diabetes and cognitive decline

A recent analysis of the Atherosclerosis Risk in Communities study (ARIC, a prospective community-based cohort study from 4 communities across the US) assessed 13351 black and white adults aged 48-67 in 1990-92, and assessed changes in cognitive function over the next 20 years.

baseline:

--population: average age 57, 56% female, 24% black, 13.3% with diabetes

--cognitive assessment: 3 tests -- delayed word recall test (DWRT, a test of verbal learning and recent memory); digit symbol substitution test (DSST, a test of executive function and processing speed), a component of the Weschler adult intelligence scale-revised; and the word fluency test (WFT, a test of executive function and language), all checked at baseline and at least 2 more times.

--they assessed stored blood for A1c values and controlled for age, race, education level, smoking, alcohol, BMI, hypertension, history of CAD, history of stroke, and apolipoprotein e4 genotype

results:

--20-year decline in cognitive function:

    --the adjusted difference in those with and without diabetes was pretty consistent: all of the tests were lower in diabetics, though only the DSST and ​WFT reached statistical significance (ie, most significant decreases were in the tests of executive function/processing speed)

    --the composite (global Z score) unadjusted was -0.15, and on adjusted analysis was -0.23 (a score of -0.15 is equivalent to the cognitive difference of a 60-year old vs a 55-year old)

    ​--there was a significant trend per A1c: the most dramatic decline was in those with A1c>7.0, less so in those with 6.5-7, and pretty similar in those either 5.7-6.4 as in those 6.5-7. all of these (including those 5.7-6.4) were significantly different from controls

    --longer duration of diabetes was associated with greater late-life decline in cognitive function

    ​--no difference between black and white patients

so, this was an observational study, making it hard to reach clearcut conclusions (ie, was the cognitive decline related to untested variables?  --  are the patients with prediabetes or diabetes fundamentally different from normoglycemic patients, such that they are predisposed to cognitive decline? are both diabetes and cognitive decline, for example, associated with lack of exercise, eating fewer anti-oxidant vegetables or other lifestyle variables??). however, there is important biological plausibility for the association between diabetes and cognitive decline: the decline in executive functioning/processing speed involves the subcortical microvasculature, causing damage in white matter pathways and subcortical gray matter, as is more likely to be prevalent in diabetics (eg, those with other microvascular diabetic disease, such as retinopathy, in some studies, have higher likelihood of cognitive dysfunction. macrovascular atherosclerotic disease can be associated with decreased cerebral perfusion).  one of my blogs done 6/21/12 reviewed a lancet review (see DM cognitive dysfunction lancet 2012 in dropbox, or Lancet 2012; 379: 2291–99), noting similar findings. data on reversal of cognitive decline by improved diabetes control are paltry (the ACCORD MIND study did not find anything, but was only a 3-year study in patients already age 63 and with frequent hypoglycemic episodes). one thing new in the current study is the association with prediabetes. this really brings up the issue that the primary goal is to prevent prediabetes and diabetes. several studies have shown that people at high risk of developing diabetes, or who have prediabetes, can significantly delay developing diabetes by around 10 years through lifestyle changes. i have personally had several patients revert from prediabetes (and even a few with diabetes) to a totally normal A1c through diet and exercise. but the "lifestyle" issue, as noted previously, is always complex, ranging from issues of personal preferences (shaped by a society that reinforces driving a few blocks instead of walking), access to safe and pleasant exercise venues (safe neighborhoods, sidewalks, etc), access to good food (issues of access in the "food deserts" of the inner city), income inequalities (less money for poorer patients to buy the more expensive fresh fruits and vegetables), etc etc etc -- ie, there really needs to be a clear social/public health imperative to focus on these "lifestyle" issues, which of course in the longrun are better for people and undoubtedly save lots of money from decreased medical costs of caring for the resultant medical disabilities.