Diet and CHD risk in diabetic women Richard M. Steingart, MD

Cardiology Review® Online, January 2004, Volume 21, Issue 1

One cannot attend a cardiovascular meeting without hearing about the epidemic of diabetes in our progressively older and overweight population. Invariably, the audience gasps when maps of the United States show, state by state, year by year, how dramatically the prevalence of obesity, and with it, type 2 diabetes, is growing.

A headline from The New York Times of September 28, 2003, speaks to the impact of the problem on one important aspect of the health care continuum.1 The article, titled “On the final journey, one size doesn’t fit all,” refers to the growing demand for oversized caskets, and the general impact of obesity on the funeral industry. Examples given include the Woodlawn cemetery in the Bronx, which recently increased the width of its standard burial plot from 3 feet to 4 feet wide to accommodate wider burial vaults. The Cremation Society of America is training its practitioners in the handling of obese bodies. Hearse makers are producing wider vehicles. Despite these changes, the reporter notes that critics say the funeral

industry hasn’t done enough…. We should all do more.

The study by Hu (page 32) makes a contribution toward solving some of the health problems associated with obesity. In his analysis of questionnaires from a well-characterized group of diabetic nurses in the Nurses Health Study with an average age between 47 and 50 years, there was a quantitative relationship between fish and fish oil consumption, coronary heart disease (CHD), and total mortality.

In addition to age, a multivariate analysis using pooled logistic regression simultaneously controlled for cigarette smoking, body mass index (BMI), menopausal status and postmenopausal hormone use, physical activity, alcohol use, history of hypertension, high cholesterol, and other cardiovascular risk factors. Relative risks ranged from 0.70 for fish consumption one to three times monthly to 0.38 for consumption five or more times weekly. The author states that further adjustment for other dietary factors did not appreciably alter the results.

The absolute event rates were low. There were 45,845 person-years of follow-up, with 326 incident cases of CHD and 468 deaths from all causes (161 from CHD or stroke). For comparison purposes, it would be interesting to know the event rates among nondiabetic nurses. Their event rate is likely too low to determine the impact of fish consumption on CHD and mortality.

From table 1 of the original paper, the average BMI of the diabetic women was 27.8 to 28.7 over the range of reported fish consumption (a BMI over 25 is considered overweight, over 30, obese).2 Women with higher levels of fish consumption exercised more, drank less alcohol, ate less trans and saturated fat, and were less often currently using estrogens, to name but a few of the differences among the fish consumption groups. Hu controlled for these differences in the regression analyses, but one is reminded of the cohort studies of postmenopausal hormone use. In those studies, women taking estrogen were consistently thinner, more fit, and from higher socioeconomic strata than those not taking estrogen.3 Although these analyses also mathematically controlled for differences between the groups, recent prospective randomized trials of estrogen therapy clearly demonstrated no beneficial effect of postmenopausal hormones; in fact, there was a potential for harm.4 These results are exactly opposite those suggested by the cohort studies.

Hu concludes that regular fish consumption can be included as part of a healthy diet for diabetic management, and acknowledges that controlled trials of fish oil supplementation for preventing CHD and mortality among diabetics are clearly needed. Certainly, we should not lose sight of the primary goal of preventing type 2 diabetes through exercise and calorie limitation. Once the disease is manifest, advising diabetic patients to exercise, eat fewer calories, and eat more fish seems prudent.