Introduction

Recurrent lower limb pains, termed growing pains or growth pains, constitute the most frequent cause of musculoskeletal pain in children. Growing pains mostly affect children between the ages of 3 and 12 years, and their prevalence varies from 2.6 % to 49.4 % [1–3]. Growing pains are usually non-articular, almost always bilateral, and often located in the shins, calves, thighs, or popliteal fossa. The pain usually appears late in the day or at night, often awaking the child, with a duration of less than 72 h and without any objective signs of inflammation on physical examination [1]. After the first definition of growing pains by the French physician Duchamp [4], many attempts at better defining this condition have been made, and many researchers have tried to set specific diagnostic criteria [5].

To explain the etiopathogenesis of growing pains, many different theories have been purported [5], but none has ever been confirmed; “growing pains” continue to be diagnosed by exclusion of other factors, and only the symptoms are treated [1].

In the literature, several studies on adults with persistent, non-specific musculoskeletal pain reported that these patients had decreased 25-OH-D levels and that their painful condition improved with vitamin D supplementation [6–12]. It is known that vitamin D deficiency is common [13, 14] also among children [15–17], but few studies have investigated the prevalence of vitamin D deficiency specifically in children suffering from growing pains [18, 19]. Moreover, only one study conducted by Friedland et al. [20] has evaluated bone mineral status in these patients, observing a reduced bone speed of sound (SOS) compared with healthy controls.

Our study is the first that assessed a possible relationship between pain intensity, vitamin D levels, and bone mineral status in children affected by growing pains.

Materials and methods

Patients

We enrolled 33 children, 15 girls and 18 boys, affected by growing pains as defined by Petersen [5]. Their mean ages were 8.1 ± 3.2 years for the boys and 10.0 ± 2.3 years for the girls. Twenty-nine patients (87.9 %) had a BMI between the 5th and 95th sex-specific and age-specific percentiles [on the growth charts of the Center for Disease Control (CDC) and Prevention—http://www.cdc.gov/growthcharts]. Two (5.9 %) had a BMI above the 95th percentile, and 2 (5.9 %) had a BMI below the 5th percentile. The mean BMI value was 16.8 ± 2.8 SD in males and 18.6 ± 4.0 SD in females. Fourteen of the 18 boys (77.8 %) and 5 of the 14 girls (35.7 %) were at a pre-pubertal stage of sexual development (Table 1).

Table 1 Characteristics of the study population by gender (mean ± standard deviation)
Full size table

Inclusion criteria were children or adolescents affected by intermittent pains in both legs (non-articular in location), with no signs of inflammation such as local tenderness or swelling, or movement limitation. Exclusion criteria were history of known bone disease; physical signs (swelling, redness, trauma, reduced joint range, limping) and lab results (blood tests, imaging) of articular diseases; treatment with drugs that could increase bone metabolism (i.e., anticonvulsants, glucocorticoids, highly active anti-retroviral therapy, and antirejection medications); a medical condition that would decrease vitamin D; and cognitive impairments which would not enable the child to understand the questionnaire and the pain scale used.

Study design

Our study was designed as a prospective pilot cohort study. The patients first presented to our Pediatric Clinic between January 2010 and May 2010; in May 2010 we decided to end the children’s enrollment because the beginning of summer would have produced a bias in the interpretation of vitamin D levels. At first evaluation (T0) all 33 patients enrolled completed a questionnaire where the intensity of their growing pains was rated on a pain evaluation scale. At the same time serum 25-OH-D, PTH, and ALP levels were measured, and a quantitative ultrasound assessment (QUS) of the last four proximal phalanges of the non-dominant hand was performed. After the first 3 months of vitamin D supplementation (T1) we re-evaluated pain intensity and 25-OH-D, PTH, and ALP levels using the same methods used in the initial evaluation. After almost 24 months (T2), when children with levels of 25-OH-D <30 ng/ml continued taking vitamin D supplements, we reassessed pain intensity, lab tests, and QUS evaluations.

Informed written consent was obtained from the parents or guardians of all participating children, and our study followed the medical protocol established by the Declaration of Helsinki.

The questionnaire

Responses to the questionnaire were collected through semi-structured interviews (a combination of open and closed questions) derived from the literature. The items on the questionnaire regarded age, sex, ethnicity, height, weight, medical history, features of growing pains (pain experiences, frequency, management, and drugs used), family history of growing pains, physical activity, diet (in particular calcium and vitamin D intake), and co-existence of other orthopedic problems.

To assess pain intensity, the questionnaire included a Wong–Baker Faces Pain Rating Scale (WBFPRS), that is, a horizontal line of 6 hand-drawn faces, scored from 0 to 10, that ranged from a smiling face on the left to a crying face on the right (hardly noticeable pain versus strongest conceivable pain). We used this scale because it had adequate psychometric backing in the literature [21].

Laboratory tests

Serum vitamin D levels [measured as 25 hydroxyvitamin D, 25(OH)D] were measured using a chemiluminescent method (LIAISON®25-OH Vitamin D Total, DiaSorin, Saluggia, Italy); serum intact-PTH levels were determined by another chemiluminescent method (LIAISON® 1-84PTH, DiaSorin, Saluggia, Italy), while serum ALP levels were measured with a standardized colorimetric test (ALP2, Roche/Hitachi Cobas c.).

Vitamin D supplements

After the first vitamin D dosage, the patients were divided into 4 categories according to their serum 25-OH-D levels. In increasing order of severity, the 25-OH-D level groups were as follows: normal vitamin D levels (25-OH-D ≥30 ng/ml), vitamin D insufficiency (25-OH-D <30 ng/ml, but ≥20 ng/ml), vitamin D deficiency (25-OH-D <20 ng/ml) and severe vitamin D deficiency (25-OH-D <10 ng/ml). This classification of 25-OH-D levels was based on data obtained from previous studies [22–24].

During the first 3 months, the patients were given different amounts of cholecalciferol (vitamin D3), according to their 25-OH-D levels: the children with 25-OH-D <10 ng/ml were given 100,000 UI of intramuscular (IM) vitamin D3 once a month; the children with 25-OH-D ≥10 but <20 ng/ml were given 100,000 UI orally once a month; the children with 25-OH-D levels ≥20 ng/ml but ≤30 ng/ml took 25,000 UI orally once a month, and those with 25-OH-D ≥30 ng/ml were not treated but just kept under control.

Patients were informed of the consequences and infrequent side effects of vitamin D supplementation [24].

QUS

To assess bone mineral status we performed a QUS with a DMB Sonic 1200 device (IGEA, Carpi, Italy). We considered two parameters: bone transmission time (BTT), which is related to cortical thickness and amplitude-dependent speed of sound (AD-SOS), which better determines mineral density [25, 26].

At each session, the reference speed of the patient’s soft tissue was measured by applying the probes to the soft tissue area between the base of the thumb and the index finger. To measure bone density, probes were positioned at the distal metaphysis of the first phalanges in proximity of the condyles. The device automatically calculated the average AD-SOS of the 4 measurements, corrected for the presence of soft tissue and AD-SOS Z score.

The phalanx was used to measure bone strength because it is one of the most metabolically active parts of the skeleton, and in this region it is possible to observe small changes in bone turn-over, such as resorption of the trabeculae at epiphyseal and metaphyseal levels and enlargement of the medullary canal [27].

We repeated QUS assessment after 24 months, referring to Mauloni et al. [28], who, considering the sensitiveness of the methodology and the variations expected over time, calculated a minimal interval of 18 months between one measurement and the next to see QUS bone variations.

Statistical analyses

The statistical distributions of continuous variables were described using traditional summary statistics: mean ± standard deviation (Table 1).

Differences of mean levels for continuous variables (25-OH-D, PTH, pain intensity, AD-SOS, and BTT Z scores) were evaluated before and after vitamin D supplementation using a paired samples t test. The associations between continuous and discrete variables were assessed using the nonparametric median test. A p value <0.05 was considered statistically significant.

Statistical analysis was performed using Stata 11 (StataCorp. 2009. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP.).

Results

At the beginning of the study (T0), the children reported a mean pain intensity of 7.5 ± 1.6 SD, in particular 7.4 ± 1.8 SD for boys and 7.7 ± 1.4 SD for girls.

The mean values of 25-OH-D levels were 15.7 ± 6.9 mg/dl (16.1 ± 5.9 mg/dl for boys and 15.2 ± 8.1 mg/dl for girls), and the mean PTH values were 57.3 ± 27.3 pg/ml (50.3 ± 26.73 for boys and 66.9 ± 26.0 for girls) (Table 1). All 33 patients had a serum 25-OH-D level below 40 ng/ml. 31 patients (93.9 %) had 25-OH-D levels below 30 ng/ml, 8 (24.2 %) below 10 ng/ml, 19 (57.6 %) between 10 and 20 ng/ml, and 4 (12.1 %) between 20 and 30 ng/ml. Only 2 children (6.1 %) had 25-OH-D levels above 30 ng/ml, suggesting vitamin D sufficiency.

At the baseline, 20 of the 33 patients (60.6 %) had serum intact-PTH levels higher than 50 pg/ml, and 11 of these 33 (33.3 %) higher than 60 pg/ml. ALP levels were in the normal range for all the patients, with a mean value of 217.7 ± 101.9 UI/l.

Regarding the QUS assessment at T0, 16 of 26 children (61.5 %) and 19 of 26 (73.1 %) had AD-SOS Z scores and BTT Z scores below the normal mean value of 0, respectively. Six patients (23.1 %) had an AD-SOS Z-score between −1 and −2 SD, and 3 (11.5 %) had scores below −2 SD. Nine patients (34.6 %) had a BTT-Z score between −1 and −2 SD, and 2 (7.7 %) had scores below −2 SD. At T0, we observed that children with severe vitamin D deficiency had mean AD-SOS and BTT Z scores lower than mean scores of children with higher vitamin D values (Table 2). Moreover, patients with PTH levels below 50 pg/ml had less negative AD-SOS and BTT mean Z scores than those with PTH levels above 50 ng/ml (Table 3). However, these differences are not statistically significant.

Table 2 AD-SOS and BTT Z score mean values (±SD) at T0 in groups of patients according to 25-OH-D levels
Full size table
Table 3 AD-SOS Z score e BTT Z-score mean values (±SD) at T0 in groups of patients according to PTH levels
Full size table

At T1, after 3 months of vitamin D supplementation, 25-OH-D mean levels increased from 15.7 ± 6.9 to 34.1 ± 17.8 ng/ml (p < 0.001), and mean PTH levels decreased from 57.3 ± 27.3 to 47.3 ± 30.6 pg/ml (p = 0.135) (Figs. 1, 2). We did not observe any significant difference in ALP mean levels (213.4 ± 88.4 UI/l, p = 0.350). Mean pain intensity decreased significantly from a value of 7.5 ± 1.6 SD to a mean value of 2.7 ± 2.2 SD (p < 0.001), and in 8 cases (23.5 %) the episodes of pain disappeared altogether (Fig. 3). An association between the total amount of vitamin D3 taken in the 3 months of treatment and the resulting pain reduction was found (p = 0.047).

Fig. 1
figure 1

Vitamin D mean levels in ng/ml (with 95 % confidence intervals) at the baseline, after 3 months, and after 24 months

Full size image
Fig. 2
figure 2

PTH mean levels in pg/ml (with 95 % confidence intervals) at the baseline, after 3 months, and after 24 months

Full size image
Fig. 3
figure 3

Mean scores of the Wong–Baker Faces Pain Rating Scale (with 95 % confidence intervals) at the baseline, after 3 months, and after 24 months

Full size image

At T2, after 24 months, the patients maintained vitamin D mean levels higher than they were at T0 (p = 0.005), but lower than mean levels at T1 (p = 0.090). We observed a further decrease in PTH mean levels. The differences between PTH mean levels at T2 and T0 are statistically significant (p = 0.004). At the same time, patients experienced less intense pain than at T0 (p < 0.001). Moreover, we noted an improvement in the QUS parameters: AD-SOS Z scores grew from −0.53 ± 1.19 SD to −0.22 ± 1.06 SD (p = 0.538), and BTT Z scores improved from −0.75 ± 0.96 SD to −0.13 ± 0.90 SD (p = 0.014) (Fig. 4).

Fig. 4
figure 4

Values of QUS parameters (AD-SOS and BIT Z scores, means with 95 % confidence intervals) at the baseline, after 3 months, and after 24 months

Full size image

Discussion

Growing pains are the most common form of episodic childhood musculoskeletal pain, concerning both males and females [1].

In 1960 Brenning et al. [5] supposed that “growing pains” were caused by a higher growth velocity rate, and it is known that the most important growth spurt occurs during puberty; nevertheless we observed that our patients were mostly prepubertal; most children did not have a growth velocity greater than their peers without growing pains, and ALP values, which are an expression of osteoblastic activity and growth, were in the normal range in all our patients. Instead, we believe that other factors should be taken into account.

Considering a possible relationship between growing pains and vitamin D levels, we observed that 94 % of our children affected by growing pains had values of vitamin D below the normal range, confirming a previous Pakistani study [19]; the prevalence of hypovitaminosis D observed in these patients seemed to be greater than in healthy children, in which it is reported with a prevalence of 40 %-50 % [15, 18].

A recent study conducted by Friedland et al. [20] described 39 children affected by growing pains, especially in tibial regions, compared with healthy controls in which they observed a reduced bone SOS (hypothesizing a local overuse syndrome; but no studies have evaluated possible relationships between growing pain intensity, vitamin D levels, and QUS assessment). By correlating vitamin D levels and QUS parameters, we observed that patients with 25-OH-D levels below 10 ng/ml had AD-SOS Z scores and BTT-Z scores lower than children with higher vitamin D values. At the same time we observed an inverse relationship between PTH and AD-SOS and BTT Z scores.

Regarding pain intensity, we did not find a correlation between initial 25-OH-D values and the intensity of pain; but pain perception is subjective, and so it may not be helpful to make a comparison between different people and pain sensitiveness; whereas it was interesting to determine a change in pain after vitamin D supplementation in the same individual.

After 3 months of vitamin D supplementation we observed a significant increase in 25-OH-D values (p < 0.004) associated with a significant decrease in pain (p < 0.0001) and PTH levels. The PTH reduction was not statistically significant, probably because 3 months of vitamin D supplementation are not sufficient to normalize PTH levels. In fact, after 24 months of vitamin D supplementation the reduction in PTH levels became statistically significant, although vitamin D levels at T2 were lower than at T1, probably as a result of a diminishing compliance.

Along with the decrease in PTH levels we observed a clear improvement in QUS parameters, in particular in BTT Z scores (p < 0.01), which assess cortical bone density; it is probable that increases in vitamin D lead to reduced PTH-mediated cortical bone resorption.

The observed reduction of pain intensity could be attributed to more mineralized cortical bone since it is reported in the literature that skeletal pain could be caused by an altered structure of cortical bone. Specifically, in a state of vitamin D deficiency, through PTH action, osteoblasts continue to deposit a collagen rubbery matrix on both the endosteal and periosteal surfaces of the skeleton; this matrix expands under the periosteal covering and could cause an outward pressure on periosteal sensory pain fibers [12].

Our study is the first that has evaluated a relationship between pain intensity, vitamin D levels, and bone mineral status in young patients affected by growing pains.

In line with our results, we feel that it is helpful to consider vitamin D deficiency in children with musculoskeletal pains, and that it would be worthwhile to evaluate their bone status with QUS assessment. Further prospective and case–control studies will be necessary to confirm the role of vitamin D supplementation in reducing the incidence of so-called “growing pains”.