Key points are not available for this paper at this time.
Abstract Background and Aims Factors affecting poor health-related quality of life (HRQoL) for people with chronic kidney disease (CKD) have been previously described in literature, and worse HRQoL is associated with CKD progression and death. Little emphasis has been placed on which factors may be potentially modifiable, and longitudinal HRQoL outcome analyses in non-dialysis dependent CKD cohorts are few in number. Baseline analysis of the NURTuRE-CKD cohort identified several associations between these factors and HRQoL and these associations can now be examined with longitudinal data. Method The NURTuRE-CKD cohort study recruited 2996 participants with non-dialysis dependent CKD between 2017-2019, collecting biochemical, anthropometric, sociodemographic, medical history and patient reported outcome measure (PROM) data. Face-to-face follow-up repeated these measures. Longitudinal worsening overall HRQoL was the main outcome of interest, represented by mapped EQ-5D-3L index value and visual analogue score (VAS). Multivariable linear mixed effects models were fitted in R using the lmer and lmerTest packages. Models for index value were fit separately for each independent variable of interest and collated, as well as fitting models with all variables of interest. Models were adjusted for available confounders, repeated measures and follow-up time. P-values were adjusted for multiple testing using false discovery rate correction. Results 2054/2996 (68.6%) had complete HRQoL data at follow-up, of whom most had CKD stage G3a-G4 (n = 1597, 77.8%). 1817 (88.5%) were white and 1196 (58.2%) were male. Mean age was 63.7 (SD ±14.5) years and the mean number of comorbidities was 3.7 (SD ±2.2). Median (IQR) time between baseline and follow-up was 518 (410) days. At follow-up, mean mapped EQ-5D-3L index value was 0.72 (SD ±0.3) and mean VAS score was 70.4/100 (SD ±20.6). 1595 (77.7%) reported problems in at least one EQ-5D-5L dimension, with 255 (12.4%) developing issues since baseline. 1203 (58.6%) reported worse HRQoL index values at follow up and 910 (44.3%) reported worse VAS. The dimension with the highest number of newly reported issues at follow-up was ‘usual activities’ with 301 (14.7%). Multivariable linear mixed methods models showed independent associations with worse HRQoL (Figs 1 and 2) for those on ten or more medications (β −0.118, CI −0.137 to −0.100, p 0.001), current smokers (β −0.036, CI −0.067 to −0.006, p = 0.0.012), obesity (β-0.059, CI −0.080 to-0.039, p 0.001), haemoglobin of 100g/L (β-0.044, CI −0.073 to-0.016, p = 0.014), hospital anxiety and depression scale scores of 8 or above for anxiety (β-0.146, CI −0.161 to-0.130, p 0.001) and depression (β-0.202, CI-0.219 to −0.186, p 0.001), worse health literacy (β −0.099, CI −0.136 to −0.063, p 0.001), and pain (β −0.148, CI −0.161 to −0.135, p ≤ 0.001), weakness (β −0.106, CI −0.120 to −0.092, p 0.001) and shortness of breath symptoms (β −0.074, CI −0.087 to −0.060, p 0.001). Conclusion Potentially modifiable factors influencing longitudinal worsening of HRQoL for people with CKD include polypharmacy, smoking, obesity, anaemia, depression and anxiety, limited health literacy, pain, weakness and shortness of breath symptoms. This highlights the need for high quality randomised controlled trials to test interventions addressing these factors with reported HRQoL outcomes. Further work in this cohort is required, examining a third timepoint for HRQoL and linked clinical outcome data.
Phillips et al. (Wed,) studied this question.