Application of bone metabolic parameters in the diagnosis of growing pains

Abstract Objective The present study aimed to assess the diagnostic significance of serum bone metabolic parameters in children with growing pains (GPs). Methods All patients diagnosed with GP and healthy controls matched with age and gender were recruited at the outpatient clinic of Children's Hospital at Zhejiang University School of Medicine from August 2016 to August 2021. In all subjects, serum levels of calcium (Ca), phosphorus (P), procollagen type‐I N‐terminal (PINP), parathormone (PTH), 25‐hydroxyvitamin D (25‐(OH)D), osteocalcin (OC), N‐terminal cross‐linked telopeptides of type‐I collagen (CTX), and tartrate‐resistant acid phosphatase type 5b (TRACP5b) were investigated. The univariate analysis, multivariate logistic regression analysis, and receiver operating characteristic (ROC) curve were used to identify the bone metabolic parameters factors for diagnosing GP. Results We enrolled 386 children with GP and 399 healthy controls in present study. The mean age of GP group was 5.319 years, and, primarily, the subjects were preschool‐age children. The gender ratio (male‐to‐female) was 1.27 in GP group. After adjusting for age and gender, we identified that the serum levels of Ca (p < 0.001, OR: 25.039), P (p = 0.018, OR: 2.681), PINP (p < 0.001, OR: 1.002), and PTH (p = 0.036, OR: 0.988) were independent diagnostic factors associated with GP. Area under curve (AUC) of the ROC curves was in the order: PINP (0.612) > Ca (0.599) > P (0.583) > PTH (0.541). A combination of independent diagnostic factors and multivariable logistic regression analysis provided a refined logistic regression model to improve the diagnostic potential, of which the AUC had reached 0.655. Conclusions Serum levels of Ca, P, PINP, and PTH could be independent diagnostic factors associated with GP. The logistic model was significantly superior to bone metabolic parameters for diagnosing GP.


| INTRODUC TI ON
Growing pains (GPs) were first described in 1823 by French physician Marcel Duchmap as the most common recurrent leg pain problem in children, which occur in about 3% to 47% of children with no organic lesions. 1,2 There is no evidence that GP is associated with particularly rapid growth. 3 GP mostly affects children aged 3-12 years, and the highest incidence is found in the 4-to 6-year age bracket. 3 GP is almost always bilateral, usually nonarticular, and often located in the shins, calves, thighs, or popliteal fossa. 4 The pain usually occurs at night, with intensity varying from mild to very severe, and resolves by morning. The severity and duration of pain varies from child to child. 5 In severe cases, the pain can occur daily. The duration of pain ranges from minutes to hours.
Some children experience regular pain, whereas others have days or months of pain-free periods between episodes, which vary from child to child. 6 Previous studies found association between GP and abdominal pains, nonspecific headaches, and behavioral disorders. 7 Meanwhile, GP is not related to serious organic lesions and usually resolves by late childhood. The physical examination showed no objective signs of inflammation. 8,9 However, frequent pain may have tremendous impacts on the children and their family's daily routines including activity, sleep, frequent use of pain relief medications, and causing anxiety. 10,11 The etiology of GP has still not been fully elucidated, despite its discovery for over 200 years ago. Many different theories have been postulated, but none have been confirmed. Unknown etiology contributes to diagnostic difficulty that the diagnosis of GP requires the exclusion of other diseases with similar clinical manifestations. 12 Thereby, GP has been more of a diagnosis of exclusion than an explicit entity with ambiguity in diagnostic criteria, which has become a hot research topic.
According to some studies, the bone mineral density of some children with GP was significantly less than that of healthy children, and hypovitaminosis D may play a role in the pathogenesis of GP. 4,13 Therefore, we speculate that altered bone metabolism may occur in children with GP. Traditionally, densitometry techniques were recognized as a gold standard for the assessment of bone status. However, these measures only reflect the static state of bone tissue. 14 The detection of bone metabolism markers can reflect the dynamic state of bone metabolism, and they show a significantly higher sensitivity at the early stage of bone metabolism changes. 15 Metabolites released by osteoblasts during bone formation and bone matrix fragments and secretions produced by osteoclasts during bone resorption enter the bloodstream and can be detected. They are divided into three types: markers for bone formation, markers for bone resorption, and calcium (Ca) and phosphorus (P) metabolism indicators. For bone formation, markers reflecting osteoblast activity are by-products of osteoblastic enzymes, matrix proteins, or collagen synthesis, including procollagen type-I N-terminal (PINP) released during the procession of type-I procollagen into collagen and osteocalcin (OC) synthesized by mature osteoblasts. [16][17][18] For bone resorption, markers reflecting osteoclast activity are degradation products of type-I collagen, such as C-terminal cross-linked telopeptides of type-I collagen (CTX) and tartrate-resistant acid phosphatase type 5b (TRACP5b) which is an enzyme secreted by osteoclasts. 13 Calcium and phosphorus metabolism indicators mainly include 25-hydroxyvitamin D (25(OH)D), parathormone (PTH), calcium (Ca), and phosphorus (P). The homeostasis of calcium and phosphate is mainly regulated by 25(OH)D and PTH, promoting bone mineralization. 19 Advances in the correlation between the above bone metabolic parameters and GP may provide potential alternative diagnostic targets in diagnostic studies of GP.
The study aimed to investigate differential bone metabolism

| Laboratory tests
Fasting venous blood were drawn from all subjects in the morning.
We measured serum concentrations of CTX and TRACP5b as bone resorption indexes and serum concentrations of PINP and OC (we

| Statistical analysis
Data analysis was performed using SPSS software (SPSS 19.0). Plots of the study were generated using GraphPad Prism 6.0 software.
Quantitative variables were reported as mean and 95% confidence interval (95% CI). Independent samples T test or nonparametric test was used to compare means between children with GP and controls.
Pearson's Chi-square test was used to analyze the categorical variables. Multivariate logistic regression analysis was used to identify independent predictors and construct the refined logistic regression

| Demographic data
From August 2016 to August 2021, a total of 386 children with GP were collected, which included 216 boys and 170 girls and the male-to-female ratio was 1.27. The mean age of the study population was 5.319 (5.118, 5.520) years (range 1-12 years).
According to age, 22 the subjects were divided into four subgroups: 0-2 years old (infancy and childhood), 3-6 years old (preschool age), 7-12 years old (school age), and more than 12 years old (puberty). Children with GP were primarily preschool-age children  Rather, serum level of PTH was higher in GP group than in healthy controls (Z = −1.981, p < 0.05), and there was no significant difference in serum OC, CTX, TRACP5b, and 25-(OH)D levels between GP group and healthy controls (all p > 0.05).

| Diagnostic value of bone metabolic parameters
For diagnosis of GP, a multivariate logistic regression analysis was performed to assess significant bone metabolic parameters variables with p < 0.1 which were derived from the univariate analy-  Table 1.

| The role of bone metabolic parameters in diagnosing GP
Diagnostic performance of bone metabolic parameters was assessed by applying the ROC curve. As is shown in Figure 3, area under ROC curve was the highest for PINP level, followed by Ca level, P level, and PTH level. With the utilization of ROC curve and Youden index, the cut-off values of serum PINP, Ca, P, and PTH levels for diagnosing GP were determined. Relevant sensitivity and specificity at the maximum of the Youden index were calculated and shown in Table 2.  Ca supplementation may increase bone strength density. 13 Simon found that Ca supplementation may ease GP. 28 Qamar et al. also found that a minority of children with GP have hypocalcemia (6%) and hypophosphatemia (3%) in their study, 4 but there was also no control group used for hypothesis tested. In this study, we found that serum concentrations of Ca and P were significantly decreased in GP than healthy controls that supported Simon's view. In addition, the present study indicated that PTH is an independent predictor for GP, and it is also an important regulator of Ca and P levels. This may suggest that the mechanism that PTH involved in the disease GP is not merely achieved by affecting Ca or P level, relevant mechanism needs to be investigated in detail.
With the change of growth rate, the serum concentration of bone markers varied with age and sex. 29 Bone formation parameters increase in infancy, and bone formation and bone resorption parameters increase in both sexes during adolescence. 30 In view of this, age and gender were considered for inclusion in the multivariate regression modeling analysis. After having adjusted for age and gender, bone formation-related marker PINP and calcium and phosphorus metabolism indicators of PTH, Ca, and P remained independent predictors for GP diagnosis.
In this study, bone resorption markers of CTX and TRACP5b were not associated with GP. This suggests that bone resorption is not involved in the pathogenesis and course of GP, and relevant research was rather scarce. In our case-control study, no association was found between vitamin D (VD) levels and GP. However, the theory that VD deficiency is related to GP is often mentioned. In a study including 120 children with GP, 86.6% cases were found to be deficient in VD. After 3 months of oral vitamin supplementation, the level of 25(OH)D was significantly increased, and the pain score was significantly decreased. 25 Morandi et al. also found that VD levels were significantly increased and pain intensity significantly decreased by VD supplementation in children with GP. 21 However, one drawback of these studies was that there was no control group for hypothesis test or some studies have smaller sample sizes. 3 In addition, because GP is

| CON CLUS IONS
In conclusion, the present study suggests that bone metabolism parameters including serum levels of PINP, Ca, P, and PTH were

ACK N OWLED G EM ENTS
We thank the patients and family members for their participation in this study. This work was funded by Analysis and Testing Foundation of Zhejiang Province (LGC19H200006) and The National Natural Science Foundation of China (8197060176).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interests.

AUTH O R CO NTR I B UTI O N S
Huamei Li and Bing Wang designed the research and wrote the article. Huamei Li, Bing Wang, Lin He, and Ran Tao completed the experimental part of the study, collected the laboratory parameters and patients' information, performed the statistical analysis, and drew the figures. Shiqiang Shang supervised the entire study and provided academic guidance throughout the study process.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.