Diabetes: lifestyle changes longer-lasting than meds

​​a recent systematic review and meta-analysis assessed the long-term sustainability of medical vs lifestyle modification (LSM) in diabetes prevention in those at high diabetes risk, finding LSM to be more effective (see 10.1001/jamainternmed.2017.6040).

Details:

--43 studies included, all RCTs in adults at high risk of diabetes. 49,029 participants, 19 tested meds/19 LSM/5 both. studies were from 1990-2015 and had to be >6 months duration

--mean age 57, 48% male

--studies lasted 0.5-6.3 years; 3 of the studies had extensive follow-up (time is from randomization): US Diabetes Prevention Program (DPP, 10 years), Finnish Diabetes Prevention Study (13 years) and the Da Qing Diabetes Prevention Study (20 years)

--meds used: weight loss meds (orlistat, phentermine-topiramate); RAS blockers (ACE-I/ARB), lipid-lowering (benzafibrate, no statins), insulin sensitizers (metformin, glitazones including rosiglitazone in 2 studies, pioglitazone in 1), insulin secretagogues (glipizide, nateglinide), insulin, hormone therapy (estrogen/progestin), and a-glucosidase inhibitors (acarbose, voglibose)

Results:

--at the end of the active intervention:

    --LSM was associated with a 39% relative risk reduction (RRR) for diabetes, RR 0.61 (0.54-0.68):  7.4 cases of diabetes/100 person-yrs vs 11.4 cases/100 person-yrs in the control group, NNT 25 persons to prevent one case of diabetes. dietary strategies alone were associated with a 32% RRR​; physical activities alone (n=2) had no significant effect, but combined diet and physical activity was associated with a 41% RRR

    --meds were associated with a 36% RRR for diabetes, RR 0.64 (0.54-0.76):​ 5.4 cases/100 person-yrs with meds, 9.4/100 person-yrs in controls, NNT 25 persons . all of the interventions except the insulin secretagogues and hormone therapy led to significant improvement: most pronounced with weight loss (3 studies, RRR 63%), insulin sensitizers (7 studies, RRR 53%), a-glucosidatse inhibitors (6 studies, RRR 38%), fibrate (1 study, RRR 32%), insulin (1 study, RRR 23%). 

--sustainability of diabetes prevention:

    -- LSM associated with 28% RRR​, RR 0.72 (0.60-0.86), after mean follow-up of 7.2 years after the intervention stopped. all of the intervention studies individually had statistically significant reductions in diabetes rates at that time. 

    --meds were associated with no sustained risk reduction, after mean follow-up of 17 weeks​. None of the interventions were statistically significant except the US DPP with metformin, but the follow-up was only 2 weeks after the study ended.

--the results were similar independent of the researchers' assessment of the quality of the individual trials

Commentary: 

--notable aspects of this study were: 

    --of the lifestyle interventions, the one that had the most effect was the combination of diet and exercise.  Exercise by itself did not do much, but there were only 2 studies, one with a significant 37% diabetes risk reduction and one with so few cases of diabetes that it could not come to any conclusion. though the important point here is that exercise does so much good for the body overall (associations with decreased heart disease/vascular disease in general, decreased cancer, longer life expectancy, better mental health.....) that it is important to stress exercise independent of the diabetes risk. And in this study, there was a synergy between diet with exercise in terms of decreased risk of diabetes.

    --of the medical interventions, the most impressive was weight loss (which, at least in some cases, can be done with lifestyle interventions.....).  then insulin sensitizers (most of which also are more cardioprotective than other medical interventions, except rosiglitazone, which is basically off the market), then the fibrate (that was the only lipid lowering agent tested, though as per other blogs statins may increase diabetes risk. a new blog to follow in a few days).

--the meds did not have lasting effect after the study stopped. this study basically confirms that diabetes meds do not by themselves alter the natural history of the progression of prediabetes. Not so surprising. One lingering concern I have is that by prescribing meds, patients may well think that lifestyle changes are not necessary (ie, we may be undercutting the LSM messages inadvertently, since the A1c is better: the patient, as well as clinician, is happy that the A1c is lowered, seeing that as the goal. Studies have shown, for instance, that patients on statins often stop their prior LSM to lower their LDL. They miss the bigger picture of the overall healthful value of LSM, focusing on that one aspect)

so, perhaps the biggest take-home message is that lifestyle changes really can be long-standing, even after the study reinforcers (clinicians, research assistants) stop hounding the patients about keeping up with these lifestyle changes. And, I think that we in primary care sometimes undercut these potential life-changing interventions. They take a lot of time (and we all know that we have very limited time…), and they often do not work (it is challenging for many patients to make basic changes:  these changes might be contrary to many individual cultural/environmental issues, such as not-safe neighborhoods to exercise because of poor infrastructure/crumbling sidewalks/unfriendly traffic patterns/being physically threatening, inability to change cooking styles, non-availability to good foods/food deserts, negative effects of advertising or strong reinforcement in the community to smoke/drink/eat junk food, etc). And I think these are especially concerning to new primary care clinicians, who do not have really long-term relationships with patients, and are not in a position to see these benefits (which can take a long time to effect).  But, as one of those long-term primary care clinicians, I can attest to the fact that some patients are really able to change for the long-term, typically after lots of back-patting, motivational interviewing, repeated messages, and time spent.  But these changes are really really satisfying and empowering both to the patients and to us.