Plasma marine n-3 PUFA levels showed a large variation in the present study, ranging from 1.35 to 23.87 wt%, with median level of 7.95 wt% and an interquartile range of 6.20 to 10.03 wt%. Patients with high plasma marine n-3 PUFA levels were generally older, less often current smokers and had lower plasma n-6 PUFA levels. Despite the higher age group, they had a lower prevalence of diabetes mellitus. Tacrolimus was the calcineurin inhibitor of choice in younger patients in the recent transplant era (2007 – 2011) and consequently patients with high plasma marine n-3 PUFA levels were more often treated with cyclosporine A than with tacrolimus. When adjusted for recipient age, choice of calcineurin inhibitor did not differ between patients with high compared with low plasma marine n-3 PUFA levels.
8.1 Overall and cause-specific mortality
During a median follow-up of 6.8 years, there were 406 deaths (20.4%). Death was caused by cardiovascular disease in 164 patients (40.4% of deaths). There were 95 deaths due to cancer (23.4%) and 101 deaths due to infectious disease (24.9%). Since plasma marine n-3 PUFA levels were associated with recipient age, we estimated mortality rates according to plasma marine n-3 PUFA levels across age groups. High plasma marine n-3 PUFA levels (≥ 7.95 wt%) compared with low levels (< 7.95 wt%) were associated with lower mortality rates in all age groups except for patients aged 60 years or more (Table 5).
Table 5. Mortality rates and mortality rate ratios according to marine n-3 polyunsaturated fatty acid levels and recipient age.
Age category
16-44 years 45-59 years 60-80 years
Marine n-3 PUFA level High Low High Low High Low
Mortality rate, cases / 1000 person-years 4.1 10.1 16.6 26.7 59.2 69.0
Mortality rate difference - 6.0 - 10.1 - 9.8
Mortality rate ratio (confidence interval) 0.41 (0.17-0.88) 0.62 (0.40-0.85) 0.86 (0.63-1.05)
Pooled estimate (confidence interval) 0.69 (0.57-0.85)
Multivariable Cox proportional hazard regression analysis was used to estimate all-cause and cause-specific mortality hazard ratios across quartiles of plasma marine n-3 PUFA levels, without (model 1) and with (model 2) plasma n-6 PUFA levels included in the Cox models.
Using model 1, mortality risk was 56% lower for patients belonging to the upper marine n-3 PUFA quartile compared with the lower (adjusted HR 0.44; 95% confidence interval [CI]
0.26 to 0.75), whereas it was 67% lower using model 2 (adjusted HR 0.33; 95% CI 0.19 to 0.58). The negative association between plasma marine n-3 PUFA levels and mortality was largely due to a negative association with cardiovascular mortality and especially the risk of death from stroke and sudden cardiac deaths. Using model 1, comparing the upper marine n-3 PUFA quartile with the lower, there was a 82% lower cardiovascular mortality risk (adjusted HR 0.18; 95% CI 0.08 to 0.42). The association between plasma marine n-3 PUFA levels and death due to infectious disease did not reach statistical significance. We found no association between plasma marine n-3 PUFA levels and death due to cancer.
8.2 Cardiovascular risk markers
Multivariate linear regression analysis was used to assess associations between plasma marine n-3 PUFAs and cardiovascular risk markers measured at ten weeks post-transplant, adjusted for the influence of other traditional and transplant specific cardiovascular risk factors. We found negative associations between plasma marine n-3 PUFA levels and resting heart rate (Unstd. βcoeff. 0.56, Std. βcoeff. 0.13, p<0.001), fasting plasma glucose (Unstd. βcoeff. -1.04, Std. β-coeff. -0.08, p=0.001) and plasma triglyceride levels (Unstd. β-coeff. -6.55, Std.
β-coeff. -0.18, p<0.001) and a positive association with plasma HDL cholesterol levels (Unstd. β-coeff. 0.30, Std. β-coeff. 0.05, p=0.05). There were no significant associations between plasma marine n-3 PUFA levels and plasma LDL cholesterol levels, SBP, DBP or pulse wave velocity.
8.3 Overall and death censored graft loss
During follow-up, 569 (28.6%) renal grafts were lost in total, either due to recipient death (n=340, 59.8% of graft losses) or death censored graft loss (n=229, 40.2%). We estimated the risk of graft loss with plasma marine n-3 PUFA levels as a continuous variable. Using multivariable Cox regression we found a significant negative association between plasma marine n-3 PUFA levels and both overall and death censored graft loss. For every 1.0 wt%
increase in plasma marine n-3 PUFA level, there was a 11% reduced risk of overall graft loss (adjusted HR 0.89; 95% CI 0.84 to 0.93) and a 10% reduced risk of death censored graft loss (adjusted HR 0.90; 95% CI 0.84 to 0.97).
8.4 Acute rejection episodes
We found a negative association between plasma marine n-3 PUFA levels and acute rejection rate beyond three months post-transplant. Patients belonging to the lower marine n-3 PUFA quartile suffered more acute rejection episodes than patients belonging to the three upper quartiles. In contrast, we found no association with acute rejection rate within the first three months after transplantation (Table 6).
Table 6. Acute rejection rates according to marine n-3 polyunsaturated fatty acid quartiles.
Marine n-3 PUFA quartiles All Q1 Q2 Q3 Q4 p
Marine n-3 PUFA level, wt% 1.35-23.87 6.20 6.21-7.94 7.95-10.02 10.03 (trend)
Number of patients 1990 497 499 495 499
Acute rejections within first three
months post-transplant,% 27.7 27.8 29.7 27.3 26.3 0.44
Acute rejections beyond the first
three months post-transplant, % 8.9 11.7 9.2 6.9 7.8 0.02
8.5 Renal graft function
We used increase in serum creatinine as a measure of decline in renal graft function during the first five years after transplantation and found significant differences across marine n-3 PUFA quartiles (Table 7).
Table 7. Renal graft function according to marine n-3 polyunsaturated fatty acid quartiles.
Marine n-3 PUFA quartiles All Q1 Q2 Q3 Q4 p
Marine n-3 PUFA level, wt% 1.35-23.87 6.20 6.21-7.94 7.95-10.02 10.03 (trend)
Number of patients 908 196 216 234 262
Mean change in serum creatinine between ten weeks and five years
after transplantation, µmol/L 9.0 18.4 7.3 5.0 6.8 0.01
8.6 Plasma eicosapentaenoic and docosahexaenoic acid levels
Plasma EPA and DHA levels are derived from the same dietary sources and were highly correlated (Spearmans r = 0.72). They were analyzed separately in multivariable Cox regression analysis.
Patients belonging to the upper EPA quartile compared with the lower had a significant lower infectious disease mortality risk (multivariable adjusted HR 0.24; 95% CI 0.09 to 0.62). In contrast, plasma DHA levels were not associated with the risk of death from infectious disease. Both plasma EPA and DHA levels were negatively associated with cardiovascular mortality.
We found negative associations with resting heart rate, plasma triglyceride levels and fasting plasma glucose levels for both plasma EPA and DHA levels. On the other hand, a positive association with plasma HDL cholesterol levels was only found with plasma EPA levels (Std.
β-coeff. 0.16, p<0.001).
Plasma levels of EPA, but not DHA, were significantly associated with death censored graft loss (EPA: adjusted HR 0.81; 95% CI 0.71 to 0.93. DHA: adjusted HR 0.91; 95% CI 0.80 to 1.02), whereas both were associated with overall graft loss (EPA: adjusted HR 0.84; 95% CI 0.77 to 0.91. DHA: adjusted HR 0.87; 95% CI 0.80 to 0.94). Associations with acute rejection rates within the first year and decline in renal graft function during the first five years after transplantation did not differ between plasma EPA and DHA levels.