Now accepting Telehealth appointments. Call the practice to schedule a Virtual Visit.

Low-fat Diets Do Not Lower Diabetes Risk

A review of an important, recently published scientific article selected by The North American Menopause Society (NAMS) and published in their newsletter.

Tinker LF, Bonds DE, Margolis KL, et al. Low-fat dietary pattern and risk of treated diabetes mellitus in postmenopausal women: the Women’s Health Initiative randomized controlled Dietary Modification Trial. Arch Intern Med 2008;168:1500-1511. Level of evidence: I. 

A low-fat dietary pattern, in and of itself, does not lower the risk for type 2 diabetes mellitus in healthy postmenopausal women. Any reduction in incidence is instead from weight loss, found the randomized, controlled Dietary Modification Trial of the Women’s Health Initiative (WHI). The trial included 48,835 healthy postmenopausal women (aged 50-79 y; average, 62 y) from across the United States, who were randomized to either a low-fat diet intervention group or a usual-diet comparison group. The low-fat diet aimed for 20% of energy intake from fat, but did not promote reduced overall energy intake, a weight loss goal, or a physical activity goal. Diet was monitored via a food frequency questionnaire; the women were followed for 8.1 years. The objective of the trial was to measure the effect of a low-fat diet alone, without the goals of exercise or weight loss, on the incidence of diabetes in healthy postmenopausal women.

At baseline, there were no differences in dietary intake between the two groups. At 1 year, the intervention group reported lower dietary fat as a percentage of energy and a higher intake of vegetables, fruits, and grain. In addition, weight, body mass index (BMI), waist circumference, systolic and diastolic blood pressures, and serum glucose levels were lower in the intervention group (< 0.001) at 1 year. After 6 years, weight and BMI remained lower in the intervention group (< 0.001). At study’s end, 7.1% of women in the intervention group and 7.4% in the comparison group reported a new diagnosis of diabetes (hazard ratio, 0.96; 95% confidence interval, 0.90-1.03; = 0.25). There was a significant effect of percentage of energy from fat on incidence of diabetes (= 0.04), but after adjustment for weight change, the effect on incidence was no longer significant. Women in the intervention group lost a mean of 2.2 kg after the first year, with a 0.8-kg weight loss maintained by the end of the study. The comparison group lost a mean of 0.1 kg.


The incidence of type 2 diabetes has increased dramatically in the United States, tripling to 1.4 million new cases in 2005 compared to 493,000 in 1980.1 This rise has been clearly linked to the increase in overweight and obesity, as even small increases in weight within the normal ranges of BMI result in an increased risk of developing type 2 diabetes.2, 3  Evidence shows that modest weight loss, as little as a 5% to 10% reduction in body weight, reduces incidence of type 2 diabetes in those with impaired fasting glucose.4, 5

What is less clear is whether a certain macronutrient composition will affect the incidence of diabetes independent of weight loss. Studies are conflicting on the effect of dietary fat. These studies, as in the Tinker study, are often confounded by weight loss, which mitigates the effect of macronutrient composition. Complicating the issue is the differential effect of the type of fat—incorporation of mono- and polyunsaturated fats in place of saturated and trans fat appears to have a protective effect on the development of diabetes. In the Prevencion con Dieta Mediterranea study,6 investigators showed fasting glucose, fasting insulin, and insulin resistance to be significantly reduced in a Mediterranean diet fortified with olive oil or tree nuts versus a low-fat diet despite an insignificant change in weight and BMI between the intervention groups, suggesting that the macronutrient content of diets may be important in the management of diabetes.

The Tinker study control group did not change percentage of fat intake from baseline (36.9% vs 37.7%), whereas the intervention group reduced fat intake to 28.6% but notably did not meet the goal of a 20% low-fat dietary pattern. The intervention group did have reduced incidence of diabetes, which was not significant when adjusted for the weight lost. The conclusion made by the authors was that weight loss rather than macronutrient composition is the dominant predictor of reduced risk of type 2 diabetes. The inability of the intervention group to reach the target despite intensive nutritional and behavioral modification training argues against the long-term feasibility of a 20% lowfat dietary intervention.

However, based on the current evidence, we cannot conclude that if this target were achieved, it would not protect against type 2 diabetes. It is fair to state that weight loss should be the main goal of medical nutrition therapy for the primary prevention of type 2 diabetes. What is not yet supported by the literature is whether a certain macronutrient composition, namely a 20% low-fat dietary intervention, has a significant effect on the incidence of diabetes.


Adrienne Youdim, MD
Director, Medical Weight Loss
Cedars-Sinai Center for Weight Loss
Assistant Clinical Professor
UCLA School of Medicine
Los Angeles, CA



  1. Centers for Disease Control and Prevention (CDC), National Center for Health Statistics, Division of Health Interview Statistics, data from the National Health Interview Survey. Available at: 6 nhis.htm. Accessed October 21, 2008.
  2. Chan JM, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. Obesity, fat distribution, and weight gain and factors for clinical diabetes in men. Diabetes Care 1994;17:961-969.
  3. Colditz GA, Willet WC, Rotnitzky A, Manson JE. Weight gain as a risk factor for clinical diabetes mellitus in women. Ann Intern Med 1995;122:481-486.
  4. Tuomilehto J, Lindstrom MS, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 2001;344: 1343-1350.
  5. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393-403.
  6. Estruch R, Martinez-Gonzalez MA, Corella D, Salas-Salvado J, Ruiz-Gutierrez V, Covas MI, et al. PREDIMED Study Investigators. Effects of a Mediterranean- style diet on cardiovascular risk factors: a randomized trial. Ann Intern Med 2006;145:1-11.
Dr. Adrienne Youdim

You Might Also Enjoy...

The Things We Can Control

The world is reminding us right now that there is much we cannot control. True, there is much we cannot control but that does not mean we are powerless.


Tis' the season for resolutions!  But creating change like creating a new habit can be difficult. Sometimes we falsely believe that change-makers always LOVE to do that thing that they do.