Glucose Metabolic Abnormalities and Their Interaction With Defective Phosphate Homeostasis in Tumor-induced Osteomalacia

Abstract

Context

Phosphate homeostasis was compromised in tumor-induced osteomalacia (TIO) due to increased fibroblast growth factor 23 (FGF23) secretion. Nevertheless, the glucose metabolic profile in TIO patients has not been investigated.

Objectives

This work aimed to clarify the glucose metabolic profiles in TIO patients and explore their interaction with impaired phosphate homeostasis.

Methods

20 TIO patients, 20 individuals with normal glucose tolerance, and 20 patients with type 2 diabetes mellitus (DM) were enrolled and underwent an oral glucose tolerance test (OGTT). Serum phosphate and FGF23 concentration were monitored during OGTT.

Results

In patients with TIO, 60% (12/20) exhibited impaired glucose tolerance (IGT) and 5% (1/20) had type 2 DM. Those with IGT or type 2 DM experienced more ambulatory difficulties (69.2% vs 42.9%), lower phosphate concentrations (0.43 ± 0.10 vs 0.53 ± 0.10, P = .042), and lower calcium concentrations (2.20 ± 0.08 vs 2.30 ± 0.40, P = .001) compared to TIO patients without these conditions. According to correlation analysis, serum phosphate levels were negatively correlated with plasma glucose levels at 60 minutes (P < .001), fasting plasma insulin levels (P < .05), and homeostasis model assessment for insulin resistance (P < .05). Those with high FGF23 levels had a higher glucose level at 60 minutes (10.5 [9.3, 12.3] vs 7.3 [6.4, 10.1], P = .048) than that of low group. After glucose loading, both FGF23 and phosphate levels exhibited a decreasing trend.

Conclusion

The development of diabetes in TIO patients may be predisposed by ambulatory issues, low phosphate, and elevated FGF23 levels. Dysglycemia might further aggravate hypophosphatemia.

Tumor-induced osteomalacia (TIO) is an acquired disease due to excessive secretion of fibroblast growth factor 23 (FGF23), primarily produced by phosphaturic mesenchymal tumors. It is featured by hypophosphatemia, appropriate normal or low 1,25-dihydroxyvitamin D [1,25(OH)₂D] level, and increased serum alkaline phosphatase (ALP) levels (1, 2). The clinical symptoms include bone pain, height loss, skeletal abnormalities, bone fracture, and muscle weakness, with the potential for severe disability and impact on life quality if the diagnosis is delayed. In addition to skeletal and muscle dysfunction, glucose abnormality often occurs in hereditary hypophosphatemic rickets (3, 4). We conducted a retrospective analysis on 104 TIO patients hospitalized in our center between 2018 and 2020 and revealed a high incidence of diabetes among TIO patients, with a rate of 11.5% (12/104) (Supplementary Table S1) (5). However, the detailed glucose metabolic profile has not been reported in TIO patients.

Low phosphate levels have been associated with disrupted glucose sensitivity in both healthy individuals (6) and type 2 DM patients (7), highlighting the potential importance of phosphate in glucose metabolism. In TIO patients, hypophosphatemia was primarily regulated by FGF23 by interacting with coreceptor Klotho (8). FGF23 is a hormone generated by bone cells including osteoblasts as well as osteocytes, and emerging evidence suggests that FGF23 exerts glucose metabolic effects (9). In humans, insulin resistance was positively related with FGF23 in patients with chronic kidney disease (10). Nevertheless, the relationship between FGF23 and glucose metabolic abnormalities has not been explored in cases of extremely elevated FGF23 in TIO patients. We hypothesized that serum phosphate, FGF23, and other phosphate-regulated hormones such as parathyroid hormone (PTH) and 1,25(OH)₂D might jointly participate in glucose regulation in TIO patients. Besides, it is still unknown whether dysglycemia is caused by osteomalacia-related symptoms or defective phosphate homeostasis in hypophosphatemia individuals.

The interaction between bone and glucose metabolic markers is crucial for comprehending the development of impaired glucose homeostasis in TIO. Hence, this study aimed to outline the glucose metabolic profiles and investigate the possible interplay between abnormal phosphate metabolism and glucose metabolic effects in TIO patients.

Subject and Methods

Subjects and Study Design

This research included all hospitalized patients with clinical suspicion of TIO from January 2021 to December 2021 at Peking Union Medical College Hospital. The criteria for inclusion were (1) the onset age over 18; (2) at least 2 low serum phosphate level records and low tubular maximum phosphate reabsorption per glomerular filtration rate (TmP/GFR); (3) osteomalacia manifestations including fatigue, bone pain, and pseudofractures or fractures confirmed by radiological imaging or bone scans; (4) pathological diagnosis of phosphaturic mesenchymal tumor. Pregnant patients were excluded, as well as patients who were previously diagnosed with type 1 diabetes or specific types of diabetes, such as pancreatogenic diabetes, glucocorticoid-induced diabetes, etc. Fasting C-peptide and glutamic acid decarboxylase antibody were tested in patients with dysglycemia. Patients with a fasting C-peptide level lower than 0.6 ng/mL or positive glutamic acid decarboxylase antibody were excluded. Finally, 20 TIO patients were recruited. They were matched with 20 individuals with normal glucose tolerance (NGT) and 20 type 2 DM patients based on similar age, sex, and body mass index (BMI). The participants were randomly selected from the Beijing suburbs cohort, and detailed information about this cohort has been described in a previous study (11). All participants provided informed consent. The local ethics committee and the Peking Union Medical College Hospital scientific research department gave their approval for this study.

Clinical and Biochemical Measurement

After fasting for at least 10 hours, each participant had an oral glucose tolerance test (OGTT). After 75 g of anhydrous glucose was administered, levels of blood glucose, insulin, and C-peptide were measured at 4 different intervals (0, 30, 60, and 120 minutes). The concentrations of FGF23 were examined at both 0 and 120 minutes. Phosphate levels were also monitored at the 4 time points mentioned earlier in TIO patients. Samples of blood and urine were analyzed in the Department of Laboratory Medicine, with the remainder stored at −80 °C. Prior to blood collection, all TIO patients refrained from taking phosphorus and vitamin D supplements. An autoanalyzer (Beckman Coulter AU5800, USA) was utilized to detect ALP, serum calcium, phosphate, glucose, total cholesterol, triglyceride (TG), low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, as well as creatinine. An electrochemiluminescence system (E601 analyzer, Switzerland) was utilized for the analysis of C-terminal cross-linking telopeptide of type I collagen, procollagen type 1 N-terminal propeptide, and 25-hydroxyvitamin D (25OHD). Nomogram was used to calculate TmP/GFR (12). Twenty-four-hour urine phosphate and 24-hour urine calcium were analyzed by an autoanalyzer (Beckman Coulter AU2700). Bone mineral density was assessed using dual-energy x-ray absorptiometry (Prodigy Advance, GE Healthcare, USA). Radioimmunoassay (DIAsource ImmunoAssays SA, Belgium) was used to test 1,25(OH)₂D levels. FGF23 was tested by iFGF23 ELISA Kit (Kainos Laboratories, Japan, RRID: AB_3105844). Clinical features including age, sex, BMI, waist circumference, smoking history, drinking history, physical activity, bone fracture, and disease duration were collected from medical records.

Clinical Feature Definition

BMI was calculated as weight/height². Normal weight was characterized as 18 ≤ BMI ≤ 24 kg/m², while overweight was 25 ≤ BMI<28 kg/m², and obesity was defined as a BMI ≥ 28 kg/m² (13). The diagnosis of abdominal obesity was waist circumference ≥ 90 and 85 cm in males and females, respectively. Dyslipidemia was described as total cholesterol ≥ 6.2 mmol/L or TG ≥ 2.26 mmol/L (14). Hypertension was determined as systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg (15). Impaired glucose tolerance (IGT) was established through fasting blood glucose (FBG) levels<7.0 mmol/L and 2-hour postprandial plasma glucose (PPG) ranging from 7.8 to 11.0 mmol/L following OGTT. Type 2 DM was diagnosed either FBG levels were ≥ 7 mmol/L, or if 2-hour PPG levels were ≥ 11.1 mmol/L during OGTT (16). Metabolic syndrome was identified by central obesity and meeting at least 2 of the following criteria: (1) TG ≥ 1.7 mmol/L, (2) high-density lipoprotein cholesterol < 1.03 mmol/L for males or < 1.29 mmol/L for females, (3) fasting glucose ≥ 5.6 mmol/L, (4) systolic blood pressure ≥ 130 and/or diastolic blood pressure ≥ 85 mmHg (17). Secondary hyperparathyroidism (SHPT) was characterized by total calcium within the range of 2.13 to 2.70 mmol/L and elevated PTH levels (>68.0 pg/mL). Low bone mass was determined by a Z-score of femoral neck, total hip, or vertebral body < –2. Ambulatory challenges included reliance on wheelchairs or crutches. 25OHD deficiency was defined as 25OHD < 20 ng/mL.

Assessment of Insulin Resistance

We adopted the homeostasis model assessment for insulin resistance (HOMA-IR) and homeostasis model assessment for β-cell function to evaluate insulin resistance. The calculation formula of HOMA-IR was [FBG (mmol/L) × fasting insulin (μU/mL)]/22.5. The homeostasis model assessment for β-cell function was calculated as 20 × [fasting insulin (μIU/mL)/(FBG (mmol/L) – 3.5)] (18).

Statistical Analyses

Statistical analyses were performed using software version SPSS 27.0, version 10 of GraphPad Prism, RStudio Server 1.4.1103, and Origin 2022. Mean ± SD was used for normally distributed continuous variables, while medians with interquartile range were used for nonnormally distributed data. Categorical variables were represented as percentages. Clinical features between TIO with and without IGT/type 2 DM were compared using chi-squared tests. The correlation between bone metabolic parameters and glucose metabolic parameters was evaluated with the Spearman correlation coefficient. The Kruskal-Wallis test was applied for nonnormally distributed continuous variables and 1-way ANOVA for normally distributed variables when comparing the 3 groups. A P-value below .05 was considered statistically significant.

Results

Baseline Characteristics for TIO Patients

Table 1 illustrates the clinical features of 20 patients with TIO. The research involved 11 males (55%) and 9 females (45%) with an average age of 45 ± 11 years. The average BMI level was 27.54 ± 4.64 kg/m². The level of serum phosphorus was 0.47 ± 0.11 mmol/L, and the TmP/GFR level was 0.39 ± 0.14 mmol/L, which were both markedly lower than the normal range. Other laboratory examinations indicated an extremely high FGF23 level (345.9 [102.82, 1113.69], pg/mL), an appropriately low 1,25(OH)₂D level (11.52 [5.95, 18.05], pg/mL), a high ALP level (258.95 ± 114.85 U/L), a high TG level (1.75 [0.99, 2.62], mmol/L), and a high UA level (399.35 ± 111.26 mmol/L). In terms of symptoms related to hypophosphatemia, 17 (85%) patients had bone pain and 15 (75%) patients had bone fractures. Additionally, 12 (60%) patients had walking difficulties and 14 (70%) patients had low bone mass. The main clinical diagnosis related to metabolic problems included impaired glucose tolerance (12/20, 60%), type 2 DM (1/20, 5%), overweight (11/20, 55%), obesity (6/20, 30%), abdominal obesity (17/20, 85%), dyslipidemia (10/20, 50%), hypertension (9, 45%), and metabolic syndrome (14/20, 70%).

OGTT Results for the 3 Cohorts

The fluctuations of glucose, insulin, and C-peptide among TIO, NGT, and DM groups during OGTT are presented in Fig. 1. The baseline characteristics and OGTT results are respectively shown in Supplementary Tables 2 and 3 (5). As shown in Fig. 1A, the overall trend of glucose levels in TIO patients was higher than NGT controls and lower than type 2 DM patients. Although FBG had no difference in the TIO and NGT groups (4.90 [4.63, 5.35] vs 5.30 [5.00, 5.40] mmol/L, P = .535), 2-hour PPG in TIO patients were considerably higher than NGT controls (8.65 [7.10, 9.48] vs 6.90 [6.13, 7.38] mmol/L, P = .042) and lower than DM patients (8.65 [7.10, 9.48] vs 15.80 [13.40, 18.68] mmol/L, P < .001). In addition, the peak secretion time of plasma insulin and C-peptide was delayed to 120 minutes during OGTT in both TIO and DM patients.

Clinical Index Between TIO Patients With or Without IGT or Type 2 DM

The comparison of biochemical parameters is shown in Table 2. Out of 20 TIO patients, 13 (65%) were diagnosed with IGT/type 2 DM, comprising 7 males and 6 females. Patients in the IGT/type 2 DM group were notably older (48 ± 10 vs 38 ± 10, P = .043) and had higher levels of glycated hemoglobin A1c (5.30 [5.18, 5.43] vs 4.90 [4.68, 5.15], P = .013) compared to the other group. Interestingly, patients with IGT/type 2 DM exhibited significantly lower serum phosphate and calcium compared to those without IGT/type 2 DM (0.43 ± 0.10 vs 0.53 ± 0.10, P = .042; 2.20 ± 0.078 vs 2.30 ± 0.40, P = .001). Additionally, the 1,25(OH)₂D concentrations in patients with IGT/type 2 DM were also lower with marginal significance (9.98 [4.59, 17.38] vs 27.53 [8.85, 46.79], P = .073). No statistical differences were found in disease duration and bone mineral density analysis values in TIO. Figure 2 provides an overview of clinical features frequency in TIO patients with and without IGT/type 2 DM. The IGT/type 2 DM group had a higher bone pain rate (92.3% vs 71.4%) and bone fracture rate (84.6% vs 57.1%) and similar low bone mass rate (69.2% vs 71.4%) compared to those without IGT/type 2 DM. Furthermore, patients with IGT/type 2 DM experienced more ambulatory difficulties in activities of daily living (69.2% vs 42.9%). It is worth noting that the prevalence of IGT/type 2 DM in TIO patients with walking difficulties was approximately 75%, which was higher than that in TIO patients without walking difficulties (50%). In the IGT or type 2 DM group, 3 individuals were diagnosed with nephrocalcinosis and 2 of them also had SHPT, conditions that were not observed in those without IGT/type 2 DM. Furthermore, the IGT/type 2 DM group had a greater occurrence of metabolic comorbidities including obesity, dyslipidemia, and hypertension, as well as metabolic syndrome compared to those without IGT/type 2 DM.

Relationship Among Biochemical Markers of Bone and Glucose Metabolism in TIO Patients

A correlation analysis (Fig. 3) indicated an inverse association between serum phosphate and the glucose metabolic parameters, encompassing glucose level at 60 minutes (P < .001), fasting plasma insulin concentrations (P < .05), and HOMA-IR (P < .05). TmP/GFR exhibited a negative correlation with glucose concentrations at 30 minutes (P < .05) and 60 minutes (P < .001). PTH demonstrated a negative correlation with glucose at 30 minutes (P < .001), even after adjusting for 25OHD and 1,25(OH)₂D. Additionally, 24-hour urinary calcium showed a positive relationship with FBG (P < .05).

Glucose Levels Under Different Levels of Phosphate-regulated Hormones in TIO Patients

Figure 4 shows the glucose fluctuations during OGTT in TIO patients across varying FGF23, PTH, and 1,25(OH)₂D concentrations. Each hormone was categorized into high and low level groups based on median values. The trends showed that glucose levels in the high FGF23, high PTH, and low 1,25(OH)₂D groups were consistently higher compared to the respective low level groups. In particular, at 60 minutes, the glucose concentrations of the high FGF23 group were noticeably higher than those of the low group (10.5 [9.3, 12.3] vs 7.3 [6.4, 10.1], P = .048), and at 30 minutes, the high PTH group exhibited notably higher glucose levels compared to the low group (9.9 [9.65, 11.25] vs 8.5 [6.6, 9.23], P = .009).

The Trend Change of Phosphate and FGF23 During OGTT

The fluctuation trend of serum phosphate and insulin are presented in Fig. 5A. During OGTT, the level of serum phosphate showed a downward trend. The levels of FGF23 at 120 minutes were also obviously decreased compared to the levels of fasting FGF23 in the TIO group (P = .015), DM group (P = .012), and NGT group (P = .009); these are presented in Fig. 5B, 5C, and 5D, respectively.

Discussion

This research is the first to describe the glucose metabolic profile and its association with defective phosphate metabolic homeostasis in TIO patients. Key findings of this study were (1) TIO patients had multiple metabolic complications including abdominal obesity and impaired glucose tolerance, (2) osteomalacia symptoms and abnormal phosphate homeostasis might be related to the occurrence of diabetes, and (3) abnormal glucose homeostasis might further exacerbate hypophosphatemia.

The majority of patients with TIO in our study experienced bone pain and fractures, primarily due to impaired bone mineralization. These skeletal issues significantly impacted their daily physical activities, with some patients facing prolonged immobility following fractures (1). Our findings revealed that 60% of TIO patients had difficulty walking and appeared to have a higher incidence of dysglycemia. This suggested that limited mobility in TIO patients might contribute to the development of glucose metabolism disorders. Interestingly, our study did not observe a correlation between disease duration and dysglycemia, indicating that the severity of osteomalacia symptoms might be a more significant factor. Additionally, factors such as aging and comorbidities like overweight, hypertension, and dyslipidemia could also increase the susceptibility to diabetes in these patients. Managing glucose metabolism in individuals with TIO necessitates a comprehensive approach that includes blood pressure, lipid levels and weight management, and appropriate physical activity.

Hypophosphatemia emerged as the most prominent biochemical feature in TIO patients (19). Consistent with previous research, our study noted moderately decreased phosphate levels in TIO patients. Phosphate is critical in various biological processes, including energy production, bone mineralization, and protein synthesis (20). Several studies have highlighted the close connection between phosphate levels and glucose metabolism. Adequate phosphorus intake has been proven to decrease postprandial glucose levels as well as improve insulin sensitivity, indicating its positive impact on glucose metabolism (21). Conversely, a deficiency in phosphate might result in elevated glucose levels and insulin resistance (22). A major factor influencing the restricted rate of glycolysis was found to be insufficient ATP caused by hypophosphatemia (23), thereby influencing glucose tolerance. Our study observed that phosphate levels significantly decreased in TIO patients with dysglycemia. In addition, negative correlations between phosphate levels and glucose levels as well as insulin resistance were found in TIO patients, indicating the potential relationship between phosphate and glucose metabolism. Alongside serum phosphate levels, TmP/GFR emerged as another significant laboratory marker in TIO, reflecting the extent of renal phosphate wasting and disease severity. However, no statistical difference was found in TmP/GFR between TIO patients with and without dysglycemia, which might be attributed to the simultaneous effect of FGF23, 1,25(OH)2D, and PTH on TmP/GFR. Therefore, long-term glucose abnormalities might not have an obvious influence on TmP/GFR. Notably, a significant negative association was identified between 1-hour PPG levels and both serum phosphate levels and TmP/GFR, indicating that both phosphate and TmP/GFR might respond to the glucose status during OGTT. Recently, the International Diabetes Federation proposed that 1-hour PPG might be a more effective indicator for detecting impaired glucose tolerance in high-risk diabetic populations compared to 2-hour PPG, due to compromised glucose sensitivity in the compensatory stage of insulin secretion (24). This suggested that 1-hour PPG could also be sensitive to the early stages of impaired glucose tolerance resulting from hypophosphatemia. Furthermore, in addition to its impact on glucose metabolism, phosphate also played a role in obesity. Our study revealed that 85% of TIO patients experienced abdominal obesity, which was in line with the amount of obesity in patients with X-linked hypophosphatemic rickets (4). The increased incidence of obesity in individuals with hypophosphatemia may be attributed to heightened food intake resulting from decreased ATP production. Moreover, a high-phosphate diet was found to aid in weight loss by promoting lipolysis (25).

In the progression of TIO, excessive serum FGF23 plays a pathogenetic part in hypophosphatemia and contributes to decreased 1,25(OH)₂D levels (26). It was shown that FGF23 takes part in glucose metabolism. A study on FGF23-deficient mice revealed improved glucose tolerance (27). Elevated levels of FGF23 have been noted in individuals with prediabetes compared to individuals with normal glucose tolerance. In patients with insulin resistance, FGF23 levels were also increased when compared with normal controls (28). Similarly, a positive association between insulin resistance and FGF23 has been found in chronic kidney disease patients (10). However, in adolescents with obesity, HOMA-IR showed a negative correlation with FGF23 levels (29). The impact of FGF23 in glucose metabolism remains inconclusive. Our research revealed that although their statistical significance was not found among FGF23 and glucose metabolic variables, TIO patients with higher FGF23 levels tended to have worse glucose status, particularly at 60 minutes during OGTT. However, we could not clarify whether dysglycemia in TIO is independently correlated with elevated FGF23 or attributes to decreased phosphate levels and reduced 1,25(OH)₂D concentrations, as FGF23 primarily acts as an upstream regulatory hormone in phosphate homeostasis. Our research found that TIO patients with dysglycemia had lower levels of 1,25(OH)₂D, suggesting a possible involvement of this hormone in dysglycemia. PTH, another important hormone that controls calcium and phosphate levels, was unexpectedly elevated in TIO patients due to either lack of vitamin D or prolonged phosphate intake (29). Emerging evidence suggested a positive relationship between PTH and dysglycemia, with a 15.3% incidence rate of type 2 DM in individuals diagnosed with primary hyperparathyroidism, surpassing that observed in the general population (30). Additionally, in White individuals, a positive relationship was found between PTH levels and the development of diabetes (31). A positive relationship between PTH and glucose levels at 30 minutes was observed in TIO patients, even after adjusting for 1,25(OH)₂D and 25OHD concentrations. Furthermore, 2 TIO patients with SHPT exhibited abnormal glucose metabolism, suggesting a potential role of PTH in glucose regulation. More investigation is necessary to elucidate the correlation between glucose metabolism and PTH. Therefore, we hypothesized that dysglycemia might be the result of the combined action of these 3 phosphate-regulated hormones.

While previous discussions have highlighted the potential impact of bone metabolism indicators, such as decreased serum phosphate, we could not rule out the bidirectional influence between them. Clinical research has indicated that inadequate glycemic management among individuals diagnosed with type 2 DM may lead to reduced bone mineralization and a heightened susceptibility to bone fractures (32). It was found that TIO patients with dysglycemia displayed diminished calcium and phosphate levels, alongside a heightened incidence of bone fractures. Although there appeared to be a trend toward diminished bone mineral density in TIO patients with IGT/type 2 DM, the absence of statistical significance might be attributed to the restricted size of the study sample. Additionally, 24-hour urinary calcium levels were positively related to glucose in TIO patients; 3 dysglycemic TIO patients had nephrocalcinosis. Hypercalciuria was recognized as a contributing factor to the onset of nephrocalcinosis (33), possibly attributed to elevated glucose levels causing osmotic diuresis, thereby leading to elevated urinary calcium excretion (34).

To further clarify the role of insulin in phosphate and FGF23, we tested their levels during OGTT. We found both phosphate and FGF23 levels decreased after glucose loading in TIO patients. In line with our findings, this downward trend was also found in healthy and overweight individuals (35, 36). Insulin was reported as a negative regulator of FGF23. In an in vivo study, when insulin deficiency was induced in mice, FGF23 obviously increased (37). Interestingly, in TIO patients, FGF23 levels were mainly regulated by tumor secretion, while in the OGTT experiment, FGF23 showed a significant downward trend, suggesting that FGF23 levels could rapidly respond to high glucose and compensatory insulin secretion in the short term rather than tumor secretion. However, due to the impact of FGF23 secreted by tumors on TIO patients, the long-term effect of FGF23 secretion under dysglycemia might be affected by the secretion status of the tumor itself, which might partially explain why there was no significant change of FGF23 observed in the TIO patients with dysglycemia. In addition, it was speculated that increased insulin levels during OGTT in TIO might induce the shift of serum phosphate to intracellular compartment, thereby having negative feedback on FGF23 secretion, which contributed to FGF23 increasement (22). However, in another study, it was proven that changes in FGF23 were earlier than phosphates during OGTT. One of the underlying mechanisms was that insulin exerted a direct effect on FGF23 through the phosphoinositide 3-kinase/protein kinase B/Akt/forkhead box protein O1 signaling pathway (37). Theoretically, FGF23 is a negative regulator of phosphate. However, during OGTT, FGF23 and phosphate levels both decreased in TIO patients in our study. In our study, even if FGF23 might decrease before serum phosphate during OGTT, it seemed that phosphate decrease was not alleviated accompanied by FGF23 decrease, which might suggest that insulin has a stronger inhibitory effect on serum phosphorus. These results might indicate that if TIO patients suffer from insulin resistance and pancreatic islets secreted a higher level of insulin to compensate, the symptoms of hypophosphatemia can be aggravated, although we cannot verify the long-term harm in the present study. Therefore, during the disease process of TIO, we might also need to pay attention to the glucose condition and observe the effect of glucose improvement on phosphate homeostasis in the future. Besides, in TIO patients, the underlying mechanism between phosphate metabolism and glucose metabolism was complicated, which needs further study.

There are still some limitations. First, the sample size was limited due to TIO being a rare disease. Except for serum phosphate, we did not find a significant difference in other bone metabolism indicators such as TmP/GFR between TIO patients with and without dysglycemia status, which might be due to the small sample. Second, we explored the relationship among serum phosphorus, FGF23, and glucose homeostasis under an extreme condition, which limited the generalizability of these findings. To further expand generalizability, we will subsequently expand the sample size and include patients with different levels of elevated FGF23 and hypophosphatemia in the future. Third, it was a retrospective study and no follow-up was conducted. Therefore, we could not determine participants’ glucose control and long-term prognosis after surgery. Fourth, because this study design was cross-sectional, we were unable to demonstrate a causal relationship between abnormal glucose metabolism and a further decrease in serum phosphate as well as symptom aggravation in TIO patients. Fifth, FGF23 measurement used frozen samples, which might have bias.

Conclusion

Patients with TIO are prone to experiencing various metabolic issues, especially impaired glucose tolerance. Factors such as restricted daily activities and disrupted phosphate balance could increase the risk of developing dysglycemia. Conversely, dysglycemia might further worsen hypophosphatemia in TIO patients.

Acknowledgments

The authors acknowledge all the participants from the Peking Union Medical College Hospital and Beijing suburb cohort.

Funding

This research was supported by the National Key R&D Program of China (2021YFC2501700), CAMS Innovation Fund for Medical Sciences (2021-I2M-1-002), National High Level Hospital Clinical Research funding (2022-PUMCH-D-006), the National Natural Science Foundation of China (No. 81970757, No. 82270938), and the nonprofit Central Research Institute Fund of the Chinese Academy of Medical Sciences (2023-PT320-10).

Disclosures

The authors have nothing to disclose.

Data Availability

Some or all datasets generated and/or analyzed during the current study are not publicly available but can be obtained from the corresponding author upon reasonable request.

References

Author notes