More Novel Effects of Diet on Blood Pressure and Lipids

Recent “feeding studies” in a highly selected population of participants who are prehypertensive and stage 1 hypertensive have demonstrated that a diet modestly reduced in salt content coupled with an increase in fresh fruits, vegetables, and low-fat dairy products (Dietary Approaches to Stop Hypertension [DASH] diet) could lower blood pressure.¹ Such a diet, with increased potassium, calcium, and magnesium content in addition to the reduced sodium intake, provided confirmation of earlier studies that had focused on a single mineral component and had indicated that reducing sodium intake or increasing potassium intake, and less consistently, increasing calcium intake, could lower blood pressure. A subsequent study testing the effect of 3 levels of sodium intake (usual, modestly reduced, and greatly reduced) with the DASH diet provided evidence of a graded influence of dietary salt restriction on blood pressure.²

Concerns have been raised regarding the effect of dietary sodium restriction on blood lipid levels, based largely on short-term studies using more severe reduction in sodium intake than that of the DASH studies, which demonstrated a reduction in low-density lipoprotein (LDL) cholesterol levels.^{3 - 4} In this issue of JAMA, Appel and colleagues⁵ report the results of the OmniHeart randomized trial, which represents the latest effort by members of the DASH Trials group to examine the effect of varying protein, monounsaturated fat, and carbohydrate intake with this basic DASH diet on blood pressure and lipid levels in a similar population.

The OmniHeart Trial used a complex crossover design and recruited 164 individuals with blood pressure between 120 and 159 mm Hg systolic and 80 and 99 mm Hg diastolic, including 45% women and 55% African Americans, 79% of whom were overweight or obese. The basic DASH diet was modified to contain 58% kcal of carbohydrate; 25% kcal of protein; or a high unsaturated fat intake (37% kcal) using olive oil, canola oil and safflower oil, and nuts and seeds. Each diet was ingested for a 6-week period. The caloric content of the diet was designed to avoid weight changes as a confounding variable.

The reduction in blood pressure, at least during the high-carbohydrate phase of the study, was similar to that reported with the DASH diets alone, implying that for that phase of the study the change in blood pressure may well have been attributable to the reduction of sodium intake (25% from baseline) in conjunction with the increased potassium (30% from baseline) and calcium intake (not reported) of the basic DASH diet. The high carbohydrate period was also associated with the smallest improvement in lipids. Thus, little benefit of a high carbohydrate intake-DASH diet beyond the basic diet can be inferred from the present study. The other 2 diet phases featured further blood pressure reduction and an improvement in the lipid profile. The high-protein diet was associated with reduced physical activity, reduced appetite, and bloating. Future studies will be needed to determine whether these issues alter compliance with the diet over a longer term than the 6-week period used for these studies and whether reduced physical activity observed with the high-protein phase is associated with weight gain, which may blunt the effects of the diet on blood pressure and lipids.

The investigators were careful to design isocaloric diets to avoid weight loss as a confounding variable. However, given the fact that the vast majority of participants in the OmniHeart Trial were overweight or obese and the fact that the national epidemic of obesity contributes to elevation of both blood pressure and lipids, is it reasonable to anticipate that the next rendition of the DASH and successor diets by the authors will focus on weight loss? This would seem to be the logical next and perhaps most important intervention.

The most surprising observation reported by the OmniHeart Trial is the reduction in high-density lipoprotein (HDL) cholesterol content after the high-protein phase. Given the potential effect of HDL cholesterol on cardiovascular disease, more information regarding this finding will be needed before such a diet can be routinely recommended. Long-term outcome trials focusing on cardiovascular events will also be required to convince both clinicians and the public that the reduction in HDL cholesterol levels associated with the high-protein diet is reflected in a decrease in actual events, which is contrary to conventionally held notions about HDL cholesterol.

Another issue inherent in such carefully conducted trials is the ability to generalize from the study population and the observations of the study to a broader population. The participants in both of the DASH trials and in the OmniHeart Trial were highly selected from a motivated group of relatively young, well-educated, overweight individuals, a majority of whom were ethnic urban dwellers. Previous studies have shown that individuals with similar characteristics are likely to respond to sodium reduction⁶ and the DASH-type diets with a reduction in blood pressure and lipid levels.⁷ Thus, the blood pressure responses of the participants in the present trial are not very surprising. Although the benefit of blood pressure reduction in those in the prehypertensive group is clear, the ability to control blood pressure adequately with diet alone among the stage 1 hypertensives studied in the OmniHeart Trial has not been presented. It is likely that more than diet will be required to reach goal blood pressures for the majority of these individuals, particularly with the lowering of blood pressure goals being advocated at present.

The provision of all meals and snacks for a period of several months may have served as an inducement for study participants to remain in the trial. This may not be applicable to unselected patients or to the general population who have to purchase and prepare their own meals. Moreover, the practical application of these findings requires the demonstration that the lifestyle changes inherent in the tested diets can be maintained for periods longer than the few weeks studied in the OmniHeart trial. The composition of and the apparent lack of commercial availability of these very carefully designed diets also may be limitations to the broader application of these findings. Finally, because the OmniHeart Trial only used the surrogate outcomes of blood pressure and lipid levels, longer trials examining actual cardiovascular event outcomes will be needed to convince a skeptical public of the benefit of yet another unique and difficult-to-achieve dietary regimen.