Impaired Physical Performance in X-linked Hypophosphatemia Is not Caused by Depleted Muscular Phosphate Stores

Abstract

Context

X-linked hypophosphatemia (XLH) is a rare genetic disease, characterized by renal phosphate wasting and complex musculoskeletal manifestations including decreased physical performance.

Objective

To characterize muscular deficits in patients with XLH and investigate phosphate stores in muscles.

Methods

Case–control study (Muscle fatigability in X-linked Hypophosphatemia [MuXLiH]) with a 1-time assessment at the German Aerospace Center (DLR), Cologne, from May to December 2019, including patients with XLH cared for at the Osteology Department, University of Wuerzburg. Thirteen patients with XLH and 13 age/sex/body weight–matched controls aged 18-65 years were included. The main outcome measure was ³¹P-magnetic resonance spectroscopy (³¹P-MRS)–based assessment of phosphate metabolites in the soleus muscle at rest. Further analyses included magnetic resonance imaging–based muscle volume measurement, laboratory testing, isokinetic maximum voluntary contraction (MVC), fatigue testing, and jumping mechanography.

Results

By means of ³¹P-MRS, no significant differences were observed between XLH and controls regarding phosphate metabolites except for a slightly increased phosphocreatine to inorganic phosphate (PCr/Pi) ratio (XLH: 13.44 ± 3.22, control: 11.01 ± 2.62, P = .023). Quadriceps muscle volume was reduced in XLH (XLH: 812.1 ± 309.0 mL, control: 1391.1 ± 306.2 mv, P < .001). No significant differences were observed regarding isokinetic maximum torque (MVC) adjusted to quadriceps muscle volume. Jumping peak power and jump height were significantly reduced in XLH vs controls (both P < .001).

Conclusion

The content of phosphoric compounds within the musculature of patients with XLH was not observed to be different from controls. Volume-adjusted muscle strength and fatiguability were not different either. Reduced physical performance in patients with XLH may result from long-term adaptation to reduced physical activity due to skeletal impairment.

X-linked hypophosphatemia (XLH, OMIM: #307800) is a rare genetic disorder caused by dominantly inherited or spontaneous pathogenic variants within the phosphate-regulating endopeptidase homolog, X-linked gene (PHEX) on the short arm of the X chromosome (Xp22.1) leading to elevated circulating concentrations of phosphatonin fibroblast growth factor 23 (FGF23) (1). XLH is the most common form of heritable rickets (2) and has an incidence of approximately 1:20 000 (3). Under physiological conditions, bone-forming cells, specifically osteocytes, are considered to be the main source of FGF23 (4). Elevated levels of FGF23 cause renal phosphate wasting by limiting NaPi IIa/c expression in the proximal renal tubules, thereby compromising phosphate reabsorption from urine (2). In addition, FGF23 interferes with vitamin D metabolism by downregulating 1-alpha-hydroxylase (CYP27B1) activity and stimulating expression of 24-hydroxylase (CYP24A1), thus reducing activation of vitamin D to calcitriol while enhancing its degradation (2). Growing evidence supports potential further effects of elevated FGF23 (3) and PHEX deficiency itself (5) on the phenotypic manifestation of XLH. All of this contributes to the clinical presentation of XLH, which is characterized by bone deformities due to rickets, delayed growth and motor development in children, short stature, gait abnormalities, oral abscesses, pain, and compromised physical performance. Particularly in adult patients with XLH, the clinical burden associated with the disease appears to be largely determined by mobility restrictions, musculoskeletal pain, and muscle weakness (6, 7), which could at least partly be interpreted as a result of the skeletal phenotype.

However, phosphoric compounds are also known to have a direct relevance for skeletal muscle function. This applies foremost to the energy-rich phosphates: adenosine trisphosphate (ATP), adenosine diphosphate, and phosphocreatine (PCr). While ATP is degraded to adenosine diphosphate as a necessary energetic step during the actin–myosin cross-bridge cycle, ATP pools are readily replenished by transphosphorylation of phosphate derived from PCr by creatine kinase (CK), thereby ascertaining near-constant ATP levels even during muscle fatigue (8, 9). In addition, phosphate plays important roles with regards to phosphorylation of glucose to glucose-6-phosphate, and in phosphorylation cascades of intracellular signaling enzymes (10). In line with this, Veilleux et al (11) have demonstrated that juvenile patients with XLH have increased bone mass and size while still presenting muscle function deficits. Accordingly, muscle weakness is considered an inherent clinical feature of XLH due to both suboptimal ergonomics resulting from skeletal deformities and compromised muscle function associated with deficient phosphate metabolism (3, 11, 12). However, it is still unknown whether the latter is truly a consequence of deficient or altered intramyocellular inorganic phosphate (Pi) and whether such an intracellular Pi deficiency affects the energy metabolism of muscles. Since the musculature of patients with XLH shows no signs of rigor, it can be assumed to contain normal levels of ATP at resting state.

The reaction of CK is typically far on the side of PCr with only low levels of Pi during a rest. So, a general Pi deficiency of a myocyte at rest might also entail reduced levels of PCr.

It has been shown in mice that hypophosphatemia leads to decreased ATP synthetic flux and may thus be associated with deficient restoration of phosphatic compounds in the dynamic state (13). This is supported by an observation of phosphofructokinase-deficient patients (14). In these patients, Pi released from PCr during muscle contraction was bound to sugar residues and could not be released due to the enzyme deficiency. During recovery from exercise, the lack of accessible Pi results in a slower and eventually incomplete recovery of PCr. Considering the above and given that serum levels of Pi are typically reduced in patients with XLH, deficient levels of intracellular Pi may also occur, affecting muscular energy metabolism. However, XLH and its effects on the skeleton may also affect the musculature indirectly by reduced physical activity causing potential adaptations in terms of atrophy and fiber type composition (15, 16).

Therefore, the primary aim of the present study was to assess intramuscular phosphate metabolites in patients with XLH at resting state. It was hypothesized that an altered phosphate ratio of the muscle would lead to greater fatiguability in patients with XLH. Furthermore, we endeavored to assess neuromuscular power, isokinetic muscle strength, and muscle volume of our patients.

Materials and Methods

Subjects

Thirteen patients with X-linked hypophosphatemia (XLH group) and 13 age, sex, and weight-matched control participants (control group) were included in this study, with 7 women and 6 men in each group. The patients were recruited from the Orthopedic Hospital, König-Ludwig-Haus in Würzburg and informed consent was obtained from the participants. Inclusion criteria comprised adult patients aged 18-65 years either diagnosed with XLH (XLH group) or healthy subjects without any major illness and specifically without a musculoskeletal disorder (control group). Exclusion criteria comprised any acute injury or chronic disease that precluded performance of the 2-legged vertical jump test or the isokinetic muscle test, any contraindication for magnetic resonance imaging (MRI) such as claustrophobia, ferromagnetic implants, pacemakers, or pregnancy, and participation in regular or organized exercise or sport activities for the control group to preclude having athletes as comparators. The genetic background on PHEX variants of subjects with XLH and phosphate supplementation can be found elsewhere (17). Further information on their bone deformities and surgical history has been previously published in an article on bone geometry (18). The study was designed and conducted before treatment with the monoclonal anti-FGF23 antibody burosumab was approved and licensed for adult patients with XLH in Europe. None of the participants received such treatment before or at the time of assessment.

Study Design

This is a case–control study conducted at the :envihab research facilities at the Institute of Aerospace Medicine, German Aerospace Center (DLR) in Cologne, Germany. All study participants were invited to the DLR premises for a 1-day testing protocol between May and December 2019, and matched XLH and control participants were tested on the same day.

Determination of Body Composition

As an extension of the standard anthropometric data including age (years), height (cm), body mass (kg) and body mass index (BMI) (kg/m²), body composition regarding body fat content, and skeletal muscle mass were determined by bioelectrical impedance analysis using the InBody S10 device (InBody Co., LTD, Seoul, Korea) with LookinBody 120.3.0.0.11-S/N:G424O1116 software. Results of 1 subject with XLH had to be excluded from the bioelectrical impedance analysis because of a technical error.

Biochemical Analyses of Blood and Urine Samples

Biochemical analysis of bone and mineral metabolism markers relevant for the pathophysiology of XLH was performed from blood and urine samples that were collected prior to the physical and technical assessments. Laboratory workup was performed by a certified clinical routine laboratory including the following parameters: c-terminal fibroblast growth factor (cFGF23), parathyroid hormone 1-84 (PTH), alkaline phosphatase (ALP), 25-OH vitamin D (native vitamin D), 1,25-OH vitamin D (active Vit D), n-terminal telopeptide of type 1 collagen (NTX), calcium (Ca²⁺), phosphate, thyroid-stimulating hormone, CK, lactate dehydrogenase (LDH), C-reactive protein, total serum protein, glucose, and serum creatinine. Urine samples were assessed for creatinine, calcium, and phosphate levels to enable computation of the glomerular filtration rate (GFR), tubular reabsorption of phosphate (TRP), and ratio of tubular maximum reabsorption rate of phosphate to glomerular filtration rate (TmP/GFR).

Considering logistical requirements such as the journey to Cologne and in order to avoid underperformance or bias due to medication deficits or hypoglycemia, participants were allowed to have breakfast and take their regular medication and supplements until 1 hour before the assessments.

Determination of Phosphate Metabolites by ³¹P-Magnetic Resonance Specroscopy

³¹P-magnetic resonance spectroscopy (³¹P-MRS) was performed using a 3 Tesla Siemens Biograph mMR system (Fig. 1), a cylindric ³¹P/¹H resonator coil with 18 cm inner diameter and 20 cm sensitive length (Rapid Biomedical GmbH, Rimpar, Germany), and the chemical shift imaging sequence provided by the manufacturer.

All MR based examinations were performed before the exercise-based assessments. The lower right leg was placed in the resonator on a 3D-printed plastic support that stabilizes the plantar flexor muscles in the center of the resonator. The right foot was kept in a neutral position on a pedal. The acquisition parameters were pulse angle 90°, repetition time TR = 1050 ms, echo time TE = 2.3 ms, 64 averages, vector size 2048. An axial matrix of 8×8 voxels was located over the right calf muscle. The voxel size was 25 mm × 25 mm × 50 mm. The voxel matrix was located with the help of 3 ¹H localizer scans to place at least 1 voxel into the soleus muscle avoiding neighboring muscles. Using the software package jMRUI v5.2 (www.jmrui.eu), the ³¹P signal of 2 voxels, which fit into the soleus muscle, were added in the time domain as free induction decay improving the signal to noise ratio. The free induction decay was filtered for noise and was converted into a spectrum by fast Fourier transformation.

The peak areas of Pi, phosphocreatine (PCr), and the beta-signal of ATP (Fig. 2) as well as the intracellular pH derived from the chemical shift of Pi relative to PCr were analyzed in arbitrary units (a.u.) using the AMARES based algorithm included in jMRUI v5.2. As no information is available on the absolute ATP concentration in muscles of patients with XLH, we did not quantify metabolites to deliver phosphorus values. Instead, we determined the endpoints as ratios of peak areas such as PCr/ATP, PCr/Pi, and Pi/ATP. Furthermore, the intracellular pH was calculated from the difference of chemical shifts of Pi and PCr.

Determination of Muscle Volume by MRI

MRI was performed on the same system as ³¹P-MRS. Five stacks of 64 axial images of the right leg were acquired from the foot to the lower part of the lumbar spine. One subject with XLH had major surgery of her upper leg musculature. In her case, we performed the MRI and isokinetic tests on the left leg. Images were recorded using a turbo 2-echo DIXON sequence with following parameters: flip angle 10°, TR 7.02 ms, TE1 2.46 ms, TE2 3.69 ms, 300 mm × 300 mm field of view at 1.2 mm × 1.2 mm × 4 mm voxel size. For the quantification of muscle volume, we used the image representing the signal intensity of water ¹H spins, which represents the muscle tissue as bright voxels surrounded by dark voxels for bone, fat, and fascia. The volumes of following muscles were evaluated using a semi-automatic software for the differential segmentation of muscle areas in the axial images (ROI-segmenter, iPattern, Institute for Pattern Recognition, University of Niederrhein, Krefeld, Germany), dissecting the following muscle compartments: musculus gluteus maximus, m. gluteus medius, m. gluteus minimus, m. iliopsoas, m. tensor fasciae latae, m. rectus femoris, mm. vasti femoris (ie, the muscle group comprising m. vastus medialis, m. vastus intermedius, and m. vastus lateralis), m. soleus, and m. gastrocnemius medialis et lateralis.

Muscle Strength and Fatigue Tests

These tests were preceded by a general warm-up exercise of 5 minutes at 75 W on a bicycle ergometer. The torque of isokinetic maximum voluntary contraction (MVC) of the right knee extensor musculature was determined at an angular velocity of 60°/second on an IsoMed 2000 device (D.&R. Ferstl GmbH, Hernau, Germany). In 1 case the left leg was examined owing to previous surgery on the right side. Subjects were seated with the upper body elevated by 80° and well fixated with a hip belt and with fixation pads placed on their shoulders. After a warm-up of 5 repetitions at submaximum torque and a break of 1 minute, MVC of knee extension was determined as the highest torque value reached within 3 repetitions.

After 3 minutes of recovery following isokinetic testing, participants performed a fatigue protocol consisting of 20 continuous cycles of knee extensions and flexions at 180°/second over the full range of motion (20°-95° knee flexion). Subjects were asked to do all contractions with maximum power predominantly trying to increase the speed of the knee extension. Fatigue of the knee extension was determined as the decrement between the maximum torque value and mean torque over the last 3 contractions (in %). Torque maximum was typically reached within the first 4 contractions.

Jumping Mechanography

All participants performed 3 countermovement jumps on a Leonardo Mechanograph Type 8N600304A (Novotec Medical GmbH Pforzheim, Germany) using the integrated software (Research Edition 4.4b01.35). Subjects were asked to stand still on the platform, and after the software-based registration of body weight and a quiet stance for 1 second, an acoustic signal was provided for the participant to start the countermovement jump. During the jumping test, the subjects kept their arms passively by the side. This arm position allowed easier balancing of the body than placement of the hands to the hips. Subjects were then asked to jump as high as possible. For safety reasons, 1 operator was standing by at each side.

The Leonardo software computed velocity, power, and elevation of the body's center of mass from the time course of the ground reaction force (19). As end points, we determined the peak power (W), peak power normalized to body mass (W/kg), peak force (kN), peak force normalized to body mass and expressed in multiples of the Earth's acceleration (g), jump height (mm), and the Esslinger fitness index (%) representing an experience based, age-adjusted, body mass–specific relative peak power (20).

Statistical Analysis

Statistical analyses were performed using the SPSS statistic software version 27.0 (SPSS Inc, Chicago, IL) and Prism 8/9 (GraphPad Software, San Diego, CA), which we also used to create graphs. Differences between groups were tested by unpaired t tests after testing for normal distributions with the Kolmogorov–Smirnov test. When significant deviations from the normal distribution were found, Mann–Whitney tests were performed. All values shown in the text, tables, and figures represent mean values ± SD.

Open-source software R version 4.1.2 was used to analyze group differences (XLH vs control) in the correlation between 2 variables, in other words volumes of quadriceps muscles referred to body height (see Fig. 6) and maximum isokinetic torque referred to volume of the quadriceps muscle or to body mass, respectively (see Fig. 7).

Results

Anthropometric Data and Body Composition

Participants of the XLH group and the control group were matched for age, sex, and body weight. There were no significant between-group differences regarding these parameters. However, while participants in the XLH group were significantly shorter than the controls (P = .002) and had significantly lower absolute skeletal muscle mass (P = .033), their skeletal muscle mass index (skeletal muscle mass adjusted to height squared) was no different from the controls (XLH 10.3 ± 1.3 kg/m², control 10.9 ± 1.5 kg/m², P = .332).

Patients in the XLH group had insignificantly higher average values for BMI (P = .390) and body fat percentage (P = .090). Details are provided in Table 1.

Laboratory Parameters of Mineral and Bone Metabolism, Renal Function, Tissue Integrity, and Energy Metabolism

Serum levels of phosphate were significantly lower in the XLH group than in controls (P < .001) while levels of cFGF23 were significantly elevated in the XLH group (P < .001). Participants in the XLH group exhibited significantly reduced TRP and TmP/GFR (P < .001). The serum levels of native vitamin D were significantly lower in controls vs XLH (P = .016), while active vitamin D was not significantly different between both groups (P = .114). Furthermore, there were no significant differences regarding serum concentrations of calcium, sodium, and potassium. However, serum ALP activity was significantly elevated in the XLH group (P = .004), along with significantly increased serum levels for PTH (P < .001) and NTX (P = .002). Serum levels for thyroid-stimulating hormone, γ-glutamyltransferase, glutamate oxaloacetate transaminase, and glutamate pyruvate transaminase as well as LDH did not show any significant differences. Similarly, there were no significant between-group differences regarding serum creatinine, CK, total protein, and blood glucose levels. See Table 2 for details.

Phosphate Metabolites in Muscle Measured With ³¹P-MRS

Results of comparative analyses regarding phosphate metabolites are illustrated in Fig. 3. One female patient with XLH with an extremely high BMI could not be examined in the MR magnet because she could not fit into the gantry without causing pain in her hips. Overall, ³¹P-MRS measurements revealed a significantly higher PCr/Pi ratio of (mean ± SD) 13.44 ± 3.22 in the XLH group vs 11.01 ± 2.62 in controls (P = .023). Further parameters and ratios did not show significant between-group differences, specifically the Pi/ATP ratio (XLH: .20 ± 0.04, control: 0.23 ± 0.05, P = .117), PCr/ATP (XLH: 2.58 ± 0.50, control: 2.39 ± 0.34, P = .324), and intracellular pH (XLH: 7.02 ± 0.03, control: 7.02 ± 0.03, P = .946) were similar among the groups. For details see Fig. 3.

Volumes of Selected Leg and Hip Muscles

Valid muscle volume scans were available for 12 patients with XLH and 12 controls. Assessments for 1 healthy male and 1 female patient were incomplete and had to be excluded from the volumetric analysis. Except for the m. tensor fascia latae, the m. iliopsoas, and the m. gluteus minimus, the volumes of all examined hip and leg muscles were significantly smaller in XLH than in controls. After adjustment for stunted growth in XLH by calculating muscle volume per height squared, this difference was no longer significant for most muscles assessed. However, muscle volume was substantially reduced and still significantly lower even after adjustment for reduced body height in XLH concerning the mm. vasti femoris group representing the predominant part of the knee extensor musculature as well as the m. soleus and m. gastrocnemius lat.: the main aspects of the musculature controlling plantar flexion and stabilization of the ankle. Details are provided in Table 3.

Maximum Torque During Voluntary, Isokinetic Knee Extension at 60°/second

Isokinetic knee extension maximum torque was significantly lower in XLH than in controls (XLH: 71 ± 30 Nm, control: 145 ± 50 Nm, P < .001). When maximum torque was normalized to the individual body mass, the performance of XLH was only 52% of the control value (XLH: 0.84 ± 0.38 Nm/kg, control: 1.71 ± 0.56 Nm/kg, P < .001), indicating for XLH an impaired knee extension force with respect to the body mass to be accelerated during motion (see Fig. 4). Furthermore, this difference was also statistically significant after adjustment for body height (XLH: 0.45 ± 0.18 Nm/cm, control: 0.86 ± 0.25 Nm/cm, P < .001).

Muscle Fatigue During 20 Times Repeated Voluntary Isokinetic Knee Extension at 180°/seconds

The torque maximum at 180°/second, which was typically reached within the first 4 repetitions, and was significantly lower in XLH than in controls (XLH: 68 ± 8 Nm, control: 112 ± 8 Nm, P < .001); this difference was still statistically significant after adjustment for body weight (XLH: 0.87 ± 0.28 Nm/kg, control: 1.35 ± 0.46 Nm/kg, P = .003) and height (XLH: 0.43 ± 0.14 Nm/cm, control: 0.66 ± 0.16 Nm/cm, P = .001). However, the indicator of relative fatigue, which was the relative decrease of torque between the initial torque maximum and the mean of the torque maxima during the last 3 of 20 repetitions, was not significantly different between groups (P = .295, see Fig. 5B). The corresponding increase in electromyography (EMG) amplitude, indicating an increase in motor unit recruitment, was also not significantly different between the 2 groups (P = .147, see Fig. 5A). However, while some of the XLH group participants exhibited a decrease in motor unit recruitment, this was not observed in controls.

Performance of Countermovement Jumps

Jumping performance results were significantly inferior in XLH compared with controls for peak power (P < .001), peak power/body mass (P < .001), peak force (P < .026), jump height (P < .001), and Esslinger fitness index (P < .001). The group difference of peak force/body mass did not quite reach significance. Details are provided in Table 4.

Linear Correlation of Knee Extensor Muscle Volume and the Body Height in Group Comparison

Considering people with XLH having short stature and specifically shorter limbs and thus shorter knee extensor muscles than controls, the correlation of quadriceps muscle volume to body height was not significantly different (analysis of variance [ANOVA] on group P = .058, Fig. 6A). A concordant correlation was also found for the rectus femoris muscle, as a component of the quadriceps muscle (ANOVA P = .341, Fig. 6C). However, focusing on the 3 heads of the vastus femoris group, we found a significant difference between XLH and controls in the correlations of muscle volume and body height (ANOVA P = .026, Fig. 6B), showing that for patients with XLH, this component of the quadriceps muscles was smaller than in controls even after adjustment for body height.

Group Comparison of Linear Correlation Between Maximum Torque, Muscle Volume, Strength, and Body Mass

In both groups, isokinetic maximum torque at 60°/second increased with the volume of the quadriceps muscle with no group difference between the correlation models (ANOVA on group P = .154, Fig. 7A). However, in XLH the strength of knee extensor muscles was significantly lower with respect to body mass than in controls, and in both groups maximum torque and body mass did not significantly correlate with each other (see Fig. 7B).

Discussion

While functional restraints and compromised physical performance are well-studied characteristics of XLH in adults, the underlying pathophysiological mechanisms are not completely resolved. In order to better understand actual muscular deficits in patients with XLH and to specifically elucidate if these are associated with deficient intramyocellular availability of Pi or phosphoric compounds, this study compared ³¹P-MRS and exercise-based muscular assessments in affected adult patients and age, sex, and body weight–matched healthy controls.

Regarding intramyocellular phosphate metabolites, the results presented here do not confirm the hypothesis of deficient concentrations of Pi or phosphoric compounds as a major cause of altered muscle energy metabolism, force development, or fatigability. In both groups, the obtained phosphorus spectra from soleus muscle at rest were of equal quality and ratios of concentrations, and the intracellular pH in general did not show significant differences between groups, except for a moderately higher PCr/Pi ratio in the XLH patient group. However, since the PCr/Pi ratio in resting muscles is known to depend on the fiber type composition (21, 22), attributing this finding to a disease-related reduction of intramyocellular Pi is unlikely, specifically considering the absolute difference is low and in case of supposed pathognomonic difference, larger disparities were to be expected. Accordingly, we would rather assume this to be better explained by a slightly higher proportion of fast twitch fibers in patients with XLH (21, 22) in order to compensate for overall reduced muscular capacity resulting from lower physical activity of patients in comparison with controls (15, 16). If this interpretation is correct, the observed deficit in jumping power in patients with XLH was of even greater concern, given that jumping performance is supposed to be superior in people with predominance of fast twitch fibers like, for instance, in sprint-trained athletes (23).

As a technical limitation, phosphorus concentrations had to be assessed in arbitrary units, since calculation of absolute concentrations would require separate measurements of standard solutions, mimicking MR conditions within the cytosol of the muscle fiber, or the beta signal of ATP as an internal standard (24, 25), and both means of quantification would require identical conditions in all participants, which is not warranted here. Consequently, the results do not allow for meaningful interindividual comparisons and would require a muscle biopsy.

Further, since the PCr/Pi ratio is considered to be an indicator of mitochondrial function (26) with levels typically declining upon exercise, the fact that we do not have data on the dynamics of this parameter upon exercise is another important limitation that should be considered in future studies. Considering animal data on reduced ATP synthetic flux in hypophosphatemia, it is conceivable that despite unremarkable findings at resting state, deficits may become apparent upon exercise with deficient regeneration of ATP and/or PCr (13). Thus, in order to unravel potential differences regarding the time course of depletion and specifically recovery of phosphatic compounds in XLH, future studies should assess these spectra at various time intervals following different types of exercise intervention.

Laboratory evaluations confirmed typical disease-specific alterations in all patients with XLH, including elevated levels of FGF23 and deficient circulating phosphate along with compromised indicators of renal phosphate absorption (1). Consequently, elevated ALP and increased levels of NTX appear indicative of increased bone turnover due to hypophosphatemic osteomalacia, specifically since active and native vitamin D levels were balanced in both groups, and native vitamin D as an indicator of appropriate supply was even higher in patients than in controls due to consequent supplementation.

Interestingly, despite the equivalent calcium levels in serum in both groups, urinary calcium concentration was lower in patients with XLH. Whether this is solely a consequence of elevated PTH or potentially enhanced by increased FGF23 levels acting on TRPV5 in the distal tubule should be evaluated in further studies. The clinical sequelae of XLH-associated rickets and osteomalacia on bone geometry in this cohort were described in a previous manuscript (18).

Baseline CK values were normal and equal in both groups, affirming that none of the participants had suffered acute muscle damage before the assessments. Similarly, there were no differences regarding LDH and hepatic transaminases.

The volume of hip and leg musculature in subjects with XLH was significantly smaller in comparison with muscles of control subjects, who were matched for age, sex, and body weight. The smaller muscle volume was largely attributable to the smaller body height of patients with XLH. However, for major and functionally critical muscle groups like the m. quadriceps femoris and specifically the vastus group as well as the mm. soleus and gastrocnemius lat., this study revealed significant reductions of muscle volume even after adjustment of body height. Considering that a relevant impact of alterations of the leg alignment and specifically the Q-angle on thigh muscle volume which could be assumed in XLH has been ruled out earlier (27), this appears most likely to be a consequence of deconditioning associated atrophy. In line with that, muscle volume of subjects with XLH was also much smaller than in controls when adjusted for body weight.

In both, XLH and controls the maximum isokinetic knee extension torque followed the same correlation with the volume of the entire quadriceps muscle. However, maximum torque and torque normalized to body mass were much lower in XLH than in controls, indicating deficient muscular power. Furthermore, maximum torque was also significantly lower in XLH than in controls after adjustment for reduced body height in XLH. Based on that, it appears reasonable to conclude that this deficit in torque of patients with XLH predominantly results from the lower muscle volume following disuse associated atrophy rather than an intrinsic muscle problem. This is further supported by the observation that characteristics of fatigue were similar in both groups and all patients with XLH, except for 1, could finish the complete set of 20 repetitions of maximum knee extension and flexion at a contraction velocity of 180°/seconds, although at distinctly lower force levels in the XLH group.

Accordingly, we hypothesize that reduced performance, specifically reduced force of hip and leg muscles in subjects with XLH, predominantly result from an adaptation to disuse rather than to a specific phosphate deficiency. However, compromised biomechanics owing to skeletal deformities and early-onset osteoarthritis in patients with XLH with both leading to inferior physical performance may also have contributed to the above differences.

In a previous study, Veilleux et al reported functional data on muscle weakness as an inherent clinical aspect of hypophosphatemic rickets (11). While confirming these functional observations and expanding the spectrum of assessments, our data could specifically show that muscular deficits in XLH are predominantly due to compromised utilization abetted by deformities, early osteoarthritis, and pain rather than a consequence of deficient muscle energetics and metabolism.

In line with that, the vast majority of patients with XLH in this study had undergone orthopedic surgery and had longstanding bone and joint problems, which all together can readily explain such behavioral adaptations of the musculature. The reduced mobility of patients with XLH would also provide a reasonable explanation for the very high BMI and increased proportional body fat content in most of the participants, which again imposes a high risk of secondary health issues related to metabolic syndrome.

Considering the musculature of persons with XLH is physiologically healthy and not directly affected by the deficiency in circulating phosphate, preventive measures including physiotherapy and muscle training appear feasible and a promising approach to avert deficits in physical performance and to avoid overweight-related and immobility-associated health issues. Accordingly, early optimal treatment to correct phosphate wasting, prevent skeletal deformities and support an active lifestyle appear pivotal to facilitate lifelong mobility.

Conclusions

We encountered no primary phosphate deficiency in XLH muscle. Decreased physical performance in XLH appears to be associated with reduced muscle volume and low muscle strength, most likely resulting from adaptive processes to lifelong disuse putatively owing to the skeletal phenotype. Accordingly, optimized metabolic control of the disease along with regular physical activity could prevent muscular deterioration in patients with XLH.

Acknowledgments

We thank Friederike Körber and Mirko Rehberg for their support in the planning phase and Wolfram Sies for his engagement in :envihab. We also thank the Institute of Medical Statistics and Computational Biology at University of Cologne for their kind support.

Funding

Aspects of this study have been supported by third-party funding from Kyowa Kirin GmbH. Additional funding included the Research Track Award of the medical faculty of university of Cologne to Johannes Alexander Serhan Kara and DLR-internal funding (cost object 2047030).

Author Contributions

Data collection: J.A.S.K., J.Z., J.R., W.S., L.S. Drafting manuscript: J.A.S.K., J.Z. Revising manuscript: J.R., O.S., L.S. Supervision: J.Z., J.R., L.S. Planning of study: J.A.S.K., J.R., J.Z., O.S., L.S.

Disclosures

L.S. has received honoraria for lectures and advice and research grants to the institution from KyowaKirin. All other authors report no conflict of interest to disclose. Aspects of this research have been supported by third-party funding from Kyowa Kirin GmbH.

Data Availability

The data that support the findings of this study are not publicly available due to them containing information that could compromise the privacy of research participants. Nonidentifying data can be made available on reasonable request from the corresponding author (L.S.).

Registration of the Study

The study was registered at the German Register for Clinical Studies (Deutsches Register Klinischer Studien) (registered January 27, 2020). Registration no.: DRKS 00016074.

Compliance With Ethical Standards

Ethics approval for the study was obtained from the ethics committees of the medical faculties at the Universities of Cologne (No. 19-1020) and Würzburg (128/19).

This research is in accordance with the principles set out in the Declaration of Helsinki.

References

Abbreviations

 
• ANOVA

  analysis of variance

 
• ATP

  adenosine trisphosphate

 
• BMI

  body mass index

 
• cFGF23

  c-terminal fibroblast growth factor

 
• CK

  creatine kinase

 
• GFR

  glomerular filtration rate

 
• LDH

  lactate dehydrogenase

 
• MRI

  magnetic resonance imaging

 
• MRS

  magnetic resonance spectroscopy

 
• MVC

  maximum voluntary contraction

 
• NTX

  n-terminal telopeptide of type 1 collagen

 
• PCr

  phosphocreatine

 
• PCr/Pi

  phosphocreatine to inorganic phosphate

 
• Pi

  inorganic phosphate

 
• PTH

  parathyroid hormone 1-84

 
• TmP

  tubular maximum reabsorption rate of phosphate

 
• TRP

  tubular reabsorption of phosphate

 
• XLH

  X-linked hypophosphatemia