Dietary Patterns and Mortality in a Multinational Cohort of Adults Receiving Hemodialysis

Abstract

Rationale & Objective: Clinical practice guidelines for dietary intake in hemodialysis focus on individual nutrients. Little is known about associations of dietary patterns with survival. We evaluated the associations of dietary patterns with cardiovascular and all-cause mortality among adults treated by hemodialysis. Study Design: Prospective cohort study. Setting & Participants: 8,110 of 9,757 consecutive adults on hemodialysis (January 2014 to June 2017) treated in a multinational private dialysis network and with analyzable dietary data. Exposures: Data-driven dietary patterns based on the GA2LEN food frequency questionnaire. Participants received a score for each identified pattern, with higher scores indicating closer resemblance of their diet to the identified pattern. Quartiles of standardized pattern scores were used as primary exposures. Outcomes: Cardiovascular and all-cause mortality. Analytical Approach: Principal components analysis with varimax rotation to identify common dietary patterns. Adjusted proportional hazards regression analyses with country as a random effect to estimate the associations between dietary pattern scores and mortality. Associations were expressed as adjusted HRs with 95% CIs, using the lowest quartile score as reference. Results: During a median follow-up of 2.7 years (18,666 person-years), there were 2,087 deaths (958 cardiovascular). 2 dietary patterns, “fruit and vegetable” and “Western,” were identified. For the fruit and vegetable dietary pattern score, adjusted HRs, in ascending quartiles, were 0.94 (95% CI, 0.76-1.15), 0.83 (95% CI, 0.66-1.06), and 0.91 (95% CI, 0.69-1.21) for cardiovascular mortality and 0.95 (95% CI, 0.83-1.09), 0.84 (95% CI, 0.71-0.99), and 0.87 (95% CI, 0.72-1.05) for all-cause mortality. For the Western dietary pattern score, the corresponding estimates were 1.10 (95% CI, 0.90-1.35), 1.11 (95% CI, 0.87-1.41), and 1.09 (95% CI, 0.80-1.49) for cardiovascular mortality and 1.01 (95% CI, 0.88-1.16), 1.00 (95% CI, 0.85-1.18), and 1.14 (95% CI, 0.93-1.41) for all-cause mortality. Limitations: Self-reported food frequency questionnaire, data-driven approach. Conclusions: These findings did not confirm an association between mortality among patients receiving long-term hemodialysis and the extent to which dietary patterns were either high in fruit and vegetables or consistent with a Western diet.