Skip the Weigh-in for Type 2 Diabetes Patients?

Gregory A. Nichols, PhD


June 29, 2015

The Obesity Paradox in Type 2 Diabetes Mellitus: Relationship of Body Mass Index to Prognosis

Costanzo P, Cleland JG, Pellicori P, et al
Ann Intern Med. 2015;162:610-618

Does overweight/obesity really protect people with diabetes? To examine this controversial issue, known as the "obesity paradox," an international group of investigators used data from England's National Health Service to identify a large prospective cohort of patients with diabetes and no known history of cardiovascular disease (CVD). They followed this cohort for a median of nearly 11 years to ascertain all-cause mortality, with a secondary outcome of hospitalizations for CVD events. The main explanatory variable was body mass index (BMI) assessed at baseline. Consistent with other studies, they categorized BMI as low (< 18.5 kg/m2), normal (18.5-24.9), overweight (25-29.9), obese (30-34.9), and very obese (≥ 35). Cox regression analyses adjusted for age, sex, duration of diabetes, baseline systolic blood pressure, smoking, and comorbid conditions, including COPD, cancer, chronic kidney disease, and previous CVD.

Increased CVD Risk but Lower Mortality Risk

Of the 10,568 patients, 20% were normal weight, 38% were overweight, 25% were obese, and 17% were very obese at baseline. Compared with patients of normal weight, overweight patients had a 13% lower mortality rate. Mortality among patients who were obese or very obese did not differ from normal-weight patients; however, risk for CVD hospitalization events, including acute coronary syndrome and heart failure, was greater for overweight/obese patients. The analyses did not account for BMI changes that may have occurred during follow-up, fitness (either at baseline or during follow-up), differences in pharmacotherapies, or other key risk factors such as cholesterol levels.


Cardiorespiratory Fitness and Incident Diabetes: The FIT (Henry Ford Exercise Testing) Project

Juraschek SP, Blaha MJ, Blumenthal RS, et al
Diabetes Care. 2015;38:1075-1081

How important is fitness in reducing the risk for diabetes? The Henry Ford Exercise Testing (FIT) project was a large-scale study of patients who underwent physician-referred stress testing between 1991 and 2009. After excluding patients with a history of coronary artery disease, heart failure, or diabetes, the researchers followed 46,979 patients for a median of 5.2 years to determine how many patients developed new onset of diabetes.

The main analysis variable was cardiorespiratory fitness expressed in metabolic equivalents (METs) as calculated from a treadmill stress test, and divided into four ascending fitness groups: < 6, 6-9, 10-11, and ≥ 12. Using nested Cox proportional hazards models, the association between fitness and incident diabetes was analyzed, adjusting for a wide range of covariates, including age, sex, race, smoking, hypertension, hyperlipidemia, physical activity, and history of obesity. One sensitivity analysis examined the association across strata of glycated hemoglobin (A1c), and another also adjusted for BMI in the subpopulation that had BMI data available.

Fitness Matters

During follow-up, 14.6% of patients were newly diagnosed with diabetes. Compared with patients with METs < 6, the hazards for developing diabetes among those with METs 6-9 were not significantly different, but those with METs 10-11 had a 23% lower risk for incident diabetes, and those with METs ≥ 12 had a 54% lower risk after full adjustment. Each additional MET unit was associated with an 8% lower risk. These results were not different among those with and without a history of obesity, and there was no statistical difference across strata of A1c. In the sensitivity analysis of those with BMI data, adjustment for BMI somewhat attenuated the results, but the findings were not meaningfully changed.


Analysis and Commentary

The "obesity paradox," in which overweight or even obese patients experience better survival than normal-weight patients, has been a puzzling phenomenon for scientists and laypeople alike. While the relationship between CVD and obesity is largely unquestioned, the paradox emerges once CVD develops, where obesity seems to confer a survival advantage.[1] These unexpected findings have been reported in general populations and in patients with diabetes, leading us to believe that the paradox must be real and applicable to specific subpopulations.[2,3,4] Indeed, Costanzo and colleagues include a table of 16 studies investigating the relationship between obesity and mortality in type 2 diabetes, the sum of which leads to inconsistent and contradictory conclusions but generally supports the notion of the paradox, or at least no association between obesity and mortality. The authors point out that each of the 16 studies contained methodologic flaws that their study does not, but that is not to say that the study by Costanzo and colleagues is flawless.

Two important shortcomings are critical to the interpretation of the study. First, obesity as measured by BMI was assessed at baseline with no accounting for BMI changes during follow-up. Although this is a common approach in epidemiology, BMI trajectories may be a more important predictor of outcomes. To assume that BMI is constant among patients who are frequently (we hope) being counseled to lose weight while at the same time being prescribed drugs that induce weight gain (such as insulin and sulphonylureas)[5] is heroic and likely inaccurate.

A second shortcoming of the Costanzo study is that it could not account for fitness level. The FIT study gives us some sense of how important this is, at least with respect to diabetes incidence. No one would argue that obesity and elevated glycemia are the two primary risk factors for diabetes.[6,7] Nevertheless, the FIT study found that greater fitness substantially reduced the risk of developing diabetes, an important finding that did not differ by history of obesity, A1c level, or BMI. The authors briefly mention that their findings were limited by the primary entry criterion—referral for a treadmill stress test—which would probably place the study population at elevated risk for CVD. Because CVD is itself a risk factor for developing diabetes,[8] it is possible that these findings would not apply to a more general population. Still, the results were consistent across all analyzed strata, suggesting an inherent robustness.

The main message to take from examining these two studies in parallel is that fitness may at least partially explain the obesity paradox, and our understandable focus on body weight could be a misleading aspect of the weight/health relationship. Patients are counseled to lose weight through diet and exercise to prevent diabetes and CVD, but weight loss is often hard to achieve and even harder to maintain. Perhaps the message should be to increase fitness regardless of whether weight loss accompanies the exercise. Ignoring the scale and striving to walk a little further and faster each day may be a more achievable and sustainable goal that leads to lower CVD and diabetes risk.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.