In our nationwide observational cohort study, we found no association of iron status with HRQoL, nor with the presence of fatigue, shortness of breath, muscle cramps or restless leg syndrome during the first year of dialysis. Importantly, our results were similar after adjusting for key baseline and time-varying confounders, such as comorbidities and CRP levels, compared to when no additional adjustments for confounders were made.

At the start of dialysis, the majority of patients had a low iron status; the median ferritin and TSAT levels were 206 ng/mL and 19%, respectively. However, only 33% of patients were using any form of iron therapy at baseline. There is considerable variation in iron use during dialysis across countries, leading to differences in ferritin and TSAT levels. For example, in the USA, over 40% of prevalent HD patients had ferritin levels above 800 ng/mL, which is much higher than in our dialysis cohort [24, 25]. While more liberal dosing strategies are now considered superior following the PIVOTAL trial, a degree of caution is warranted in patients with excessively high iron status. Our cutoffs for ferritin and TSAT were based on current leading evidence from the PIVOTAL trial and the KDIGO guideline and reflect the latest prescribing practices. Although the lower cutoffs to define iron deficiency have been scrutinized, these definitions generally rely on thresholds of 200 ng/mL for ferritin and 20% for TSAT in HD patients [3, 26], which are the cutoffs we used.

Maintaining adequate iron levels is crucial to prevent anemia, increased EPO-stimulating agent dose requirements, and blood transfusions. However, it is also important to consider HRQoL when managing iron levels. Since iron plays a crucial role in muscle, brain, and cardiovascular health, it could be speculated that iron status would be an important factor determining HRQoL in the current patient setting [27, 28]. Surprisingly, in our study, higher ferritin and TSAT levels were not associated with better HRQoL, and levels exceeding safety thresholds had no impact on HRQoL either. Importantly, our results remained consistent after adjusting for Hb levels, suggesting that iron, either directly or through Hb, does not impact HRQoL in dialysis patients. An observational study in non-dialysis CKD patients from France, USA and Brazil (CKDopps) showed that ferritin levels ≥ 300 ng/mL, and TSAT levels ≤ 15% were associated with a worse physical HRQoL [10]. It is well conceivable that the effect of iron differs between non-dialysis and dialysis CKD patients, as the latter experience more severe health impairments and symptoms [29]. However, there are a few other important differences between our study and the study in non-dialysis CKD patients. For instance, the cutoffs used in the CKDopps study were based on the heart failure population and are therefore more conservative than ours, which could result in different findings. Moreover, another possible difference that could help explain our different findings is that ferritin and TSAT levels were treated as time-fixed in the non-dialysis CKD study, despite iron levels varying considerably over time, which could introduce bias [30]. Ultimately, very modest associations were found in HRQoL (< 2 points) in this study that may also be attributed to the larger sample size of 2513 [10].

While HRQoL is widely recognized as a key measure of patient well-being, it is comprised of various factors, including symptoms that may affect overall quality of life. To our knowledge, no prior research has investigated the association between iron status and specific symptoms in either dialysis or non-dialysis CKD patients. In our study, we also found no association between iron status and the commonly reported anemia-related symptoms like fatigue, shortness of breath, muscle cramps and restless legs. Although iron plays important physiological roles [4, 5], its effects do not appear to translate into improvements of symptoms in dialysis patients.

Our study has several strengths. We used longitudinal data on ferritin and TSAT, which allowed us to investigate longitudinal associations in a large cohort of HD and PD patients. We used robust statistical approaches to adjust for both baseline and time-varying confounding. Moreover, we based the cutoffs for ferritin and TSAT on the latest evidence, and thus chose those that are most useful for current clinical practice. However, our study also has a few limitations. Excluded patients had lower mental component summary scores, likely due to more missing data, but given that selection was not related to the exposures, a selection bias appears unlikely. Despite adjusting for key confounders, we cannot rule out residual confounding due to the observational nature of our study, as not all potential confounders, such as nutritional status, were measured. Due to the variability of missing data across time points and the nature of our analytic approach, we were unable to conduct a complete case analysis using non-imputed data. We were also unable to achieve convergence with item-level imputation for HRQoL and had to impute at the outcome level, which may introduce some bias. Similarly, despite the use of strong auxiliary variables for imputation, we cannot rule out that some data may be missing not at random, especially in the case of patient-reported outcomes. However, the number of missing responses per item varied by no more than 5% (Table S1), suggesting that the risk of bias may be limited. Although serum ferritin and TSAT are the most widely used markers for assessing iron status in clinical practice, we recognize that accurately determining iron stores remains challenging and that more invasive diagnostic tests (e.g., bone marrow biopsy) may be more reliable but not clinically feasible. Lastly, since we only focused on the first year of dialysis, we were not able to capture potential effects of iron on patient-reported outcomes beyond this period.

Future research could therefore explore longer follow-up periods, such as two years. Given that we observed a trend toward lower mental component summary scores with increasing ferritin levels, future studies could explore the impact of using even higher cutoff thresholds for ferritin. Additionally, since our study did not compare treatment strategies, future studies could examine the effects of iron treatment strategies on patient-reported outcomes, ideally through a randomized trial or an emulation thereof. Lastly, future studies should preferably include a greater number of older patients, those with limited life expectancy and those not waitlisted for transplantation, as these subgroups may experience greater symptom burden and reduced quality of life.

To conclude, we found no association of iron status with patient-reported outcomes in more than 1000 patients during the first year after dialysis initiation. Our findings suggest that there may be no benefits of having a higher iron status for patient-reported outcomes, and that decisions on iron therapy in dialysis patients should therefore be guided by clinical and biochemical outcomes like Hb levels and EPO-stimulating agent requirements.