Do Self-Experiments Tell Us Much?

Eric Robinson (Liverpool University) carried out a self-experiment on the impacts of a vegan and a non-vegan diet on his physical and mental attributes. The results were published in Physiology and Behaviour (the journal that accepted my own first paper in the 1960s). Robinson appears to have done his best with the necessarily-limited dataset. Have the results, however, any real messages for society? (https://theconversation.com/i-self-experimented-to-compare-a-vegan-diet-with-eating-meat-this-is-what-i-found-out-212213). Robinson recorded what he did each day (including exercising, eating out and drinking alcohol) but also rated his enjoyment of the foods he consumed. On a weekly basis, he was weighed, had his waist circumference measured and completed self-report questionnaires on depression and anxiety symptoms. Robinson had a month on a strictly vegan diet, before switching to a non-vegan alternative (apparently making a real effort to eat meat and dairy-containing foods). He then reversed the sequence. Robinson's self-experiment didn't lead to any 'light-bulb moment', although he's resumed a vegan habit. There were, however, some interesting observations made. Neither his enjoyment of food nor his depression/anxiety measures appeared changed by the diet. He also clearly lost weight on the vegan diet. Interestingly, this may be linked to vegetarianism acting as a 'red light' to unnecessary eating. Snacks and sweets required thought, when on this diet. There may be an element of self-censorship, when one has to consider the potential ingredients before eating. Robinson did note that people found him more 'fun', when he was on a non-vegan diet. This may, however, simply reflect those folk finding it more relaxing, when not having to concern themselves with Robinson's diet. The basic troubles with self-experiments involving one individual are that a) subjective, self-reported measures like depression and anxiety cannot be validated; b) doing anything like of 'gold standard' double blind experiment is clearly out of the question and c) the results may only tell us about the personal odyssey of the subject. There appears, however, to be scope for more substantial investigations on suggestions coming from this study.