Dietary intakes by levels of glycemic control for black and white adults with non-insulin dependent diabetes mellitus (NIDDM).

OBJECTIVES

The relationship between diet and glycemic control was examined among a racially mixed population of male and female adults with non-insulin dependent diabetes mellitus (NIDDM).

METHODS

Data from 3-day dietary records and glycosylated hemoglobin (HbA1c) were analyzed for two-hundred eighty two patients of a Family Practice Ambulatory Care Unit and a community-based health center in Winston-Salem, North Carolina. Correlations were calculated for individual nutrients to determine their strength of association with glycemic control. Analyses by tertiles of HbA1c were also conducted for each race/gender group. Regression analysis was used to determine independent dietary predictors of HbA1c.

RESULTS

For all subjects, energy, energy per kilogram of body weight, fat, carbohydrates saturated fat, and cholesterol were significantly correlated with HbA1c. Nutritional differences across tertile levels of HbA1c for all subjects were not significant. For black females, consumption of energy, protein and fat was significantly higher among upper tertile subjects compared to the lowest tertile; and intake of energy, protein, fat, and saturated fat was significantly correlated with HbA1c in this group. For black males, energy intake was highest among upper tertile subjects compared to those in the middle tertile, while energy per kilogram of body weight, and percent of calories from protein, were significantly correlated with HbA1c. For white males, energy intake expressed as a function of body weight was highest among subjects in the upper tertile and a significant positive correlation with HbA1c was observed. No relationship between nutritional intake and HbA1c was found among white females. Racial differences in nutrient intake were also compared for males and females in the upper tertile of HbA1c. Black females in the upper tertile consumed significantly more energy, protein, and significantly less dietary fiber per 1000 kilocalories. No significant differences were observed between black and white males in the upper tertile, although higher cholesterol consumption in black males compared to white males approached significance. Regression analysis revealed that total energy intake significantly predicted HbA1c for all subjects and all white subjects, while a similar observation was made for total fat intake among all black subjects and among black females.

CONCLUSIONS

These findings confirm that diet, especially total energy, is an important contributor to glycemic control. Dietary fat is also associated with glycemic control among blacks, especially black females, who are especially prone to more dire health consequences of NIDDM. Strict monitoring of diet should lead to improved glycemic control and less mortality and morbidity in this population.