigorous contrary evidence, failed to follow to a discipline-wide prohibition on the use of M-BMs energy-intake data, and presented reported memories of consumption despite the fact that these data were irrelevant to actual dietary intake. Thus, the conclusions made by Pase et al are not credible and their article should be corrected or retracted.

DOI: 10.1161/STROKEAHA.117.018032

Cite this paper

@article{Archer2017LetterBA, title={Letter by Archer Regarding Article, "Sugar- and Artificially Sweetened Beverages and the Risks of Incident Stroke and Dementia: A Prospective Cohort Study".}, author={Edward Archer}, journal={Stroke}, year={2017} }