Comparison of the effects of non-calcium and calcium-based phosphate binders on fibroblast growth factor-23 levels in patients with CKD–mineral and bone disorder: a meta-analysis of randomized controlled trials

Abstract

Background

Phosphate binders are used to manage hyperphosphatemia in patients with chronic kidney disease-mineral and bone disorder (CKD-MBD) and may influence fibroblast growth factor 23 (FGF-23), a biomarker associated with disease progression and mortality. However, the comparative effects of calcium-based phosphate binders (CPBs) and non-calcium-based phosphate binders (NCPBs) on FGF-23 levels remain unclear. We conducted a meta-analysis of randomized controlled trials to compare the effects of NCPBs and CPBs on FGF-23 levels in patients with CKD-MBD with hyperphosphatemia.

Methods

A systematic search of PubMed, EMBASE, and the Cochrane Library was performed for studies published up to August 2024. Only randomized controlled trials were eligible; non-randomized controlled trials, studies with overlapping patient populations, pediatric or animal studies, and trials without quantitative FGF-23 outcomes were excluded. Language restrictions were not imposed. Data were synthesized as standardized mean differences using a random-effects model, and the risk of bias was assessed using the Cochrane Risk of Bias 2 tool.

Results

Eleven randomized controlled trials involving 791 patients were included. Overall, NCPBs were associated with a greater reduction in FGF-23 levels compared with CPBs (standardized mean difference −0.56, 95% confidence interval −0.95 to −0.17, p = 0.005). Subgroup analyses suggested consistent trends across binder types (sevelamer and lanthanum), dialysis status, and treatment duration. However, the subgroup differences were not statistically significant, and heterogeneity remained substantial in several subgroups, indicating that these findings should be interpreted with caution.

Conclusions

In CKD-MBD patients with hyperphosphatemia, NCPBs were associated with greater reductions in FGF-23 levels compared with CPBs. The subgroup results suggested possible trends by binder type, CKD category, and treatment duration. However, given the heterogeneity and lack of statistically significant subgroup differences, these findings should be considered hypothesis-generating. Further well-designed trials are needed to confirm these results and clarify the underlying mechanisms and clinical implications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12882-026-04964-4.

Background

Chronic kidney disease (CKD) is a progressive condition that compromises the body’s ability to maintain homeostasis and results in a wide range of complications. Among these, mineral and bone disorder (CKD-MBD) is a common and serious condition characterized by abnormalities in phosphate, calcium, parathyroid hormone (PTH), and vitamin D metabolism. As renal function declines, phosphate excretion becomes impaired, leading to hyperphosphatemia. This biochemical disturbance contributes not only to vascular calcification and skeletal fragility but also to increased morbidity and mortality [1, 2]. To mitigate these risks, phosphate binders are widely used to control serum phosphate levels and restore balance in mineral metabolism. Calcium-based phosphate binders (CPBs) have traditionally been used as the first-line therapy; however, concerns about hypercalcemia and cardiovascular calcification have led to the increased use of non-calcium-based phosphate binders (NCPBs), such as sevelamer and lanthanum carbonate [3].

Fibroblast growth factor-23 (FGF-23), which is secreted by osteocytes and regulates the bone–parathyroid–kidney axis, has emerged as a central biomarker in CKD-MBD patients with disordered phosphate homeostasis. PTH, 1,25-dihydroxyvitamin D₃, and elevated dietary and serum phosphorus levels are known to stimulate FGF-23 production [4]. Calcium may also indirectly modulate FGF-23 levels by altering parathyroid hormone secretion and vitamin D metabolism. Furthermore, accumulating evidence indicates that iron deficiency can upregulate FGF-23 transcription via stabilization of hypoxia-inducible factor-1α (HIF-1α) signaling in osteocytes, resulting predominantly in increased C-terminal FGF-23 levels through enhanced cleavage [5]. Despite its complex regulation and pleiotropic effects, FGF-23 primarily acts on the kidney and parathyroid gland by promoting renal phosphate excretion and reducing 1,25-dihydroxyvitamin D₃ production, thereby indirectly influencing calcium homeostasis [6]. Importantly, the calcium load from CPBs may offset the beneficial effect of phosphate reduction on FGF-23 levels, raising concerns about their long-term effects compared to those of NCPBs.

Previous studies, including a cohort study of patients with CKD-MBD in South Korea, confirmed that FGF-23 levels increase as renal function declines [7]. Several clinical trials and meta-analyses have reported variable effects of phosphate binders on FGF-23. Some demonstrated significant reductions in FGF-23, particularly with NCPBs [8–10], whereas others observed no effect [11]. The variability in findings may stem from factors such as a small sample size, limited follow-up duration, or differences in the comparator groups. Regardless, the potential influence of calcium loading on FGF-23 regulation remains unclear [8, 12, 13].

Taken together, the above findings highlight a critical gap in our understanding of the differential effects of CPBs and NCPBs on FGF-23 expression in CKD-MBD. Addressing this gap is essential to developing therapeutic strategies that optimize both biochemical control and long-term patient outcomes. Therefore, this meta-analysis aimed to systematically evaluate and compare the impact of NCPBs and CPBs on FGF-23 levels in patients with CKD-MBD to clarify mineral metabolism management in this population.

Methods

This study adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines for conducting systematic literature reviews and meta-analyses, with each stage performed independently by three researchers to ensure the reliability and validity of the findings. Any disagreements during the process were resolved through consensus; if consensus was not reached, a majority decision was made by arbitration with a fourth reviewer when necessary.

Search strategy

We conducted a comprehensive search of PubMed, EMBASE, and the Cochrane Library, covering publications up to August 2024. Search terms combined controlled vocabulary (e.g., MeSH) and free-text keywords, including “chronic kidney disease-mineral and bone disorder” [14] OR “CKD-MBD”; “sevelamer” OR “lanthanum” (NCPBs); “calcium carbonate” OR “calcium acetate” (CPBs); and “fibroblast growth factor-23” OR “FGF-23.” Given the complex and multifactorial regulation of FGF-23, NCPBs were limited to sevelamer and lanthanum to minimize mechanistic heterogeneity. Iron-based phosphate binders were excluded due to their direct effects on iron metabolism, which may independently influence FGF-23 levels, and tenapanor was also excluded because of its distinct mechanism of action as an intestinal sodium–hydrogen exchanger inhibitor. Boolean operators (AND/OR) were used to combine the terms, and search strings were adapted to each database. Language restrictions were not imposed. The filters for publication type limited the retrieval to randomized controlled trials (RCTs). Manual searches of reference lists and citation tracking were performed to identify additional eligible studies.

Study selection

A two-step process was used for screening: (1) title and abstract screening, followed by (2) full-text review. Three reviewers independently screened all records. Conflicts were resolved by consensus or the involvement of a fourth reviewer. The screening process was managed using Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia). The study selection process was documented using a PRISMA 2020 flow diagram (Fig. 1).

Fig. 1.

Flow diagram of study selection

Inclusion and exclusion criteria

The eligible studies were limited to RCTs reporting FGF-23 as an outcome. Studies were excluded if they did not report FGF-23 values at baseline, follow-up, or post-treatment, involved interventions where phosphate binders were combined with other active treatments, had inappropriate comparator groups (non-binder agents), or lacked sufficient follow-up (minimum of 4 weeks required). The comparison groups included calcium carbonate, or other CPBs.

Data extraction

The following data were extracted from each eligible study: author(s), year, country, study design, number and characteristics of patients, average age, interventions, comparators, and reported outcomes. FGF-23 values were collected at baseline and at the end of the longest available follow-up period.

Risk of bias assessment

The risk of bias for each included study was independently assessed by three reviewers using the Cochrane Risk of Bias 2 (RoB 2) tool, which evaluates five domains: (1) randomization process, (2) deviations from intended interventions, (3) missing outcome data, (4) measurement of the outcome, and (5) selection of reported results.

Statistical analysis

Meta-analyses were performed using Review Manager (RevMan, version 5.4.1; The Nordic Cochrane Centre and The Cochrane Collaboration, 2020). Standardized mean differences (SMDs) were chosen as the primary effect measure because the included studies utilized different assays and measurement units for FGF-23. We calculated SMDs with 95% confidence intervals (CIs) using the inverse-variance method. A random-effects model was applied a priori to account for the anticipated clinical and methodological heterogeneity across the included trials (e.g., variation in dialysis modalities and patient characteristics).

Heterogeneity was assessed using Cochran’s Q test and the I² statistic, where I² values of 25%, 50%, and 75% were considered to represent low, moderate, and high heterogeneity, respectively. Statistical significance was defined as a two-tailed p-value < 0.05. We considered the increased risk of type I errors to address multiple subgroup analyses (by drug type, dialysis status, and treatment duration). Although no formal adjustment (e.g., Bonferroni correction) was applied, the results from the subgroup analyses were interpreted with caution, and this limitation is acknowledged.

Data synthesis

When the mean and standard deviation (SD) values of the change in FGF-23 levels were unavailable, medians and quartiles (Q1 and Q3) were converted using established methods [15–17]. When studies reported multiple follow-up time points, we used the longest available duration. The FGF-23 assay units were harmonized across studies for comparability. Missing or non-derivable data were handled by contacting the study authors or applying validated statistical approximations.

Sensitivity and publication bias analyses

Two sensitivity analyses were performed: (1) excluding studies with a high risk of bias and (2) leave-one-out analysis, in which one study was sequentially removed to assess its influence on pooled estimates. Publication bias was evaluated using the meta-package in R (version 4.5.1), with funnel plots and Egger’s regression test.

Results

Study selection

A total of 138 articles were retrieved from PubMed, EMBASE, and the Cochrane Library. After removal of duplicates and screening of titles and abstracts, full-text articles were assessed for eligibility. Studies were excluded for reasons including overlapping populations, non-randomized or animal study designs, pediatric populations, irrelevant outcomes, and inappropriate comparators. Ultimately, eight randomized controlled trials were included in this analysis (Fig. 1). Additionally, three studies were identified through a manual search of the reference lists.

Characteristics of the included studies

Among the 11 RCTs included in this study, 2 were post-hoc analyses, and 2 were RCT crossover studies [10, 12, 18–26]. A total of 791 patients from 11 different countries participated, with a minimum study duration of 6 weeks. Six studies used sevelamers and five used lanthanum as experimental agents, whereas calcium acetate and calcium carbonate served as control agents (Table 1). Notably, the FGF-23 levels in all studies were measured using intact FGF-23 immunoassays, ensuring consistency in outcome measurements.

Table 1.

Characteristics of the selected studies

Author (Year)	Country	Study Design	Study Duration	Sample Size	Study Population	Intervention	Main Results	Vitamin D management	Phosphate control
Mason et al. [20]	USA (Single - center study)	Randomized, prospective, open-label, parallel group study	Fifteen weeks in total, which included a screening phase, a washout phase, and a 12-week treatment phase	Fifty patients were recruited, with 30 completing the study	Patients with CKD stages 3 and 4 who were not expected to start dialysis for 8 months. Participants were 18 years or older with an eGFR of 15–60 mL/min/1.73 m2 and had serum intact PTH < 500 pg/mL during the screening period.	Patients were randomized to receive either sevelamer carbonate at a dose of 1600 mg three times daily with meals, or calcium acetate (CA) at a dose of 1334 mg three times daily, for a duration of 12 weeks.	Sevelamer treatment resulted in a significant decrease in the levels of FGF-23 (p < 0.05), calcidiol, and calcitriol, whereas FGF-23 and calcitriol levels remained unchanged in the CA group.	Fixed-dose supplementation (cholecalciferol 400units daily)	Not clearly specified
Kovesdy et al. [21]	USA (Multi - center study) (2)	Randomized, controlled, parallel design, open-label, active comparator clinical trial	One year	One-hundred and twenty patients were recruited, with 108 completing the study	Patients with CKD stages 3 or 4 (eGFR 15 to 59 mL/min per 1.73 m2). Participants had to have evidence of abnormal phosphorus metabolism, defined as serum phosphorus > 4.6 mg/dl, plasma intact PTH > 65 pg/mL, or tubular reabsorption of phosphorus < 80%.	Patients were randomized in a 1:1:1 ratio to one of three groups for one year: oral lanthanum carbonate, oral CA, or a dietary intervention focused on phosphorus restriction.	Serum levels of FGF23 were significantly higher at month 12 compared with baseline. Changes in FGF-23 levels from baseline to month 12 were not statistically significant in either the lanthanum carbonate group (p = 0.15) or the CA group (p = 0.4)	Continued prior fixed-dose supplementation therapy	Dietary restriction included
Zhang et al. [19]	China (Single-center study)	Randomized, prospective, open-label, parallel-group pilot study	Twelve months	Ninety-two patients	Maintenance hemodialysis (MHD) patients aged 20–65 years with diabetes and adynamic bone disease (ABD). ABD was diagnosed based on clinical criteria of an intact parathyroid hormone (iPTH) level < 150 pg/mL and bone alkaline phosphatase (b-ALP) <20 ng/mL.	Patients were randomized on a 1:1 basis to receive either oral lanthanum carbonate (LC) or calcium carbonate.	LC treatment significantly reduced FGF-23 levels from baseline (p < 0.001), whereas the calcium carbonate group showed no significant change (p = 0.242), leading to lower FGF-23 levels in the LC group at 12 months (p < 0.001)	Not reported	Not specified
Chang et al. [7]	Taiwan (Multi-center study) (5)	Prospective, randomized, open-label, controlled trial	Twenty-eight weeks, which included a 4-week screening and washout period followed by a 24-week active treatment period	Forty-six patients were recruited, with 25 completing the study	Chronic hemodialysis patients older than 18 years, with serum iPTH levels between 150 and 600 pg/mL. Patients with diabetes mellitus were excluded.	Patients were randomized in a 1:1 ratio to receive either LC or calcium carbonate as monotherapy for 24 weeks.	LC treatment significantly decreased FGF-23 levels by the 24-week mark (p = 0.013), whereas the calcium carbonate group showed no significant change (p = 0.732).	Not reported	Not specified
Soriano et al. [18]	Spain (Single-center study)	Randomized, controlled study	Four months of active treatment, following a 1-month observation period	Thirty-two patients	Clinically stable, non-dialysis patients with CKD stages 4–5. Key inclusion criteria were serum phosphorus > 4 mg/dL and 25(OH)D levels > 30 ng/mL. Patients on vitamin D receptor activators were excluded.	Patients were randomized on a 1:1 basis to receive either oral LC (up to 3,000 mg/day) or calcium carbonate (up to 2,500 mg/day) for a four-month period.	LC significantly decreased FGF-23 levels from baseline (p < 0.001), whereas the calcium carbonate group showed no significant change, leading to significantly lower FGF-23 levels in the LC group at 4 months (p = 0.002).	Not reported	Not specified
Covic et al. [22]	Europe (Multi-center study in Romania, Germany, Poland, Portugal, and Spain)	A post hoc analysis of a controlled, randomized study (the CALMAG study)	Twenty-five weeks including a 24-week intervention phase	Two-hundred and four patients completed the study per the protocol	Patients aged 18–85 years with CKD stage 5 undergoing hemodialysis (HD) or online hemodiafiltration (HDF).	Patients were randomized to receive either CA/magnesium carbonate (CaMg)(n = 105) or sevelamer-hydrochloride (sevelamer-HCl) (n = 99) for 24 weeks.	Both CaMg and sevelamer-HCl significantly reduced intact fibroblast growth factor-23 (iFGF-23) levels from baseline (p < 0.0001 for both groups), with no significant difference found between the two groups at 25 weeks (p = 0.0705).	Continued prior fixed-dose supplementation therapy	Controlled in protocol
Yilmaz et al. [23]	Turkey (Single-center study)	Randomized, prospective, open-label trial	Eight weeks	One-hundred patients	Patients with stage 4 CKD and hyperphosphatemia (serum phosphate ≥ 6.0 mg/dL). Key exclusion criteria included diabetes mellitus, a history of coronary heart disease, smokers, and patients using statins, renin-angiotensin blockers, or vitamin D.	Patients were randomized to receive either sevelamer (n = 47) or CA (n = 53) for 8 weeks.	Sevelamer treatment significantly decreased FGF-23 levels by 27.1%, whereas the CA group showed no significant change (mean increase of 3.5%), leading to a statistically significant difference in FGF-23 changes between the groups after 8 weeks (p = 0.002).	Not reported	Not specified
Toida et al. [16]	Japan (Single-center study)	Randomized, open-label, crossover study	Two 3-month treatment periods, each preceded by a 2-week washout period	Fifty patients were recruited, with 42 completing the study	Patients on maintenance HD for longer than 3 months.	Patients were randomized to receive either lanthanum carbonate (LC) or calcium carbonate (CaC) for 3 months.	The LC group showed a statistically significant decrease in log FGF-23 levels (p < 0.01). In contrast, the CaC group showed no significant change in their FGF-23 levels (p = 0.125).	Continued prior therapy of dose titrating	Not specified
Vlassara et al. [17]	USA (Single-center study)	Randomized, open-label, crossover study	Two 2-month (8-week) treatment periods, separated by a 1-week washout period	Twenty patients	Patients with stage 2–4 diabetic CKD and proteinuria. Exclusion criteria included current treatment for hyperphosphatemia or hypercalcemia.	Patients were randomized to receive either sevelamer carbonate (1600 mg three times daily) or CaC (1200 mg three times daily) for 2 months.	SC treatment decreased FGF-23 levels while CaC treatment increased them, leading to a statistically significant difference between the treatments (p = 0.04). In the subgroup of patients with high baseline FGF-23 levels, SC significantly reduced FGF-23 levels (p = 0.02), while CaC did not.	Fixed-dose supplementation (cholecalciferol 400units daily) or continued prior therapy	Not specified
Cancela et al. [9]	Brazil (Multi-center study)	A post-hoc analysis of a randomized clinical trial (the BRIC study)	One year	Seventy-two patients completed the study	Maintenance hemodialysis patients (CKD stage 5) with hyperphosphatemia (serum phosphate > 5.5 mg/dL) after a 2-week washout period from phosphate binders and calcitriol.	Patients were randomized to receive either sevelamer or CA for 1 year. Other concurrent CKD-MBD therapies were also adjusted	Sevelamer treatment significantly reduced FGF-23 levels (p < 0.001), whereas the CA group did not show a significant decrease (p = 0.062). Sevelamer, non-use of calcitriol, and low-calcium dialysate were independently associated with greater FGF-23 reduction.	Not reported	Dialysate Ca + binder controlled
Oliveira et al. [15]	Brazil (Multi-center study)	Randomized, open-label trial	Eight weeks, which included a 6-week treatment period and a 2-week washout period	Forty-two patients were recruited, with 40 completing the study	Normophosphatemic patients with stage 3 or 4 CKD. Key exclusion criteria included diabetes mellitus and intact PTH levels of 500 pg/mL or higher.	Patients were randomized to receive either sevelamer hydrochloride or CA for a 6-week period with escalating doses.	Sevelamer hydrochloride significantly decreased FGF-23 levels from baseline (p < 0.05), whereas the CA group showed no significant change, leading to a statistically significant difference in the reduction of FGF-23 between the two groups after 6 weeks (p < 0.05).	Not reported	Early phosphate control strategy

Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; MBD, mineral and bone disorder

The effects of reducing FGF-23 levels were categorized and compared based on the type of NCPB, differentiating between sevelamer and lanthanum. We extracted the sample size, mean change in FGF-23 levels from baseline, and corresponding SD from each study for the meta-analysis, and SMDs were reported using a random-effects model.

Effect of the intervention on FGF-23

The pooled analysis of all 11 studies demonstrated that NCPBs resulted in a statistically significant reduction in FGF-23 levels compared to CPBs (SMD −0.56, 95% CI −0.95 to −0.17, p = 0.005). However, substantial heterogeneity was observed (I² = 84%, p < 0.00001) (Fig. 2). Given this level of heterogeneity, the interpretability of the pooled effect estimate is limited. This heterogeneity may be attributable to differences in patient characteristics, CKD stage, binder dosage, assay methods, and baseline phosphate levels.

Fig. 2.

Meta-analysis of the effect of phosphate binders on FGF-23 levels from baseline. CI, confidence interval; FGF-23, fibroblast growth factor-23; SD, standard deviation

In the subgroup analysis by NCPB type, sevelamer was associated with a significant reduction in FGF-23 levels (SMD −0.33, 95% CI −0.56 to −0.10, p = 0.005) with low heterogeneity (I² = 29%). Lanthanum was also associated with a significant reduction (SMD −1.13, 95% CI −2.10 to −0.15, p = 0.02), although heterogeneity was considerable (I² = 93%). The test for subgroup differences indicated no statistically significant difference between the binder types (p for subgroup difference = 0.12) (Fig. 3a).

Fig. 3.

Subgroup meta-analysis of the effect of phosphate binders on FGF-23 levels from baseline. (a) By phosphate binder type, (b) by dialysis status, and (c) by treatment duration. Covic et al. reported data from two follow-up durations, 9 weeks and 25 weeks, which were evaluated separately in our analysis. CI, confidence interval; SD, standard deviation

In the subgroup analysis based on the dialysis status, a statistically significant reduction in FGF-23 levels was observed in non-dialysis patients (SMD −0.98, 95% CI −1.77 to −0.19, p = 0.02), although heterogeneity was high (I² = 90%). In patients undergoing dialysis, the reduction was not statistically significant (SMD −0.32, 95% CI −0.68 to 0.04, p = 0.08), with substantial heterogeneity (I² = 70%). There was no statistically significant difference between subgroups (p = 0.13) (Fig. 3b).

In the subgroup analysis by treatment duration, studies with a treatment duration < 12 weeks showed a significant reduction in FGF-23 levels (SMD −0.39, 95% CI −0.64 to −0.13, p = 0.003) with low heterogeneity (I² = 30%). Studies with a treatment duration ≥ 12 weeks also demonstrated a significant reduction (SMD −0.65, 95% CI −1.24 to −0.07, p = 0.03), although heterogeneity was high (I² = 90%). There was no statistically significant difference between subgroups (p = 0.42) (Fig. 3c).

Risk of bias assessment

The risk of bias results are summarized in Fig. 4a. Among the included studies, only one was judged to have a low risk of bias. Five studies were considered high-risk, and five were rated as having “some concerns.” The primary sources of bias were missing outcome data and selective reporting (Fig. 4b).

Fig. 4.

Summary of the risk of bias assessment. (a) The individual judgments for the 11 included studies and (b) the overall distribution of bias risk for each domain

Sensitivity analysis

We performed two sensitivity analyses. First, after excluding the five studies judged to be at high risk of bias, the pooled analysis of the remaining six studies showed that the effect of NCPBs on FGF-23 levels remained significant, with a larger effect size than that observed in the primary analysis (SMD −1.04, 95% CI −1.79 to −0.29, p = 0.007) (Fig. 5a). Next, leave-one-out analysis was conducted. When the study by Soriano et al. [21], identified as the largest contributor to heterogeneity, was excluded, the pooled effect remained statistically significant (SMD −0.35, 95% CI −0.59 to −0.11, p = 0.004) (Fig. 5b). Further influencing diagnostics, including the identification of potential outlier studies, could provide additional insights into the heterogeneity.

Fig. 5.

Sensitivity analyses to assess the robustness of the main findings. (a) Forest plot of the meta-analysis after excluding studies with a high risk of bias and (b) forest plot of the meta-analysis after excluding the study by Soriano et al., the largest contributor to heterogeneity. CI, confidence interval; CPB, calcium-based phosphate binder; NCPB, non-calcium-based phosphate binder; SD, standard deviation

Publication bias

Potential publication bias was evaluated using the meta package in R (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria). Visual inspection of the funnel plot revealed asymmetry, with one study positioned as an outlier with a large standard error (Fig. 6). Egger’s regression test confirmed this finding (p = 0.040). To further explore this bias, additional analyses, such as trim-and-fill or contour-enhanced funnel plots, could be applied to test whether the pooled effect size remained stable after adjustment. It is also important to interpret these findings cautiously, recognizing that publication bias may reduce the certainty of the conclusions. As such, the results should be viewed as supportive and hypothesis-generating evidence, rather than as definitive evidence.

Fig. 6.

Funnel plot to assess potential publication bias

Discussion

This meta-analysis demonstrated that NCPBs are significantly more effective than CPBs in reducing serum FGF-23 levels among patients with CKD-MBD (SMD −0.56, 95% CI −0.96 to −0.16, p = 0.005). This finding is clinically relevant, as reductions in FGF-23 may reflect improvements in disordered phosphate homeostasis and have been associated with improved cardiovascular outcomes in patients with CKD [27].

The apparent differences between CPBs and NCPBs in their effects on FGF-23 levels may be partly explained by their distinct mechanisms of action. CPBs bind phosphate in the gastrointestinal tract but also provide an exogenous calcium load, which may enhance intestinal calcium absorption and suppress PTH secretion via calcium-sensing receptors [28]. However, impaired renal phosphate excretion in CKD allows hyperphosphatemia to persist, which in turn stimulates osteocytic secretion of FGF-23 [29]. In contrast, NCPBs selectively bind to intestinal phosphate without supplying calcium, effectively lowering serum phosphate levels without directly influencing PTH levels [28, 29]. Although these mechanistic differences are biologically plausible, they should be interpreted as hypotheses rather than established causal explanations. Alternative contributors, including variability in FGF-23 assays and differences in baseline phosphate control, dietary intake, and residual renal function, may have also influenced the observed differences. In addition, this meta-analysis included only sevelamer and lanthanum among NCPBs to reduce mechanistic heterogeneity. Therefore, the findings may not be generalizable to other agents, such as iron-based phosphate binders or tenapanor, which have different biological effects. Acknowledging these uncertainties highlights the need for further mechanistic studies to confirm the causal pathways instead of relying solely on inferences from clinical trial outcomes.

The beneficial effect of NCPBs on FGF-23 levels was consistent across treatment durations, ranging from 8 weeks to 12 months. Although the subgroup analysis suggested a more pronounced reduction in FGF-23 levels among non-dialysis patients, the interaction with dialysis status was not statistically significant (p = 0.13). These findings suggest that the advantages of NCPBs may extend across dialysis statuses, although the potentially stronger effects in earlier CKD stages warrant further investigation.

FGF-23 levels increase sharply from the earliest stages of CKD [25] and serve as an independent predictor of left ventricular hypertrophy, cardiovascular disease, and all-cause mortality [27, 30, 31]. Importantly, the 2017 Kidney Disease: improving Global Outcomes (KDIGO) CKD-MBD guidelines recognized FGF-23 as a key mediator of mineral metabolism, preceding changes in phosphate, calcium, and PTH levels, and emphasized its role as a central regulator of CKD-MBD pathophysiology [32]. In this context, our study provides the first meta-analysis of RCTs that directly compared the effects of NCPBs and CPBs on FGF-23 levels, thereby identifying a clinically meaningful treatment target beyond conventional biochemical markers.

Our findings are consistent with evolving perspectives on CKD-MBD management. The 2023 KDIGO Controversies Conference highlighted the importance of managing CKD-MBD not only from a bone health perspective but also with a strong focus on cardiovascular complications, such as left ventricular hypertrophy and vascular calcification [14]. The observed superiority of NCPBs in lowering FGF-23 levels aligns with these updated therapeutic priorities.

This study has several methodological strengths, including restriction to RCTs, which minimized confounding factors and enhanced internal validity. However, many of the included trials were judged to have “some concerns” or a “high risk” of bias, which tempered the certainty of the conclusions. Furthermore, although the included trials represented multiple countries and patient populations, our analysis did not systematically evaluate regional or demographic subgroups. Therefore, claims regarding geographic and population diversity should be interpreted with caution and should not be overstated.

This study also has some limitations. First, substantial clinical heterogeneity was observed among the included studies. For example, baseline FGF-23 levels varied markedly, particularly between dialysis-dependent CKD stage 5 patients and earlier stages, reflecting fundamental pathophysiological differences such as impaired renal clearance, persistent hyperphosphatemia, and limited FGF-23 removal by dialysis [33]. To mitigate this, we used SMDs and performed subgroup analyses stratified by dialysis status, confirming the overall robustness of our findings. In addition, variability in serum phosphate control across studies may have influenced FGF-23 levels, making it difficult to distinguish whether the observed reductions were attributable to the intrinsic effects of phosphate binders or secondary to phosphate lowering. Although C-terminal FGF-23 fragments may have distinct biological activity and potential clinical relevance, as suggested by prior studies [34] our analysis was limited to measurements of intact FGF-23. Therefore, we were unable to evaluate the specific contribution or clinical impact of C-terminal FGF-23 fragments, which should be addressed in future studies. Second, concomitant therapies, such as vitamin D supplementation and dietary factors, may also influence FGF-23 levels. In particular, variations in vitamin D management across studies—including continuation of prior therapy, discontinuation, or protocol-driven administration—may have contributed to residual confounding. Finally, the risk of bias assessments indicated “some concerns” or “high risk” in several trials, and the funnel plot asymmetry and Egger’s test (p = 0.04) suggested potential publication bias. Although we applied a random-effects model and conducted sensitivity analyses, excluding high-risk studies, both of which supported the robustness of our results, these limitations necessitate cautious interpretation.

Future studies should address these issues. First, large-scale RCTs with rigorous control of concomitant therapies are needed to confirm the differential effects of CPBs and NCPBs on FGF-23 levels. Second, further investigation of dose–response relationships and potential modifiers, such as residual renal function and dialysis modalities is warranted. Third, real-world evidence is needed to determine whether the reduction in FGF-23 levels with NCPBs translates into improved long-term clinical outcomes, including reduced cardiovascular events, mortality, and cost-effectiveness in clinical practice.

Conclusions

This meta-analysis demonstrated that NCPBs are more effective than CPBs in lowering serum FGF-23 levels in hyperphosphatemia patients with CKD-MBD. These findings underscore the potential clinical advantages of NCPBs in managing disordered mineral metabolism beyond conventional biochemical markers.

However, the results should be interpreted with caution due to substantial heterogeneity, risk of bias in several of the included trials, and the possibility of publication bias. Although the subgroup analyses suggested potential variations in treatment effects by drug type, dialysis status, and treatment duration, none of these interactions were statistically significant. Therefore, these findings should be considered exploratory and hypothesis generating rather than definitive.

Future research should focus on large-scale, high-quality RCTs with rigorous methodological designs to confirm the comparative effects of CPBs and NCPBs on FGF-23 levels. Further investigation is warranted to clarify the underlying mechanisms, explore potential effect modifiers, such as the CKD stage and residual renal function, and evaluate whether reductions in FGF-23 translate into improved long-term clinical outcomes, including reductions in cardiovascular events and mortality.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

CI

Confidence interval

CKD

Chronic kidney disease

CKD-MBD

Chronic kidney disease–mineral and bone disorder

CPB

Calcium-based phosphate binder

FGF-23

Fibroblast growth factor 23

NCPB

Non-calcium-based phosphate binder

PTH

Parathyroid hormone

RCT

Randomized controlled trial

SD

Standard deviation

SMD

Standardized mean difference

Author contributions

SHC, CWJ, and HK conceptualized and designed the study. SHC, CWJ, MYB, and HK performed the investigation and methodology. SHC and MYB wrote the original draft. HK supervised the study and reviewed and edited the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) [NRF-2021R1F1A1052976]. The funders had no role in the study design, data collection, analysis, interpretation of data, or writing of the manuscript.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.