cognitive function: improves with vitamins?
A recent study found that multivitamin-mineral supplementation was associated with improved cognitive function (see cognitive function better with MVI AmJClinNut2023 in dropbox, or doi.org/10.1016/j.ajcnut.2023.12.011). This study was highlighted in a recent New York Times article:
Details:
-- 21,442 adults in the US aged at least 60yo were in a 2 x 2 factorial trial of cocoa extract (500 mg flavanols/day with/or without daily multivitamin-mineral supplement (MVM), in the COSMOS study (Cocoa Supplement and Multivitamin Outcomes Study)
-- 492 patients participated in the COSMOS-Clinic subcohort of the overall study (see description below) and had in-person detailed neuropsychological assessments at baseline and at two years. There are also two other subgroups of the study including 2158 non-overlapping participants receiving annual telephone-based cognitive assessment for three years in the COSMOS-Mind substudy, and 2472 participants in the annual computer-based cognitive assessments for three years in the COSMOS-Web substudy
-- for the COSMOS-clinic study:
-- mean age 69, 49% female, 60% finished college; non-Hispanic white 98%
-- BMI 27, leisure time physical activity median 21.5 hours per week
-- adherence to a healthy diet (AHEI-2010, see details below):
-- lowest tertile (15.5-38.5 points): 28%
-- middle tertile (39.5-47.5 points): 34%
-- highest tertile (48.5-72.5 points): 38%
-- smoking status never smokers 55%/past or current 45%; alcohol use never or rarely 21%/monthly 8%/weekly 39%/daily 32%; regular use of multivitamin supplement 39%
-- history of hypertension 46%/CVD 6%/diabetes 9%/taking anti-lipid medication 40%/depression 30%
-- presence of subjective cognitive complaints in 32%
--exclusion criteria include history of MI or stroke; diagnosis of cancer; other serious illnesses in the patients
-- patients in the COSMOS-Clinic component had in-person health evaluations, including a 45-minute neuropsychological evaluation at baseline and at two years
-- this on-site neurophysiologic evaluation included hearing screening, Clock-in-the-Box test, a rapid cognitive screening test followed by a Modified Mini-Mental State as a measure for general cognition. Those without serious hearing impairment were given the CERAD test (Consortium to Establish a Registry for Alzheimer’s Disease) immediate total learning, delayed recall, and cognition; immediate and delayed recall trials of the East Boston Memory Test; two category fluency tests (naming animals and vegetables); Trail Making Tests A and B; and Digit Span Backwards
-- primary outcome: change in cognitive evaluations, comparing MVM and placebo from the baseline assessment to reassessment two years later
-- this study was powered to detect an effect of at least 0.12 standard deviation units (SU) in the global cognition composite score
-- secondary outcomes included episodic memory and executive function assessments
Results:
-- 98% of the participants returned study pill adherence questionnaires, with 96% reporting adequate pill adherence of at least 75% of the study pills per month at year 1, and 91% at year 2
-- Those taking MVM versus placebo, at the two-year follow-up:
-- global cognition (primary outcome):
-- MVM: mean score 0.02 at baseline, increasing to 0.19 at year 2
-- placebo: -0.01 at baseline, increasing to 0.09 at year 2
-- adjusted mean difference: 0.06 SU (standard deviation units), confidence intervals -0.0003 to 0.13, strong trend to being significant but not statistically so
-- secondary outcomes:
-- episodic memory (a composite of four tests):
-- MVM: mean score 0.01 at baseline, increasing to 0.36 at year 2
-- placebo: -0.01 at baseline, increasing to 0.23 at year 2
-- adjusted mean difference: 0.12 SU (0.02 to 0.23)
-- executive function or attention (a composite of five tests):
-- MVM: mean score 0.03 at baseline, increasing to 0.05 at year 2
-- placebo: -0.03 at baseline, decreasing to -0.05 At year 2
-- adjusted mean difference: 0.04 SU (-0.04 to 0.11), not statistically significant
-- global cognition: adding in the COSMOS-Mind study of 2158 participants (with cognition assessed by telephone at baseline and annually) to the COSMOS-clinic trial (492 participants) increased the SU differential from the statistically non-significant COSMOS-clinic group finding a 0.06 SU difference to 0.07 (0.03 to 0.11), p=0.0009, highly statistically significant. Of note, this improvement was of only 0.1 SU for their primary outcome of global cognition….
-- episodic memory: though statistically significant by COSMOS-clinic study alone, addition of COSMOS-Mind (2158 participants) plus the COSMOS-Web (another 2472 people in the computer-based wing) led to an SU differential of 0.06 , much less than the 0.12 in the COSMOS-clinic trial but more statistically significant at p=0.0007)
-- the reason that these latter analyses reached such high statistical significance is from the large number of participants in the combined groups
Commentary:
-- as we know only too well, cognitive decline/dementia is extremely common, and this will only increase as the population increases in age (it is anticipated that by 2060, one in four Americans over 65 will be so afflicted)
-- and, we know that cognitive decline tends to get worse and is not only has a huge effect on the patient, family, and community but is also associated with significant medical problems and early death
-- the prior Multiethnic Cohort Study between 1993 and 1996 found that multivitamin/material supplements did not change the risk of mortality from all causes, cardiovascular disease, cancer incidence, or morbidity overall, however there was no information provided on exactly which supplement they were taking, and this was not a randomized controlled trial:
-- A cohort study of more than 1 million adult Americans in 1982-1989 found that multivitamin users had heart disease and cerebrovascular disease mortality similar to nonusers. Overall mortality from all cancers combined was increased among male current smokers who took multivitamins. But cancer was decreased in males taking the combination the combination of a multivitamin along with most of vitamins A, C, or E in combination: https://pubmed.ncbi.nlm.nih.gov/10909952/
-- a few older studies had found that males who smoked and were taking beta-carotene, an antioxidant, actually had an increased lung cancer incidence (eg, two studies found that there was an 18% -28%increased risk: https://www.mayoclinic.org/drugs-supplements/beta-carotene-oral-route/precautions/drg-20066795#:~:text=Use%20of%20beta%2Dcarotene%20has,for%205%20to%208%20years). The editorialist for one of the studies (Dr Walter Willett) made the astute comment that there are 100s of carotenoids known, and that it was much more important to eat healthy foods which have naturally occurring combinations of carotenoids and other vitamins/minerals, than to take the reductionist road of attempting to find a single ingredient formulated as a pill to improve outcomes
-- A 12-year study of 6000 men greater than 65 years old in the Physician’s Health Study II found that the equivalent of Centrum Silver versus placebo in this selected group of people did not find any difference in cognitive function on telephone assessment; however, in this highly-educated group, cognitive testing may be less sensitive to detect small changes: https://gmodestmedblogs.blogspot.com/2013/12/multivitamins-not-helpful.html
-- these studies on multivitamin supplements were observational and not RCTs as the current study
-- A study confirmed that physical activity as well as social engagement (social activity and support) were associated with improved cognition or slower rate of cognitive decline: https://gmodestmedblogs.blogspot.com/2015/12/lifestyle-interventions-and-cognition.html
-- a few studies on the Mediterranean diet found that this diet was associated with higher brain volume, especially in those consuming more fish and less meat. And the PREDIMED study found a high extra-virgin olive oil intake was associated with improved cognition: https://gmodestmedblogs.blogspot.com/2015/11/mediterranean-diet-and-brain-volume.html
-- This COSMOS-Mind study is a subsequent study after the initial COSMOS study finding benefit from Centrum Silver, but with more intensive on-site neurophysiological evaluation (the whole study was a randomized double-blinded placebo-controlled study assessing the role of Centrum Silver for prevention of cardiovascular disease and cancer: https://gmodestmedblogs.blogspot.com/2022/09/multivitamins-may-improve-cognition.html
-- this current COSMOS-clinic study was powered to detect an effect of at least 0.12 standard deviation units (SU) in the global cognition composite score. Both the MVM and placebo groups had significant improvements by year two, likely reflecting the fact that scores on neurophysiologic testing can improve on retesting, after the participants have already been through one battery of tests previously
-- it is a bit curious that the only way that they achieved statistical significance for global cognition was to add in thousands of people from their non-in-person substudies…..
-- and, for the primary outcome of global cognition, there was only a 0.01 SU difference
-- also, it was a bit strange that they added the COSMOS-Mind as well as the COSMOS-Web studies to the episodic memory recalculation and only the COSMOS-Mind to the global cognition recalculation
-- the AHEI-2010 is composed of 12 components: 9 components assessed dietary adequacy (foods that people should consume more of: total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and ratio of poly- and mono-unsaturated fats to saturated fats) and 3 components assess moderation (foods that people should consume less of: refined grains, sodium, and empty calories)
Limitations:
-- this study had a very select group of people: almost all were white and 60% had finished college, limiting generalizability to other groups
-- they comment that the relatively small degree of improvement they found in this study translates to being equivalent to reducing the effects on global cognition by two years of aging. They state “to assess the clinical importance of findings, we compared the observed estimate of the treatment effect with the estimate within the multivariable model for a one-year increase in participant age, because age is the most robust predictor of cognitive decline”. However, the measured rate of cognitive decline does vary by the individuals assessed in the groups studied (as noted, these tools are less sensitive for higher cognitive functioning individuals, such as in this COSMOS group). Since COSMOS itself did not assess how the differences in cognition between the 2 groups would translate into a difference in years of cognition, relying on historical changes from other studies with different people with different educations/comorbidities/diet/exercise is a bit fraught in making their 2-year comment
-- the AHEI-2010 is an accepted instrument for assessing the quality of food, though there were pretty broad ranges in their three tertiles, which may have obscured some differences in their categorizations
-- it was notable that on subgroup analysis, there was a trend to decreasing effect on cognitive health associated with MVM in those with increasingly healthy food consumption; and there was zero benefit in those who ate the healthiest!!!!
-- perhaps a more conclusive study would be to compare patients on MVM to patients on a Mediterranean diet plus supplemented EV00 olive oil, which has documented benefits for cognitive impairment, as noted above
-- their adding the COSMOS-Mind and COSMOS-Web individuals into their recalculation of the statistical benefit of MVM is really an apples/oranges thing: the COSMOS-clinic group had aggressive cognitive assessments with an array of on-site tests; adding in very large numbers of participants who had just telephone-based or computer-based evaluations as if they were equivalent seems a bit of a stretch. A prior paper mentioned that the COSMOS-Web group, for example, had 3 neuropsychological instruments, the ModRey, ModBent, and Flanker tests (https://www.sciencedirect.com/science/article/pii/S0002916523489046?via%3Dihub ), very different from the large array of tests given to those having on-site evaluations. So adding results of data from participants having very different platforms (on-site vs web-based) and using very different neuropsych tools seems to be pretty strange. And they did not comment on this in the paper
So, it is clear that the foods consumed do affect cognitive health. This is not particularly surprising since we know various components of foods and vitamins are necessary for proper cerebral functioning. The results of this trial can be interpreted and applied in different ways:
-- we could suggest that everyone over age 60 should be on a Centrum Silver or the equivalent (making sure that it was really equivalent), to hopefully prevent cognitive decline
-- or, we could consider doing that at an earlier age, since thinking quickly and the ability to recall information rapidly starts to decline around age 20
-- or, we could consider doing that at age 45, when researchers in France and the UK started to find cognitive decline (https://www.bmj.com/press-releases/2012/01/05/cognitive-decline-can-begin-early-age-45-warn-experts )
-- or, we could have a robust public health initiative to really promote healthy eating at a young age, making it considerably less attractive to eat junk food or drink sodas for example. this of course leads to backlash from the highly politically-connected fast food industry putting pressure on politicians to leave them alone (McDonalds did have a huge and expensive campaign to try to prevent eliminating trans fats from their cooking (though trans fats are multi-fold more atherogenic than saturated fats), citing that trans fats both improve flavor and increase the shelf-life of foods. The flaw in the argument was that they did a study in France, as I remember, showing that there was no detectable taste or shelf-life difference conferred by trans fats, and they changed to serving trans fat-free meals in France while persisting in trans fats in their American restaurants for many years….).
-- and, we do have a role in primary care (and medicine in general) in connecting with patients and helping them develop healthier eating habits, perhaps best through motivational interviewing; this is a significant uphill battle without a strong public health initiative, since significant numbers of patients are overexposed to advertising for quite delicious/unhealthy junk foods, or live in food deserts
-- per https://www.ers.usda.gov/webdocs/publications/45014/30940_err140.pdf , noting that “23.5 million people live in low-income areas that are further than 1 mile from a large grocery store or supermarket, and that 11.5 million of these people have low incomes themselves”, and even those not living in food deserts often have much easier access to corner stores/bodegas that are remarkably depleted of healthy foods
-- and, one really unfortunate development, is that fewer people seem to be cooking at home
-- a recent survey found that those 60% of Americans cook at home more than five days a week, and 84% rate hamburgers and cheeseburgers are the most popular dishes: https://www.testhut.com/how-many-americans-cook-statistics/
-- and, bottom line, healthy nutrition is extraordinarily important for the whole body, and not just for cognitive protection
-- MVM certainly may have a role for those who do not eat healthfully for whatever reason (cost of veges and fruits are much too high; cheaper to buy calorie-dense junk foods that are tasty and filling, etc. etc…..), but it is not clear from this study that those eating healthfully have cognitive benefit
geoff
-----------------------------------
If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu
to get access to all of the blogs:
go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order
-- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category
-- 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
if you would like to see the articles in this blog, please email me.
please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list
Comments
Post a Comment
if you would like to receive the near-daily emails regularly, please email me at gmodest@uphams.org