This was a hospital based, cross-sectional, case-control study, conducted at Department of Paediatric Medicine, Sir Padampat Mother and Child Health Institute (SPMCHI) attached to SMS Medical College, Jaipur, Rajasthan, India. Ethical clearance was taken from the Review and Research Board, SMS Medical College, Jaipur, Rajasthan, India.

The study group included 42 (28 males, 14 females) children of spastic quadriplegia (GMFC score 4 and 5) [11], and 42 (24 males, 18 females) age and sex matched healthy controls were taken. Both cases and controls were in the age group of 2 to 5 years and nutritional assessment was done using Indian Academy of Paediatrics (IAP) classification of malnutrition [12]. Children with metallic implants, family history of bone diseases or genetic disorders were excluded from the study.

The evaluation component included a detailed systemic and neurological examination, records of use of antiepileptic drugs, nutritional status was assessed by height, weight, weight/height and malnutrition grading according to IAP criteria [13]. IAP classification is based on weight for height values. The standard used in this classification for reference population was the 50^th centile of Harvard standards. Each patient and control completed a questionnaire, asking about their physical therapy status, daily calories and calcium intake, exposure to sunlight, a radiograph of the left hand and wrist, and a measurement of bone mineralisation in subjects. A detailed systemic and neurologic examination and measurement of bone mineralisation of controls was done. Informed consent was obtained from parents. Serum phosphorus, alkaline phosphatase, calcium (by calorimetric method), parathyroid hormone and vitamin D were measured (by two site immunoradiometric assay).

For each patient, bone maturation as bone age from the left hand and wrist radiographs were taken and assessed for bone age by using the sex specific standards of Greulich and Pyle [14] and compared with the controls.

Bone mineralisation in both the patient and control group was measured (total body mineral content) [15] with a dual energy x-ray absorptiometry (Hologic DQR-1000/w scan). The BMD (bone mineral density) is the bone mineral content (in grams) per unit area (cm²). Dual Energy X-ray Absorptiometry (DEXA) is the most widely used method for assessment of BMD and is considered the gold standard [15,16].

Statistical analysis was performed using EPI INFO software. Results were analysed for statistical significance by using student t-test and ANOVA test for continuous variable and chi-square test for discrete variables. We studied BMD z-score with other parameters by using Pearson’s correlation coefficient method. Multivariate analysis was performed to evaluate factors predictive of low BMD z-score. Variable with significant correlation were entered into stepwise linear regression model and model with greatest r-value was finally considered.

Most of the cases as well as controls were in the age group of 4-5 years followed by 2-3 years age group (mean age for cases was 4.2 and for controls was 4.1 years). [Table/Fig-1] summarized the laboratory parameters. Among the subjects, most of the males (10/28) were in Protein Energy Malnutrition (PEM) grade 2, and females (6/14) were in PEM Grade 3. Eight subjects were not malnourished, whereas 20 subjects were in PEM grade 1&2 and 14 in PEM grade 3&4.

BMD in study population (in both males and females) was lower than the controls (p<0.001). Our results showed that the mean BMD in male and female subjects was lowest in upper extremities. Higher values were found in lumbar spine, followed by pelvis and lower extremities. The total mean BMD was higher in male compared to female subjects, but this difference was statistically insignificant (p>0.05). BMD values of upper extremities were compared with lumbar spine, pelvis and lower extremities, and we found significant difference in values of upper limb and other region (p<0.05) [Table/Fig-2]. p-value was <0.001 and was very highly significant when the BMD values of male and female subjects was compared with sex matched controls using Z scores [Table/Fig-3].

[Table/Fig-4] showed BMD values of subjects taking more than one anticonvulsants compared with other subjects who were taking one anti convulsant or who were not on medication.

No statistically significant difference was found in total BMD values of male and female subjects taking anticonvulsants compared with subjects not on anticonvulsant (p-values for males 0.854 and for females 0.063) [Table/Fig-4].

BMD values of male subjects with regular physiotherapy compared with other subjects was not significant (p-0.448) while in females, it was statistically significant (p 0.002).

Bone age of male and female subjects with non severe malnutrition was delayed in 21 subjects out of 28 children (in 75 % cases) and with severe malnutrition, it was delayed in 14 subjects out of 14 children (in 100% cases).

For identification of predictors of BMD z score among spastic quadriplegic patients, step wise multivariate regression analysis [Table/Fig-5] was done considering BMD z score as dependent variable.

Age, sex, diet (k.cal/day), calcium intake, history of medication use, regular physiotherapy, S.calcium, S.phosphorus, S.alkaline phosphatase, S.parathyroid hormone, S. Vit.D, and PEM grade were entered as independent variables as predictors of low BMD in spastic quadriplegic [Table/Fig-5].

PEM grade showed significant criteria of inclusion as predictor, while rest of the other independent variables were excluded from the model.

The present study shows that spastic quadriplegic children have significantly lower values of S. Calcium, S. Alkaline phosphatase, while no significant difference in between S. Phosphorus, S. PTH and S. Vit D levels as compared to controls. BMD values were lower in cases and it was found that nutritional status independently contributes to lower BMD values, while physiotherapy and use of anticonvulsant drugs did not correlate significantly with BMD values. Bone age was lower than chronological age in most of the subjects.

In our study, we aimed to identify factors that may affect bone mineralization and to assess their association with BMD. Results of the present study demonstrates that decrease in BMD is significant in children with spastic quadriplegia. The suggested mechanisms responsible for the reduction in bone density in these children are poor nutritional status, insufficient calcium intake, immobilization and anticonvulsant use [17]. Reduced BMD values (measured by Dual-Energy X-ray Absorptiometry) in the spastic quadriplegic group as compared to the healthy controls in the study, further confirmed the presence of inadequate mineralization in children with cerebral palsy. King and colleagues also found that Lumbar spine BMD was markedly reduced in children with spastic quadriplegia [18]. In another series, Coppola G and colleagues confirmed that a severe mental retardation and spastic quadriplegia are significantly correlated to an abnormal bone mineral density with or without epilepsy [19].

Antiepileptic drug use may affect skeletal mineralisation adversely. Sato et al., and Tsukahara et al., found that long term antiepileptic treatment induces a state of decreased bone turn over and low bone mineral density [20,21]. In our study, mean BMD was lower in the subjects taking anticonvulsants but this difference was statistically insignificant (t-test 0.155 and 0.206 in male and female respectively, & Partial correlation = -0.230). Demet Yardimci et al., found no statistical correlation between the duration of anticonvulsant and bone mineral density and bone mineral content values in both sexes [16].

Poor feeding skills may lead to inadequate caloric, protein and calcium intake. In our study, mean calorie and calcium intake was significantly lower in study population as compared to controls (p 0.0001). 92.6% cases and 85.4% of controls in present study had subnormal vitamin D levels. Vitamin D deficiency has been associated with inadequate exposure to sun, vegetarian diet and absence of Vitamin D fortification in milk and food. A study comparing the outcomes of treatment with Vitamin D and bisphosphonate therapy on BMD values showed that bisphosphonate therapy was effective for patients who presented with secondary osteoporosis due to cerebral palsy [22]. Stallings et al., reported that malnutrition and growth failure were common in children with quadriplegic cerebral palsy [23]. Tandon et al., reported that a greater body mass index gain in childhood and adolescence is associated with higher peak bone density [24]. Another study by Henderson et al., showed that BMD z score was low in subjects with feeding problems [25]. Our data showed that, mean BMD z-score in severe malnutrition group was lower than non severe malnutrition group. Linear significance was 0.018 and significance of quadratic regression was 0.006.

Another important correlate of low BMD in children with CP is physiotherapy. Our data indicate that BMD values in the subjects with regular physiotherapy were higher compared to those without physiotherapy but this difference was statistically insignificant in males (Anova=0.448), while in females it was significant (Anova=0.002). {Correlation coefficient was 0.130 for physiotherapy and BMD}. The significant difference found in females was probably because of co-existing severe malnutrition. Bülent Ünay et al., found no significant difference in the bone mineral density values of children on regular physiotherapy or who were not on physiotherapy [26]. Many similar studies had been conducted to assess the relation of BMD with other parameters and various therapeutic interventions and had shown almost similar conclusion which suggest that BMD values were much lower in children with CP due to the various underlying problems mentioned in their studies [27–33] [Table/Fig-6].

Our data indicate that bone age was delayed in 35 (83.33%) subjects (75% in non severe malnutrition and 100% in severe malnutrition). Seven children had normal bone age compared to chronological age, all of them belonged to non severe malnutrition group. Our results shows that nutritional inadequacies is one of the important factors for skeletal maturation, other factors like negative neuropathic effect, severity, type of the cerebral palsy and low bone mineral density may shows additive effect. Study by Demet Yardimci et al., showed that skeletal maturation is frequently delayed in children with spastic cerebral palsy [16].

We did not follow up our patients to compare the results of various medical and physical interventions they received.

Skeletal maturation and mineralisation are significantly influenced by nutritional status in children with cerebral palsy and adequate steps should be taken to maintain appropriate nutrition in these children.

Abbreviations: BMD, bone mineral density; CP, cerebral palsy; DEXA, dual energy X ray absorpsiometry; GMFC, gross motor functional classification; IAP, Indian academy of paediatrics; PEM, protein energy malnutrition; PTH, parathyroid hormone; SD, standard deviation; SMS, sawai man singh.