PT155 · S4 · Q23 Researcher: Consumption of turmeric

It's hard to say whether this counts as any support; the fact that they've always suspected something doesn't make it true. At a criminal trial, it doesn't add anything to the prosecution's case to say, "Even before this recent murder that Eddie has been accused of, we always thought Eddie would probably murder someone some day".

This provides a possible alternate explanation. If the curry-eaters are more likely to be highly educated, then maybe it's the high levels of education, not the turmeric in the curry, that is slowing down cognitive decline.

We already knew that the study found a correlation between people who ate curries regularly and better performance on the cognitive test. It doesn't matter whether the people who eat curries regularly are 10% of the data set, 40%, 70%, etc. If they're 10% of the data set but 15% of the people who got top scores on the cognitive test, then we would still have our correlation. So similarly, it doesn't matter if the Indian ethnicity segment of the data set is higher or lower than 50%. Either way, we found that within the segment of the data set who were of Indian ethnicity, the correlation between regularly eating curry and ranking higher on the cognitive test was especially strong.

It's irrelevant how Singapore residents compare to residents of other countries in terms of how often they eat curries. We could have done this study in Mexico, where people definitely eat curries less than residents of other countries, and it wouldn't affect how we think about the data set at all. If I say that "in a study we did at a certain prison, people who were left-handed were much more likely to believe religion than were people who are right-handed", it makes no difference whether left-handed people are more/less common than right handed people, and it makes no difference whether the residents of this prison are more/less left handed than those of other prisons, etc. Correlations just mean disproportionate representation. Whatever the underlying prevalence of left handed people are, they are more likely to be associated with religious beliefs. Similarly, whatever the underlying prevalence of frequent curry eaters were, they were more likely to be associated with better cognitive performance.