3 ms·

>I take this to mean that overall, the risk do develop dementia is 2.5/10 = 0.25 times the base risk, so overall, the Alzheimer's risk is only 4 times lower wit

by carlmr 8y ago

>I take this to mean that overall, the risk do develop dementia is 2.5/10 = 0.25 times the base risk, so overall, the Alzheimer's risk is only 4 times lower with herpes medication, when compared with the general population. Looking at Wikipedia "According to the World Health Organization 67% of the world population under the age of 50 have HSV-1." and according to the source there 11% have HSV-2. There's probably some good amount of overlap between the two, but we have somewhere between 67 and 78% of the population with some form of herpes. In any case the "general population" is closer to having herpes than not having herpes, so we'd probably arrive at a factor somewhere closer to the 10x.

credit_guy 8y ago
In most cases that HSV-1 or HSV-2 is dormant, and the person is considered healthy. In this study, the people diagnosed with HSV were considered sick, and in some cases there was a need for the so-called "aggressive antiviral therapy". Now, the study seems quite well conducted, there was a large enough group of people diagnosed with HSV (about 8k people) and a control group of about 25k people who were aged-matched with the first group, but healthy. The two group were followed for 9 years. The main conclusion of the group, that the people with HSV were 2.5 times more likely to develop dementia appears sound, and the power of the experiment probably many times higher than the one necessary to determine statistical significance. The secondary conclusion, that the people "aggressively treated with antivirals" is not as strong, in my opinion. First of all, it's not clear how many in the primary group were subject to that treatment. In an extreme scenario, maybe we are talking only about 50-100, in which case, we may end up with an intended or unintended case of p-hacking. Second, the researchers may have looked for hundreds of things, like how more prone to Alzheimer people who drank more alcohol were, or people who watched too much TV, or other hypotheses. And we may end up again with p-hacking (like in the acne-causing-green-jelly-beans xkcd [1]). [1] https://www.xkcd.com/882/ https://www.xkcd.com/882/

by carlmr 8y ago

credit_guy 8y ago

In most cases that HSV-1 or HSV-2 is dormant, and the person is considered healthy. In this study, the people diagnosed with HSV were considered sick, and in some cases there was a need for the so-called "aggressive antiviral therapy". Now, the study seems quite well conducted, there was a large enough group of people diagnosed with HSV (about 8k people) and a control group of about 25k people who were aged-matched with the first group, but healthy. The two group were followed for 9 years. The main conclusion of the group, that the people with HSV were 2.5 times more likely to develop dementia appears sound, and the power of the experiment probably many times higher than the one necessary to determine statistical significance. The secondary conclusion, that the people "aggressively treated with antivirals" is not as strong, in my opinion. First of all, it's not clear how many in the primary group were subject to that treatment. In an extreme scenario, maybe we are talking only about 50-100, in which case, we may end up with an intended or unintended case of p-hacking. Second, the researchers may have looked for hundreds of things, like how more prone to Alzheimer people who drank more alcohol were, or people who watched too much TV, or other hypotheses. And we may end up again with p-hacking (like in the acne-causing-green-jelly-beans xkcd [1]). [1] https://www.xkcd.com/882/ https://www.xkcd.com/882/