(Shostakovich, Festive Overture, Op 96 in an excellent recording from the Nobel Prize awards from 2009. We played this one in college and it was a lot of fun.)
Long-time readers will know I'm not a huge fan of statins in the water. That is to say I don't like primary preventative treatment of the general population with a poweful liver, muscle, and brain irritating cholesterol-killing drug without knowing if it is really worth it. The anti-platelet and anti-inflammatory effects may be worth it for a lucky few (while the unlucky few develop diabetes and muscle damage… seriously, click that link, and then go watch a Crestor commercial)... that said, if you are a middle aged man with a couple of heart attacks or unstable plaque and you eat Ho-Hos and Doritos and smoke two packs a day, hey, take the Zocor. It may well prevent an earlier death (though even cardiologists will admit this magical effect is not due to lowering cholesterol, but through the aforementioned anti-inflammatory and anti-platelet effects). Believe it or not, not everyone who comes to my office is excited about giving up her bag-a-day chip habit. C'est la vie.
I also have bizzare and outlandish views about cholesterol. I think it might be important to have enough. And my idea of a "normal" total cholesterol is more along the lines of 220, not "as low as possible." Though I have no doubts that super-high cholesterol (as in familial hypercholesterolemia) in the present Western food environment comes with a greatly increased risk of heart disease as there is a crapload of oxidized LDL hanging out in the blood.
On the other hand, super low cholesterol is almost always associated with increased risk of death from various causes, including suicide and violence. Those risks begin to climb below a serum level of 160, and start to really spike around 130. It's a correlation, not necessarily causative, but given the importance of a nice fluffy supply of fresh free cholesterol in the brain, it is certainly plausible that low cholesterol could be problematic. For the details, visit my blog article here. But let's be realistic. Mr. Smith with advanced diabetes and a history of stroke with a total cholesterol of 330 is unlikely to be plummeted to below 160 by a statin alone (though they might try to get him down there in a heroic effort to make the LDL as low as pharmacologicaly possible as current guidelines reccomend an LDL<70).
Anyway, I have suspicions of statins and the brain, but where the rubber meets the road is the data. (Or perhaps in the interpretation of the data ;-). But I do like the Heart and Soul study. Lots of participants, a long time…though they do use a mere 9-item "Patient Health Questionnaire" (known in the biz as the PHQ-9*) to determine depressive symptoms and not a standardized clinical interview like those careful Germans.
All the participants in Heart and Soul were folks with prior heart disease (that means we are talking about secondary prevention of early death with statins, for which there is a better track record than for primary prevention). 1024 patients (mostly male, as two of the three recruitment sites were VA hospitals) enrolled, 59 died in the first year, but 965 were able to give at least 2 or more annual measures of depressive symptoms to be included in the analysis. All right, blah blah, analysis, confounders, statistics, etc (you can read the full paper from the first link if you like):
Statin users at baseline (about two thirds of the sample) had a lower rate of depression, but were also less likely to develop new depressive symptoms over the next 6 years. Non-statin users were more likely to have depression at baseline and develop more depression symptoms later on. These correlations were preserved by removing a number of confounders (such as "nonadherence" to medication regimen, which occured in 6.1% of the statin users and 12.2% of the non-statin users**) All told, the odds of developing "depression" (meaning a PHQ score greater than or equal to 10) were 34% less among statin users than non-statin users, and those who were depressed at baseline had a 38% decreased odds of having depression at a later measured date. This number is consistent with the only other prospective cohort study done and with 2 large, retrospective case-control studies.
But here's the rub:
To date, however, randomized trials have failed to demonstrate a beneficial effect of statin therapy on psychological well-being. In 1 randomized trial of 1,230 patients with existing coronary heart disease, those assigned to pravastatin had lower subsequent depression scores (measured by the Global Health Questionnaire) during 4 years of follow-up than those assigned to placebo, but these differences were not statistically significant. Another trial in individuals with hyperlipidemia but without known coronary artery disease found no improvement of depressive symptoms…during almost 3 years of follow up among patients assigned to simvastatin versus those assigned to placebo.The authors guess that patients who take statins were "more likely to exhibit healthy other behaviors that would decrease their risk of depression." They are certainly less likely to admit to nonadherence! It's also possible that the anti-inflammatory effects of statins could help balance out (my suspicion) pummeling cholesterol metabolism when it comes to mental health states. The authors think maybe we need larger trials of longer duration to find the beneficial effects…
One more cute quote and then I'm signing off for the day: "in the previous observational study of patients with coronary heart disease, the beneficial effects of statins on depression appeared to be independent of lipid-lowering effects."
*From time to time I will get a letter from an insurance company. One of my patients will go to his or her primary care doctor's office and be given the PHQ-9. A positive score will prompt the insurance company to let me know that my patient might be depressed, and urges me to consider treating that patient. Insurance premiums well-spent?
**I can't tell from this paper how "nonadherence" was determined. It says "medical history" was determined by self-report. I can tell you that both the 6.1% and 12.2% numbers for nonadherence seem incredibly low to me depending upon the definition of "nonadherence," particulary for medicines such as statins or blood pressure medicines that one doesn't notice anything right away if you miss a dose)