A Study to Evaluate Clinical and Hormonal Profile in Short Statured Children of Age Between 6 to 12 Years at a Tertiary Care Centre

Full text

Text extracted from the article PDF for reading and search. Tables, figures and equations are in the PDF, which remains the authoritative version.

1. Introduction

Short stature — height below −2 standard deviations or the 3rd percentile for age and sex — is among the commonest reasons for referral to paediatric endocrine services. It is a clinical sign rather than a diagnosis, arising from normal growth variants such as familial short stature and constitutional growth delay as well as from endocrine, nutritional, systemic and genetic disease [1, 2]. The clinical task is to separate children who need only reassurance and monitoring from those with a treatable pathological cause, because growth hormone (GH) deficiency and hypothyroidism respond to replacement therapy and delayed recognition costs adult height [3, 4].

That separation is difficult in practice. Referral cohorts are already selected for severe height deficit, so degree of shortness is unlikely to remain informative once a child has crossed the referral threshold, and there are limited region-specific data on which routinely available parameter actually discriminates between etiological groups. We therefore studied consecutive short-statured children aged 6 to 12 years — a pre-pubertal window in which growth velocity should be steady and endocrine disease is most readily detectable — with the objectives of defining the etiological and hormonal profile of this population and determining which clinical, anthropometric, hormonal and radiological parameters distinguish endocrine from non-endocrine causes.

2. Materials and Methods

2.1. Design, setting and participants

This prospective observational study was conducted in the Department of Paediatrics, Saraswathi Institute of Medical Sciences, Hapur, over 18 months. Children aged 6–12 years with height-for-age below the 3rd percentile on WHO growth charts were enrolled consecutively from the outpatient department and wards. Children with systemic cardiovascular, renal or hepatobiliary disease, those with contractures or kyphoscoliosis preventing height measurement, and those already on treatment were excluded. Informed written parental consent was obtained. Sample size was calculated as n = Z²Pq/E² (Z = 1.96, P = 5.5%, q = 0.945, E = 5%), giving 83.16; adding 20% for non-response gave a target of 100.

2.2. Assessment and investigations

A structured proforma recorded demographic details, birth history, family history of short stature or pubertal delay, consanguinity, diet and chronic illness. Height was measured to the nearest 0.5 cm on a fixed height board and weight on a calibrated scale. The lower segment was measured from pubic symphysis to heel and the upper segment derived by subtraction; the upper-to-lower segment ratio classified stature as proportionate or disproportionate. Mid-parental height was calculated in every child and expressed as a standard deviation score (SDS), height velocity recorded in cm/year, and height, weight and BMI converted to SDS on CDC references. Pubertal status was assigned by Tanner staging. First-line investigations in all children were complete blood count with ESR, urine examination, serum creatinine, SGPT, thyroid function tests, serum IGF-1, serum cortisol and radiographic bone age (Greulich–Pyle). GH stimulation testing was performed for poor height velocity, low IGF-1 or severe short stature, with GH deficiency defined as peak GH below 10 ng/mL; karyotyping and skeletal survey were performed on clinical indication. Each child received a single final diagnosis and was grouped as normal variant, endocrine, nutritional, chronic/systemic or genetic.

2.3. Statistical analysis

Analyses were performed on patient-level data. Continuous variables are given as mean ± SD with median and range, categorical variables as counts and percentages. Groups were compared by one-way ANOVA with Levene's test for homogeneity of variance, Kruskal–Wallis testing as a distribution-free check and Tukey HSD for post-hoc comparison, with effect size as η². Attained and mid-parental height SDS were compared within each child by paired t-test and Wilcoxon signed-rank test. Two-group comparisons used independent-samples or Welch t-tests, and categorical associations the Pearson chi-square test with the Fisher– Freeman–Halton exact test where expected counts were below five. Correlations are Pearson r with Spearman ρ. Two-sided p < 0.05 was considered significant.

3. Results

3.1. Cohort characteristics and etiological distribution

One hundred children were enrolled with complete data. Mean age was 8.61 ± 1.54 years, 64 (64.0%) were male and all were pre-pubertal. Mean height SDS was −2.89 ± 0.38, with 38 (38.0%) below −3 SDS, and stature was proportionate in 90 (90.0%). Undernutrition was documented in 79 children but judged the primary cause in only 21, making it a coexisting rather than causative finding in 58 of the 79 (73.4%). Mean height velocity was 3.75 ± 0.79 cm/year, with 95 children (95.0%) growing at under 5 cm/year and 57 (57.0%) under 4 cm/year, confirming active growth faltering (Table 1).

Table 2: Etiological categories and final diagnoses (N = 100). Etiological category and diagnosis n % Normal variant 36 36.0 Constitutional growth delay 23 23.0 Familial short stature 13 13.0 Endocrine 28 28.0 Growth hormone deficiency 18 18.0 Hypothyroidism 6 6.0 Multiple pituitary hormone deficiency 2 2.0 Other endocrine 2 2.0 Nutritional 21 21.0 Chronic / systemic illness 9 9.0 Genetic 6 6.0 Fig 1: (A) Etiological categories in the whole cohort. (B) Diagnostic subtypes within the endocrine group.

3.2. Attained height relative to genetic target

Mean mid-parental height SDS was −1.26 ± 0.45 against a mean attained height SDS of −2.89 ± 0.38. The within-child deficit relative to genetic target averaged 1.62 SDS (95% CI 1.51–1.73; paired t = 28.99, df = 99, p < 0.001; Wilcoxon p < 0.001; Cohen's d = 2.90). Eighty-six children (86.0%) were more than 1 SDS below target and 32 (32.0%) more than 2 SDS below, confirming that the cohort was short relative to family background and not only relative to the population (Figure 3A). Mid-parental height SDS itself differed across etiological groups (F = 3.220, p = 0.016), being lowest in the normal variant group (−1.39 ± 0.44) and highest in the nutritional group (−0.99 ± 0.49).

3.3. Which parameters distinguish the etiological groups?

Table 3: Comparison of growth, hormonal and skeletal parameters across etiological categories (one-way ANOVA; df 4,95). On Tukey post-hoc testing of height velocity the endocrine group differed from every other group (all p ≤ 0.010). Parameter Normal variant (n = 36)

3.4. Hormonal evaluation and growth hormone testing

Mean IGF-1 SDS was −2.21 ± 0.64. Thyroid function was normal in 92 children, with subclinical hypothyroidism in 8, all of whom fell within the endocrine group (6 hypothyroidism, 2 multiple pituitary hormone deficiency). Serum cortisol was low in only 2 children, both with multiple pituitary hormone deficiency. GH stimulation testing was performed in 28 children (28.0%) for poor height velocity (12), low IGF-1 (9) or severe short stature (7), confirming GH deficiency in 18 — an overall yield of 64.3% (95% CI 44.1– 81.4%), similar across the three indications (66.7%, 55.6% and 71.4%). Peak GH clearly separated the outcome groups (5.10 ± 1.86 vs 12.05 ± 2.65 ng/mL; Welch t = 7.34, p < 0.001; Figure 3B), but IGF-1 SDS did not differ between GH-deficient children and normal responders (−2.29 ± 0.54 vs −2.13 ± 0.57; t = −0.76, df = 26, p = 0.452; Mann–Whitney p = 0.457) and correlated only weakly with peak GH (r = 0.220, p = 0.260). Across the cohort height SDS correlated with neither bone age delay (r = 0.119, p = 0.239) nor IGF-1 SDS (r = −0.036, p = 0.719); age correlated inversely with bone age delay (r = −0.295, p = 0.003).

Fig 3: (A) Within-child height deficit relative to mid-parental target; the dotted line marks the genetic target and the solid line the mean deficit. (B) Peak growth hormone by stimulation test result; the dotted line marks the 10 ng/mL threshold.

4. Discussion

The central finding is that the severity of a child's height deficit carried no information about its cause, whereas height velocity did. Height SDS was almost identical across all five etiological categories (group means −2.70 to −3.06, p = 0.449), and neither weight SDS, BMI SDS, IGF-1 SDS nor deficit from genetic target separated normal variants from pathological causes. Height velocity produced a large and specific effect (η² = 0.306) that post-hoc testing localised entirely to the endocrine group, which grew about 1 cm/year more slowly than every other group. This is physiologically coherent: once a referral threshold of the 3rd percentile has been applied, attained height is a cumulative measure already constrained by that selection criterion, whereas velocity reflects the growth process at the time of assessment and remains free to vary with active endocrine disease. The practical implication is that velocity, not degree of shortness, should drive the decision to investigate — and indeed poor height velocity was the commonest trigger for GH stimulation testing here, yielding a diagnosis in 8 of 12 children so tested. Because 95% of the cohort was already growing at under 5 cm/year, absolute cut-offs derived from healthy populations were of limited use; the discriminating signal lay in the relative reduction between groups. The second clinically consequential finding concerns IGF-1. Mean IGF-1 SDS was low across the whole cohort (−2.21 ± 0.64), including in children with normal variants, and it did not distinguish GH-deficient children from those with a normal stimulation response (p = 0.452), nor did it correlate meaningfully with peak GH (r = 0.220). Peak GH separated the two groups cleanly. IGF-1 is therefore useful for selecting children for dynamic testing but cannot substitute for it, and a low IGF-1 in isolation should not be read as evidence of GH deficiency — consistent with published guidance that biochemical confirmation requires stimulation testing [5,6]. Comparing attained with mid-parental height added a third dimension: children fell 1.62 SDS below their genetic target on average, with 86% more than 1 SDS below, so the cohort was genuinely short for its families rather than short in a familial context. Mid-parental height SDS was lowest in the normal variant group, as expected where familial short stature contributes, and highest in the nutritional group, where an environmental insult has displaced a child from comparatively normal genetic potential. Bone age delay behaved less consistently, reaching significance only on non-parametric testing; the inverse correlation between age and bone age delay (r = −0.295, p = 0.003) indicates that delay was proportionally greater in younger children and cautions against interpreting an absolute delay in years without reference to chronological age. The etiological distribution — 36% normal variants against 64% pathological causes, GH deficiency commonest among the latter — is broadly consistent with other Indian tertiary-centre series [7,8] and confirms that referral to such a centre substantially enriches for treatable disease. The gap between documented undernutrition (79%) and nutritional short stature as a primary diagnosis (21%) is a reminder that undernutrition frequently coexists with, and may be secondary to, another cause, and should not close the diagnostic process [9]. This study is limited by its single-centre, referral-based design and by small numbers in the chronic/systemic (n = 9) and genetic (n = 6) groups, which leaves those comparisons underpowered; the absence of a difference there should not be read as evidence of equivalence. Height velocity was recorded over a single interval rather than a prospectively standardised observation period, karyotyping was performed selectively in 10 children so genetic causes may be underestimated, and there was no follow-up relating the baseline profile to treatment response or adult height. Strengths are the prospective design, a uniform diagnostic protocol applied to every child, complete data capture, and analysis at the level of individual patient records.

5. Conclusion

Almost two-thirds of pre-pubertal children referred with height below the 3rd percentile had a pathological cause, and growth hormone deficiency was the commonest endocrine diagnosis. The degree of height deficit did not indicate the underlying cause, but height velocity did, and was selectively reduced in endocrine disease — making it the parameter that should determine which children are investigated further. A single IGF-1 measurement did not separate growth hormone deficient children from normal responders and cannot replace stimulation testing. Assessment of short stature in this age group should rest on serial height velocity and comparison with mid-parental target, with dynamic endocrine testing reserved for children in whom velocity is disproportionately reduced.