I read an article in The Hindu today that reported a study that was presented at the The Endocrine Society’s annual meeting at San Francisco. The study claimed that in post menopausal women, reduction in weight led to improvement in memory. The study had very specific criteria of people who are eligible: post-menopausal women whose BMI (Body Mass Index) is greater than 27.
I tried to look for the source of this article. The results seem to have been presented at a meeting, and I could not claim my hands on the study (not for lack of access, I don’t see it published anywhere online). The closest I could get was to find the meeting abstract. I also found this article on Huffington Post that gave more details about the study.
I have multiple problems with the study (based on what I could gather from the articles and the abstract linked earlier).
What prompted me to find the source of the article was the sample size. Of late, I have learnt how to read science articles a little bit better. One of the key things to notice in such studies is the sample size. This study was done on 20 women. Yes, just 20. The Huffington Post article referred to an earlier study that claimed that obesity in post menopausal women leads to memory impairment. That study had more than 8000 subjects.
The second problem was the ease with which it has been suggested that it was the reduction in weight that led to better performance. The memory test given was to remember picture-name pairs and to recollect them later. The study says that after reduction in weight, FMRI scans revealed higher activity while storing the memory, and lower activity while retrieving it, suggesting that once weight was reduced, it was storing well (using more resources) and recollecting easily (using very little effort). There is nothing to indicate that it was weight loss that caused this improvement. What if the women were feeling more confident and happy after losing weight and so performed better? What if they knew that this was a post-evaluation and so had to do better as they felt obliged to show improvement? What if the post-weight loss memory assessment showed better performance merely because they had already done the same activity before losing weight and so were better at it a second time?
I am also surprised that all the 20 women they chose did indeed stick to their diets and lost weight. The probability of that happening, appears very less to me. If it were so easy to induce people to stick to a diet (this study seems to have done it with 100% success), there would be a lot less number of fat people. If this 20 was only a subset of the people they chose (the subset that did indeed lose weight), then why was that not mentioned in the abstract? Also, if they chose the people who lost weight, did they exclude anyone after post-weight loss assessment?
All of this makes me wonder whether the study established anything at all. I would say no. Of course I have not seen the study itself, but a sample of 20 is too low to establish or even suggest any sort of causality, much less suggest something about our famously complicated brains.