Improved A1c control was associated with lower mortality and cardiovascular morbidity

A recent large cohort study found that good risk factor control, and especially diabetes control, significantly lowered the morbidity and mortality risk of diabetes, in a study funded by the Swedish Association of Local Authorities and Regions (see dm control not inc mortality nejm2018 in dropbox, or DOI: 10.1056/NEJMoa1800256).

Details:

--271,174 patients with type 2 diabetes in the Swedish National Diabetes Register were matched with 1,355,870 controls on the basis of age, sex, and county

-- diabetic patients were assessed by: age categories and according to the presence of five risk factors:

    -- HgbA1c >7%

    -- systolic BP >140 or diastolic >80 mmHg

    -- albuminuria (micro or macro)

    -- smoking: being a current smoker

    -- LDL >97 mg/d

-- mean age 61, 49% women

-- median follow-up was 5.7 years, with 175,345 deaths: 37,825 in those with diabetes (13.9%) and 137,520 in controls (10.1%)

-- those diabetics with no risk factors had mean: A1c=6.21%, LDL 76.5 mg/dL, BMI 29.5, BP 123/70, and 62% on statins, 61% of antihypertensives

-- those diabetic with all 5 risk factors had mean: A1c=8.60%, LDL 141.1 mg/dL, BMI 30.9, BP 148/85, and 73% on statins, 38% of antihypertensives

Results:

-- Among patients with diabetes who had all five variables within target ranges, as compared with controls (where risk factors were not known):

    -- death from any cause: hazard ratio (HR) 1.06 (1.00 to 1.12)

    -- acute myocardial infarction: HR 0.84 (0.75 to 0.93)

    -- stroke: HR 0.95 (0.84 to 1.07)

    -- hospitalization for heart failure: HR 1.45 (1.34 to 1.57).

-- Among patients with type 2 diabetes, the excess risk of outcomes decreased stepwise for each risk-factor variable within the target range.

-- In patients with type 2 diabetes:

    -- HbA1c >7.0% was the strongest predictor of stroke and acute myocardial infarction, including in those with just this one sole risk factor

    -- smoking was the strongest predictor of death. low physical activity was second strongest predictor, followed by marital status and A1c

Commentary:

-- this study found an impressive role for improved diabetes control in cardiovascular disease: though they did use a low A1c cutpoint (7%), their cutpoints for other risk factors was also low (LDL 96 mg/dL, any smoking, any microalbuminuria, DBP <80). their lowest risk group (no risk factors) was pretty remarkably low risk, and even their highest risk group (all 5 risk factors) was not so bad

-- they also found the not-so-surprising result that the youngest cohort had the most cardiovascular and mortality relative risk attributable to their various combinations of risk factors

-- one problem with this study is that there were no data on risk factors in the control population, so diabetic patients with very good risk factor control were compared with a general nondiabetic population with potentially lots of other cardiovascular risk factors (ie: one cannot conclude that well-controlled diabetics have the same mortality as well-controlled non-diabetics; diabetics could still have an increased mortality associated just with well-controlled diabetes, see further comment below. also they did not consider changes in risk factors over the course of the study, and in evaluating a risk factor, they did not take into account whether the patient was on treatment or not

-- the data set had lots of holes: only 96,673 patients with diabetes had all five risk factors assessed (about 1/3 of their diabetics). there may well be a bias here: perhaps those diabetics with more risk factors were more likely to have the details of their risk factors recorded??

-- the excess risk of hospitalizations for heart failure is a bit surprising. on their more detailed analysis, however, they found that the predictors for heart failure  hospitalizations were atrial fibrillation, high BMI, A1C, and renal function. On looking at the sole risk factors associated with heart failure hospitalizations (eg, all other risk factors werre okay): BMI was overwhelmingly first, followed by A1c (ie, A1c >7% was still bad) and low physical activity. atrial fibrillation and BMI were not part of their identified pre-analysis risk factors; so their finding of higher hospitalization for heart failure may be because of many potential unaccounted for risk factors.

-- one concern i have always had about glucose intolerance/diabetes and cardiovascular disease is that it seems that even low levels of A1c is associated with increased cardiovascular risk. for example, the Atherosclerosis Risk in Communities study (ARIC) found that coronary heart disease risk was increased significantly by 23% in those with A1c from 5.5-6, and by 78% in those with A1c 6-6.5% (see https://www.nejm.org/doi/full/10.1056/NEJMoa0908359​ ). similar increased risk was found for stroke. As a  result of this and a couple of other epidemiological studies, i have been using even marginally elevated A1c levels as (yet another) marker of increased atherosclerotic risk. perhaps in the above study if we had data to compare diabetics with nondiabetics at similar low-risk by other risk factors, there would have been a residual risk for diabetics with A1c<7.0% (in fact, there was a marginal 6% increased overall mortality in those diabetics with all risks factors controlled vs the nondiabetics). But, the bottom line from this study (and an important advance in our knowledge base) is really that controlling A1c is perhaps the most profound risk-factor reducing intervention in diabetics, both in those with and without other risk factors.

So, this study does add some important information for us clinicians and patients.  

-- We do know that improved diabetes control is associated with decreases in microvascular outcomes (retinopathy, nephropathy).  

-- And we know now that improving the A1c has perhaps the most significant effect on the variety of cardiovascular outcomes (other than hospitalization for heart failure, though they did not control for the key risk factors they found).  

-- But, there may still be some residual cardiovascular risk with the diabetes itself.  I would expect that part of the issue is the focus on A1c levels, and that it still makes sense to broaden this perspective to focusing on diabetes medications which seem to decrease cardiovascular risk (which seem to be: metformin, GLP-1 agonists, probably pioglitazone, and probably SGLT-2 inhibitors though the likely increase in amputations is of concern). Insulin and sulfonylureas are probably much less good (see http://gmodestmedblogs.blogspot.com/2018/09/high-glycemic-index-diet-causes-obesity.html which documents some of the issues with hyperinsulinemia, typically associated with these meds). See http://gmodestmedblogs.blogspot.com/2018/03/loosening-a1c-goal-is-low-a1c-really.html , which defines my atherosclerosis-oriented approach to diabetes control in a bit more detail. 

One other personal note: even with an A1c level cutpoint of 7%, i am really surprised/impressed how many of my patients with A1c levels >9-10% range are <7% when i added a GLP-1 agonist to their baseline therapy (metformin, or also metformin/insulin), and in fact several have come completely off pretty high doses of insulin (in the 30-60 unit range)​

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 blogs since 8/15/17:

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 type in a name in the search box and get all the blogs with that name in them

to access older blogs from the BMJ website, from October 2013 until 8/15/17: go to http://blogs.bmj.com/bmjebmspotlight/category/archive/ 

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