stress decreases cognitive function

 

A recent article confirmed a strong relationship between stress and cognitive decline in older US adults (see stress dec cognitive function JAMA2023 in dropbox, or doi:10.1001/jamanetworkopen.2023.1860)

 

Details:

-- 24,448 participants in the Reasons for Geographic and Racial Differences in Stroke (REGARDS study), a national population-based cohort of Black and White participants at least 45 years old sampled from the US population, recruited between 2003-2007; all had ongoing annual follow-up and analysis from 2021-2022

    -- the REGARDS group was selected from a commercial nationwide list from Genesys  (a commercial database of the US population)

    -- REGARDS oversampled Black participants as well as those living in the southern US states where the stroke risk is highest (“stroke belt”) and the coastal plains of North Carolina, South Carolina, and Georgia, where there is a higher stroke mortality (the “stroke buckle”), given their primary mission of assessing stroke incidence. And they wanted to make sure enough Black participants were involved in order to have a statistically sufficient sampling

-- 60% women, median age 64, 42% Black/58% White, 12% not finish high school/26% graduated HS/27% some college/35% graduated college, annual income 18% <$20K/24% $20-34K/30% $35-74K/16% >$75K

-- CVD risk factors: hypertension 59%, diabetes 22%, dyslipidemia 56%

-- physical activity none 35%/1-3x a week 36%/more 29%,  BMI normal 24%/overwt 36%/obese 39%, smoking current 15%/past 39%/never 47%; alcohol heavy 4%/moderate 33%/none 63%

-- 5589 (23%) reported elevated levels of stress

-- data was collected by telephone, self-administered questionnaires, and in-home examinations

-- perceived stress was measured using a 4-item version of the Perceived Stress Scale (PSS), assessed at baseline visit and during one follow-up visit

    -- this 4-item scale assessed stress indicators: inability to control important things in one’s life, not confident in ability to handle personal problems, feeling that things were not going one’s way, feeling that difficulties were accumulating and could not be overcome

-- cognitive function was assessed with the Six-Item Screener (SIS)

    -- SIS is a brief measure of global cognitive status: 3-word recall and orientation to year, month, and day of week, so total score of 0-6 (it is validated, and a score of <4 has sensitivity of 74-84%, specificity of 80-85% for cognitive impairment in community-based individuals)

    -- incident cognitive impairment was defined as a shift from an SIS score >4 at the first assessment to <= 4 subsequently

    -- Prevalent cognitive impairment was score <=4 at both times

 

Results:

-- perceived stress and poor cognition: adjusting for sociodemographic variables (age, sex, education, race, income), CVD risk factors (hypertension, diabetes, hyperlipidemia), lifestyle factors (exercise, BMI, smoking, alcohol), and depressive symptoms:

    -- 37% higher odds for poor cognition, aOR 1.37 (1.22-1.53), evaluating perceived stress as a binary variable (<= 4 vs >4)

    --  4% higher per unit change in PSS, aOR 1.04 (1.03-1-06), with PSS score as a continuous variable

    -- stratifying by sex and race: White men, White women, Black men, and Black women all had an adjusted OR of 1.04-1.05 [ie no significant difference by sex or race in the adjusted model]

-- an increasing PSS score was associated with an increasing cognitive impairment:

 

-- Incident cognitive impairment (ICC), PSS score > 4 at baseline to <= 4 at follow-up (data available for 16,150 individuals):

    -- adjusting for all of the above, in binary model of PSS: 39% increased odds, aOR 1.39 (1.22-1.58)

        -- dietary data was available in 21,000 participants:  adjusting for nutrition did not alter the results

    -- adjusting for all variables: for each one unit of change of perceived stress, there was associated a 4% increased odds of cognitive decline, a OR 1.04 (1.02-1.06).

        -- adding depressive symptoms to the adjusted model was not associated with any changes in cognitive function

        -- there was a dose-response in participants with persistent stress (aOR 1.24 (0.95-1.63)), or new stress (aOR 1.03 (0.81-1.32)), versus no stress (aOR 0.81 (0.68-0.97))

-- and, there was no interaction between perceived social stress score and age, race, or sex

 

Commentary:

-- estimates are that >57 million people globally have dementia, with estimates of a staggering increase to 153 million by 2050. Up to 40% of dementia risk may be associated with modifiable risk factors

    -- there are estimates that a 10-25% reduction in modifiable risk factors (including behavioral ones) globally might prevent >1.3 million cases of Alzheimer's disease.

-- there have been many studies over the past 6 decades on the relationship between stress and disease:

    -- http://gmodestmedblogs.blogspot.com/2021/11/mental-stress-and-heart-disease.html , which reviews a couple of recent articles on the cardiovascular effects of stress, including likely mechanisms:

        -- the elaboration of stress hormones (catecholamines and cortisol) 

        -- the effect of cortisol on pretty much all other hormones in the body (i.e. cortisol alters essentially all of the other hormonal systems), with their attendant effects 

        -- many potential direct effects on the heart: increased platelet activation, increased fibrinogen levels, stress cardiomyopathy (“acute myofibrillar degeneration”), increased myocardial sensitivity to catecholamines by higher cortisol levels, increased ventricular ectopy (presumably from sympathetic stimulation), coronary artery vasoconstriction, endothelial dysfunction, inflammation, increased ventricular mass (for example a study in workers put in stressful jobs showed increased left ventricular mass, presumably related to sympathetic stimulation): see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2633295/ for a review of stress and cardiovascular disease 

        -- effects on the immune system: changes in cytokine production, decreased antibody production in response to vaccines, decreased cellular responses, decreased natural killer (NK) cell activity (important for surveillance/prevention of cancer and infections): see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1361287/ or https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3019042/ ) 

        -- A review article found that hypnosis and relaxation improve immune responses (see hypnosis relaxation and improved immune function stress2002 in dropbox, or Gruzelier JH. Stress (2002) vol 5 (2): 147-163) [there are many articles finding that strong support systems decrease the physiological effects of stress; this article reinforces that relaxation and hypnosis help] 

        -- an impressive old study found that intradermal injection of a sensitizing agent (which typically led to a type 4, cell-mediated response, as is found with a positive PPD) in one arm vs saline in another led to the expected reaction with the sensitizing agent in the control group. But, in those hypnotized to believe that the placebo saline injection was the actual sensitizer, there was a typical appearing positive PPD-like response on the saline arm (though biopsy of the other side with the true sensitizer did show appropriate cellular infiltration but without the apparent PPD-like bump/erythema). Pretty interesting…..  (I believe the study was in the 1960s, though cannot find the actual reference in this well-before-internet article) 

        -- and, perhaps more relevant to the issue of cognition, stress is associated with major changes in the brain hippocampus (important for learning, memory encoding, memory consolidation, etc), including even large-scale decreases in hippocampal volume, which seems to be related to cortisol elevation in many (but not all) studies (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4561403/ ). Perhaps related, there is also hippocampal atrophy in those with depression (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2577751/ ), which shares with stress both the frequent increase in cortisol levels, as well as both being inflammatory states (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7327519/ ) 

        -- as noted in  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4561403/, the role of cortisol is somewhat murky. There may well be multiple consequences of stress that impact cognition: 

            -- perhaps catecholamines play a role. Or perhaps cortisol can be a mediator of other hormone changes (which it is), and these other hormone changes (which might also occur through stress, independent of cortisol), are the bad actors here 

            -- stress can be associated with inflammation, both peripherally and in the brain (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5476783/ ), including depletion of neurons in the hypothalamus, which can lead to decreased memory formation. 

                -- a very recent article found that air pollution may be associated with dementia (see air pollution and dementia BMJ2023 in dropbox, or https://www.bmj.com/content/bmj/381/bmj-2022-071620.full.pdf ). Air pollution, and specifically small particulate matter, is associated with chronic inflammation and heart disease (eg see http://gmodestmedblogs.blogspot.com/2016/06/air-pollution-and-heart-disease.html ). So, this study reinforces that chronic inflammation may be another major culprit leading to dementia.  

 

-- other potential mediators of memory loss with stress 

    -- Another potential mechanism here is through sleep deprivation, which is a not-so-uncommon consequence of perceived stress, and is associated with changes in hippocampal function and memory loss (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3768199/  ) 

        -- my guess is that many of us trained in medicine have experienced some significant sleep deprivation, and perhaps noted some decrease in memory recall or encoding new memories.....

    -- chronic stress can also be associated with worsening of unhealthy behaviors, such as smoking (nicotine does have a relaxing effect), eating less well, decreasing exercise, decreasing medication adherence...

-- the issue of stress is certainly complex. Different people experience similar events (in the lab or real world) differently. Some of the reason is that it is clear from many prior stress studies as noted above that social support is a strong moderator of perceived stress, with much lower levels when support systems are functional. This is an important reason in stress studies that the focus not be on the quality of the stress but on the perception of the stress by an individual, which has been usually been assessed through questionnaires or sometimes by biological markers (increases in catecholamines or cortisol levels)

 

-- This current article found a few things:

    -- there is an independent relationship between perceived stress and cognition

    -- this relationship has a dose-response: increasing perceived stress is associated with increased cognitive impairment

    -- though Black individuals were found to have more cognitive dysfunction, this was attributable to the increased perceived stress in their lives

    -- there was an increase in cognitive impairment both in those entering the study with some deficits but also in those who developed cognitive impairment (incident cases) associated with perceived stress

-- previous studies have found similar results, though they involved many fewer participants, were cross-sectional, and did not have enough of a racial mix to address the role of race (the current study did oversample Black individuals, comprising 42% of their participant database, in order to assess this relationship more rigorously). Other findings in the prior studies include:

    -- the levels of stress increase in a linear manner with increasing age

    -- middle-aged adults with greater job strain/stress (decreased control over their job) had higher risk of cognitive impairment later in life

    -- other publications from the REGARDS study have found that perceived stress is associated with incident coronary heart disease, all-cause mortality, and atrial fibrillation

 

Limitations:

-- There was a recruitment bias in this study: potential participants were recruited through a mailing, follow-up telephone calls, and then by in-home visits: a process that would not include all comers

-- the symptoms were self-reported, so might not be so accurate

-- many of the assessments were binary: eg alcohol consumption as moderate (1-14 drinks for men, 1-7 for women) or heavy (more than those limits), and people were lumped into these categories (so, is there really a big difference between 14 and 15 drinks?  My guess is that 1 vs 14 drinks would reveal a bigger difference)

-- the participation rate in this study was only 49% (similar to rate in other large national cohorts): this might limit the generalizability to the population overall

-- there was only one post-baseline assessment of their perceived stress instrument (SIS), and perceived stress does change over time; and they used a 6-point scale for cognitive decline (though the sensitivity and specificity for mild cognitive impairment with SIS is similar to the full min-mental state examination (MMSE))

-- as with all such large studies finding an association with specific outcomes:

    -- there could be unmeasured confounders that play a strong role, so we are not able to project causality, only an association

    -- there could be reverse causation: did cognitive decline lead to increased perceived stress??? The fact that this was a long-term detailed study, that they found that incident cognitive decline was also associated with perceived stress, and that there was a dose-response curve (the more stress, the more cognitive decline), makes reverse causation less likely

 

So, several issues here:

-- Stress is omnipresent in our society, and is clearly associated with many different medical conditions, as above

-- social supports do provide an important moderator of the effects of stress on our bodies (and minds)

-- there is also good evidence that exercise helps, as well as several other interventions (mindfulness, cognitive behavioral therapy, yoga, tai chi….)

-- one important finding above is the further confirmation that Black/White outcome differences are related to structural racism in our society; perceived stress is at a much higher level in people of color, independent of their socioeconomic status. When controlling for the many variables related to stress, the cognitive outcomes in the above study were no different by race: ie, further supporting that race is a social construct in our society and the adverse consequences of stress are not a genetic issue. other studies have also found that to be true, for example for hypertension. So, the overall imperative is for us to change that adverse/perverse social construct……

 

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 (2 options):

1. 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

2. go to https://bucommunitymed.wpengine.com/, a website from the Community Medicine section at Boston Medical Center.  This site does have a very searchable and accessible list of my blogs and is the easiest to view blogs and displays more at a time

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