Missed opportunities for providing low-fat dietary advice to people with diabetes

Abstract

Introduction: Because cardiovascular disease is closely linked to diabetes, national guidelines recommend low-fat dietary advice for patients who have cardiovascular disease or are at risk for diabetes. The prevalence of receiving such advice is not known. We assessed the lifetime prevalence rates of receiving low-fat dietary advice from a health professional and the relationship between having diabetes or risk factors for diabetes and receiving low-fat dietary advice.

Methods: From 2002 through 2009, 188,006 adults answered the following question in the Medical Expenditure Panel Survey: "Has a doctor or other health professional ever advised you to eat fewer high-fat or high-cholesterol foods?" We assessed the association between receiving advice and the following predictors: a diabetes diagnosis, 7 single risk factors for type 2 diabetes, and total number of risk factors.

Results: Among respondents without diabetes or risk factors for diabetes, 7.4% received low-fat dietary advice; 70.6% of respondents with diabetes received advice. Respondents with diabetes were almost twice as likely to receive advice as respondents without diabetes or its risk factors. As the number of risk factors increased, the likelihood of receiving low-fat dietary advice increased. Although unadjusted advice rates increased during the study period, the likelihood of receiving advice decreased.

Conclusion: Although most participants with diabetes received low-fat dietary advice, almost one-third did not. Low-fat dietary advice was more closely associated with the total number of diabetes risk factors than the presence of diabetes. Increasing rates of diabetes and diabetes risk factors are outpacing increases in provision of low-fat dietary advice.

Figure

Likelihood of receiving low-fat dietary advice, by number of risk factors for type 2 diabetes and by diabetes status, Medical Expenditure Panel Survey, 2002–2009 (n = 188,006). Adjustment covariates were population characteristics (age, sex, race, ethnicity, education, federal poverty index, and geographic region), smoking status, dummy variables for each potential number of diabetes risk factors, and the modified clinical comorbidity index (25). The reference group includes respondents who had no diabetes and no risk factors for diabetes type 2. Error bars represent 95% confidence intervals. [Table: see text]