乌鸦传媒

Different kinds of dietary fats were associated with varying risks of mortality, according to a new study.

Researchers analyzed data from more than 126,000 participants with a 32-year follow-up and found that total dietary fat intake was inversely associated with total mortality when compared with total carbohydrate intake (hazard ratio for top versus bottom quintile 0.84, 95% CI 0.81-0.88; P<0.001).

But not all of the fats had equal associations, according to , at the Harvard T.H. Chan School of Public Health, and colleagues writing in .

Hazard ratios of total mortality when top and bottom quintiles were compared varied by type of fat (all P<0.001):

• 1.08 (95% CI 1.03-1.14) for saturated fat
• 0.81 (95% CI 0.78-0.84) for polyunsaturated fatty acid
• 0.89 (95% CI, 0.84-0.94) for monounsaturated fatty acid
• 1.13 (95% CI, 1.07-1.18) for trans fat

Relying on data from the long-running Nurses' Health Study and the Health Professionals Follow-up Study, the authors estimated that when 5% of energy from saturated fats was replaced with equivalent energy from polyunsaturated or monounsaturated fats, total mortality would be reduced 27% (HR 0.73, 95% CI 0.70-0.77) and 13% (HR 0.87, 95% CI 0.82-0.93), respectively.

"Replacement of saturated fats with unsaturated fats can confer substantial health benefits and should continue to be a key message in dietary recommendations," Wang and colleagues concluded. "These findings also support the elimination of partially hydrogenated vegetable oils, the primary source of trans-fatty acids."

Study participants who reported eating the most omega-6 polyunsaturated fatty acids were less likely to die than those who reported eating the least (HR 0.85, 95% CI 0.81-0.89; P<0.001), and omega-6 was associated with a lower incidence of mortality for most individual causes. Omega-3 polyunsaturated fatty acid intake was associated with a slightly lower mortality (HR 0.96, 95% CI 0.93-1.00; P=0.002).

Trans fat has previously been shown to be linked to worsened cardiovascular outcomes, but little research has been done on its relationship to mortality, wrote Wang and colleagues. In this study, trans fat intake was associated with a higher risk of cardiovascular mortality (HR 1.20, 95% CI 1.08-1.33; P<0.001) when extreme quintiles were compared.

Patients with a history of diabetes, cardiovascular disease, or cancer were excluded from the study. For the Nurses' Health Study, the baseline year was set at 1980; for the Health Professionals study, it was 1986. Dietary information was collected via semi-quantitative food frequency questionnaires, which asked the participant how often, on average, they consumed a specific portion size of a food during the last year.

At baseline, participants who reported higher saturated fat and mono-unsaturated fat intake had higher total energy and dietary cholesterol intake. They were also less likely to use multivitamins, exercise, or report a history of hypertension or hypercholesterolemia.

Total fat intake was initially positively and significantly correlated with mortality, but after the authors adjusted for confounding variables they found a significant negative correlation. Intake of saturated fat wasn't associated with cardiovascular mortality, and there was an inverse relationship between mono-unsaturated fat intake and cardiovascular mortality, particularly among women.

Among specific poly-unsaturated fats that the authors examined, linoleic acid intake was most robustly related with improved cardiovascular mortality. When saturated fat replaced carbohydrate intake, patients had a slightly higher cancer mortality risk (HR 1.07, 95% CI 0.98-1.17; P=0.02), but other types of fat were not significantly associated with cancer mortality.

Limitations of the study include reliance on food frequency questionnaires, which can be inaccurate. The design of the study does not allow for inference about causation, and reverse causation might have explained the findings, the authors noted -- i.e., people with poor health might have changed their habitual diets. In addition, the authors could not rule out confounding by unmeasured factors.

Comment