Associations of Dietary Fats with All-Cause Mortality and Cardiovascular Disease Mortality among Patients with Cardiometabolic Disease

Abstract

Previous studies have shown distinct associations between specific dietary fats and mortality. However, evidence on specific dietary fats and mortality among patients with cardiometabolic disease (CMD) remains unclear. The aim of this study was to estimate the association between consumption of specific fatty acids and survival of patients with CMD and examine whether cardiometabolic biomarkers can mediate the above effects. The study included 8537 participants with CMD, from the Third National Health and Nutrition Examination Survey (NHANES III) and NHANES 1999-2014. Cox proportional hazards regression, restricted cubic spline regression, and isocaloric substitution models were used to estimate the associations of dietary fats with all-cause mortality and cardiovascular disease (CVD) mortality among participants with CMD. Mediation analysis was performed to assess the potential mediating roles of cardiometabolic biomarkers. During a median follow-up of 10.3 years (0-27.1 years), 3506 all-cause deaths and 882 CVD deaths occurred. The hazard ratios (HRs) of all-cause mortality among patients with CMD were 0.85 (95% confidence interval (CI), 95% CI, 0.73-0.99; p trend = 0.03) for ω-6 polyunsaturated fatty acids (ω-6 PUFA), 0.86 (95% CI, 0.75-1.00; p trend = 0.05) for linoleic acid (LA), and 0.86 (95% CI, 0.75-0.98; p trend = 0.03) for docosapentaenoic acid (DPA). Isocalorically replacing energy from SFA with PUFA and LA were associated with 8% and 4% lower all-cause mortality respectively. The HRs of CVD mortality among CMD patients comparing extreme tertiles of specific dietary fats were 0.60 (95% CI, 0.48-0.75; p trend = 0.002) for eicosapentaenoic acid (EPA), and 0.64 (95% CI, 0.48-0.85; p trend = 0.002) for DPA and above effects were mediated by levels of total cholesterol (TC), triglycerides (TG), low density lipoprotein cholesterol (LDL), and high density lipoprotein cholesterol (HDL). Restricted cubic splines showed significant negative nonlinear associations between above specific dietary fats and mortality. These results suggest that intakes of ω-6 PUFA, LA, and DPA or replacing SFA with PUFA or LA might be associated with lower all-cause mortality for patients with CMD. Consumption of EPA and DPA could potentially reduce cardiovascular death for patients with CMD, and their effects might be regulated by cardiometabolic biomarkers indirectly. More precise and representative studies are further needed to validate our findings.

Keywords: cardiometabolic disease; dietary fats; mediating effect; mortality.

Figure 1

The flow chart of selecting a full analysis set.

Figure 2

Multivariable hazard ratios (HRs) of all cause mortality and CVD mortality by isocaloric substitution of specific fatty acid for saturated fatty acids among participants with CMD in NHANESIII and NHANES 1999–2014.

Figure 3

Dose–response relationships of effective dietary acids with all-cause and CVD mortality among patients with CMD in NHANES III and NHANES 1999–2014 ^a. ^a: Dose–response relationships between effective dietary acids with all-cause and CVD mortality among patients with CMD, (A) ω-6 PUFA in all-cause mortality (p nonlinear = 0.0454), (B) LA in all-cause mortality (p nonlinear = 0.0447), (C) DPA in all-cause mortality (p nonlinear < 0.0001), (D) EPA in CVD mortality (p nonlinear < 0.0001), (E) DPA in CVD mortality (p nonlinear < 0.0001) in NHANES III and NHANES 1999–2014 were estimated by multivariable adjusted Cox regression models based on restricted cubic splines. Solid splines represented estimated hazard ratios, and cloud areas represented the corresponding 95% CI. The multivariable-adjusted model was adjusted for gender, age, race, educational levels, PIR, BMI, smoking status, alcohol drinking status, physical activity status, self-reported health status, baseline history of hypertension, hyperlipidemia and cancer, total energy intake, energy intake derived from carbohydrate, protein intake (in tertiles), cholesterol levels (in tertiles), and energy intake derived from remaining fatty acids (SFA, MUFA, and PUFA) where appropriate. Abbreviation and acronyms: CMD, cardiometabolic disease; CVD, cardiovascular disease; ω-6 PUFA, ω-6 polyunsaturated fatty acids; LA, linoleic acid; EPA, eicosapentaenoic acid; DPA, docosapentaenoic acid.