statins dec dementia after concussion

As an addendum to the prior blog on atorvastatin improving renal function (see http://gmodestmedblogs.blogspot.com/2019/05/atorvastatin-retards-renal-function.html), a recent study came out suggesting that statins decrease the risk of dementia following a concussion (see concussion statin dec dementia jamaneuro2019 in dropbox, or doi:10.1001/jamaneurol.2019.1148).

Details:
-- population-based study in Ontario Canada from 1993 to 2013, then continued until March 2016 for follow-up
-- older adults diagnosed with a concussion, excluding severe cases resulting in hospitalization and those with a prior diagnosis of dementia or delirium
-- they measured the long-term incidence of dementia, comparing those who had a statin prescription written within 90 days of the concussion vs not [this is a simple database linkage, so no information about why they were on a statin]
-- 28,815 patients were diagnosed as having a concussion
-- median age 76 (30% were >80 yo), 61% female, socioeconomic status: 38% higher, 20% middle, 42% lower; 85% urban
-- cardiovascular drugs: more patients on statins were on ACE inhibitors (45% vs 25%), ARBs (18% vs 7%), beta blockers (38% vs 18%), calcium blockers (37% vs 23%), diabetes medications (26% vs 10%)
-- neuropsych medications: no difference between groups: 20% on benzos, 3% antipsychotics, 0.5% cholinesterase inhibitors, 24% antidepressants
-- 7058 (25%) received a statin, 21,757 (75%) did not
-- as a background, the annual incidence of dementia was 19 cases per 1000 patients for adults >65 yo in Ontario

Results:
-- 4727 subsequently developed a diagnosis of dementia, over a mean follow-up of 3.9 year: incidence of one case per 6 patients
    -- those on statins accounted for 1050 dementia cases over 28,129 patient years: incidence of 37 cases per 1000 annually (this is twice the population norm)
    -- those not on statins had 3677 dementia cases over 85,339 patient years: incidence of 43 cases per 1000 patients annually
-- overall, those who received a statin had a 13% reduced risk of dementia, RR 0.87 (0.81-0.93), p<0.001
    -- number-needed-to-treat (NNT) of about 50 patients
-- this became a 16% reduction after adjusting for age, sex, socioeconomic quintile, cardiovascular medications, neuropsych medications, total prescriptions, physician visits, hospital admissions, with aRR 0.84 (0.78-0.90)
-- other predictors of developing dementia included: increasing age, lower socioeconomic status, urban home location, use of antipsychotics/antidepressants/cholinesterase inhibitors.
-- Of note, no relation between dementia and non-statin lipid-lowering drugs, ACE inhibitors, ARBs, beta blockers, calcium blockers; though there was a 21% increase in those on diabetes medications [ie, the risk reduction associated with statins was not found with other cardiac drugs, suggesting it was the statin and not underlying cardiovasc condition]
-- specific statins:
    -- rosuvastatin, 1418 patients, 22% decreased risk of dementia, RR 0.78 (0.67-0.91)
    -- atorvastatin, 3457 patients, 8% decrease, RR 0.92 (0.84-1.00)
    -- simvastatin, 1156 patients, 7% decrease (0.81-1.08)
    -- overall:
        -- hydrophilic statins, 2024 patients, 24% reduction, RR 0.76 (0.67-0.86)
        -- lipophilic statins, 5034 patients, 9% reduction, RR 0.91 (0.84-0.98)
-- there was no difference by higher vs lower dose of the specific statins
-- statin timing:
    -- those who were already on a statin had a 14% decrease, RR 0.86 (0.80-0.93)
    -- those who initiated a statin within 90 days after a concussion, 15% decrease, RR 0.85 (0.74-0.99)
    -- those who discontinued a statin, not statistically significant 9% decrease RR 0.91 (0.78-1.06)
-- there was no difference in the depression risk after concussion, comparing those receiving a statin vs not
-- for comparison: 307,891 patients had an ankle sprain, 25% received a statin [as with concussions, this is just a database link of those with ankle sprains who happened to fill a statin script or not]
    -- 25,956 patients developed dementia, 6239 were on statins, an annual incidence of 19 cases per 1000 patients
    -- 19,717 developed dementia not on a statin, an annual incidence of 20 cases per 1000 patients
    -- this translated to a 5% decreased risk of dementia in those on statins, an NNT of about 220 patients [ie, much less benefit than in those who had a concussion]

Commentary:
-- more than 1 million Americans have a concussion each year, disproportionally in older adults
-- non-human studies suggest the statin might decrease injury-related brain edema, oxidative stress, amyloid protein aggregation, and neuroinflammation
-- although statins do not seem to improve cognition in those already diagnosed with dementia, they may decrease the risk of dementia and age-related cognitive decline (see below)
    -- Statins might prevent subsequent neurologic dysfunction by neuron preservation and neural stem cell activation
    -- however, statins might also have significant effects on neural transmission, adversely affecting neurologic function
-- in this study, the lipophilic statins (e.g. simvastatin) were associated with decreased risk of dementia, but this was significantly less than with the hydrophilic ones (e.g. rosuvastatin)
-- it is notable in the study that:
    -- the decrease in dementia was not found in those on any other cardiovascular medications, including non-statin lipid-lowering meds
    -- rosuvastatin, a hydrophilic statin, had the greatest benefit
    -- there was no dose-response curve, i.e. no benefit from larger doses
    -- those who received a statin both before and after the concussion explained most of the risk reduction
    -- and there was a 5% risk reduction for developing dementia with statins in a group having an ankle sprain, suggesting that there was some small effect of statins in decreasing dementia, but much less so than those who had a concussion
-- 4 other studies have looked at statins in those with traumatic brain injuries with conflicting results, 2 finding positive protective effect of statins on neurocognitive outcomes and 2 reporting no significant effect
-- limitations to the study include: this is not a randomized trial but a large observational database, so cannot assess causality; they did not have information about other important issues which could contribute to dementia, including smoking, daily exercise, drug adherence; this study involved older adults and may not be applicable to sports injuries in young people; they only looked at patients with relatively mild concussions, not requiring hospitalization; it should be recognized that both concussion and dementia are broad categories which include potentially many different types of patients (and, therefore, there might be variable responses depending on the specific subcategories of each). Also, there may have been fundamental differences/unknown confounders in those patients who happen to be put on a statin vs not, and this could introduce a significant bias in their results. BUT, overall this was a really well-done study, including lots of patients followed for a long time, and linkages of patient databases and filled prescriptions (all available since Ontario has a universal health care system, with large and pretty complete databases. Would be great to have that here….), so not relying on patient recall

-there have been related concerns about the differences in different statins with CNS outcomes, perhaps attributable to differences in lipophilicity:
    --depression: no head-to-head comparison of studies, but a large commercial claims database study of 300K statin initiators did not find any increased risk of depression comparing hydrophilic to lipophilic statins, with adjusted HR of 1.05 (0.94-1.16). though, just looking at the most lipophilic statin (simvastatin), there was a small (??clinically significant) 7% increase, HR 1.09 (1.02-1.16): see statin depression jaffectdis2018 in dropbox, or 10.1016/j.jad.2018.06.021. also, no effect on depression found in above Ontario study
    --cognitive impairment: there have been studies finding either benefit (postulated through decreased inflammation, improved endothelial function, decreasing oxidation, etc) vs harm (postulated through cholesterol depletion, where cholesterol is an important for myelin sheath formation, mitochondrial function, neurotransmitter receptor expression, synapse development, etc). A really detailed review suggested the likely scenario: lipophilic statins (simvastatin, atorvastatin) may actually prevent dementia in people under 70-80 yo (likely from the first mechanism of decreasing inflammation etc), yet led to reversible short-term cognitive impairment in some (likely through the second mechanism of interfering with neural transmission), and this latter concern may be augmented in those on meds or genetics that decrease statin metabolism. see statin dementia lipophilic transneurodegen2018 in dropbox, or doi.org/ 10.1186/s40035-018-0110-3
        -- see blog reviewing the complexity of cardiac drug interactions with statins:  http://gmodestmedblogs.blogspot.com/2016/11/cardiac-drug-interactions-with-statins.html

         -- and recent blog looking at the effect of profound cytochrome P450 inhibitors and dramatic increases in cushings from inhaled or intraarticular steroid injections: http://gmodestmedblogs.blogspot.com/2019/05/hiv-meds-local-steroids-and-cushings.html

     --aggression: in a 6-month RCT comparing simvastatin 20mg vs pravastatin 40mg, men had decrease in aggression (esp men <40yo and low baseline aggression levels) and esp in those on simvastatin; this correlated with decreases in testosterone levels and was independent of changes in LDL levels. In women, aggression increased (esp in postmenopausal women with low baseline aggression levels), though pravastatin had a higher association than simvastatin; but those >55yo with baseline aggression did worse on simvastatin. [one issue here is that the lipophilic statins are more associated with sleep disturbances/nightmares than the hydrophilic ones]. see statin aggression plos2015 in dropbox, or doi:10.1371/journal.pone.0124451
    --and a recent study finding that simvastatin plus vitamin d were effective in reducing migraine recurrences (see http://gmodestmedblogs.blogspot.com/2016/01/migraine-prophylaxis-with.html  )

So,
-- it does seem that statins do have pretty substantial effects on the CNS, with some differences between lipophilic vs hydrophilic ones
    --rosuvastatin (a hydrophilic one) seems to be more effective in decreasing the risk of dementia post-concussion. and the statin effect seems real, since other cardiac meds did nothing, and a simultaneously assessed group of patients with ankle fractures had a much less protective effect from statins
    --simvastatin (the most lipophilic one) may have a small increased risk of depression
    --lipophilic statins overall may be better at preventing dementia, though may lead to small increases in short-term cognitive impairment
    --and, perhaps, lipophilic statins may decrease aggression in men, though did worse in women especially if they had higher levels of baseline aggression
--the data on the CNS effects seem to be variable and not completely consistent (there are a lot of different statin attributes which could either make things better or worse), depending on the specific effect. though it is important to note that many of the reports are retrospective or secondary analyses of cardiac trials
--given the aging population it would be great and appropriate to have actual RCTs with primary outcomes of short-term and long-term neurologic effects of statins for depression, cognitive impairment, anxiety/agitation, etc

How should this information change practice??
--it seems reasonable to put people >60yo or so who have a concussion on a statin, preferentially rosuvastatin.  [though, probably many/most should be on a statin anyway for cardiac protection]
--in terms of which statin to choose, the biggest benefit overall is cardiovascular, and the lipophilic/hydrophilic differences don’t seem to matter much.  So, I think one should choose a statin primarily based on LDL goals, which may mean using lots of rosuvastatin 40mg as the most potent of them (esp for high risk patients, where there are some pretty compelling observational data suggesting that the lower the better: http://gmodestmedblogs.blogspot.com/2018/08/very-low-ldl-levels-benefit-without-harm.html
--but, it might make sense to incorporate the lipophilicity/hydrophilicity as a secondary consideration, as above: preferentially hydrophilic (eg rosuvastatin) in those with concussion, increased risk of depression, women who develop more aggression on lipophilic statins, and those with poor sleep/nightmares on the lipophilic ones; and lipophilic statins (eg atorvastatin) in those with underlying kidney dysfunction/proteinuria, higher risk of dementia, men who are at low risk of aggression who get worse on a hydrophilic one, and migraine [though overall would be nice to have more studies confirming this approach….]

geoff

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org


to get access to all of the blogs:
1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order
2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category
3. or you can just click on the magnifying glass on top right, then  type in a name in the search box and get all the blogs with that name in them

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list