Is ghrelin a biomarker of early-onset scoliosis in children with Prader–Willi syndrome?

Abstract

Background Adolescents with idiopathic scoliosis display high ghrelin levels. As hyperghrelinemia is found in patients with PWS and early-onset scoliosis (EOS) is highly prevalent in these patients, our aims were to explore (1) whether ghrelin levels differ between those with and without EOS and correlate with scoliosis severity, and (2) whether ghrelin levels in the first year of life are associated with the later development of EOS. Methods We used a case control study design for the first question and a longitudinal design for the second. Patients with PWS having plasma ghrelin measurements recorded between 2013 and 2018 in our database were selected and 30 children < 10 years old with EOS and 30 age- and BMI-matched controls without EOS were included. The Cobb angle at diagnosis was recorded. In addition, 37 infants with a ghrelin measurement in the first year of life were followed until 4 years of age and assessed for EOS. Total ghrelin (TG), acylated (AG) and unacylated ghrelin (UAG), and the AG/UAG ratio were analyzed. Results EOS children had an AG/UAG ratio statistically significantly lower than controls. The Cobb angle was positively correlated with TG and UAG. TG and AG in the first year of life were higher in infants who later develop EOS without reaching a statistically significant difference. Conclusions Our results suggest that ghrelin may play a role in the pathophysiology of EOS in PWS. Higher ghrelinemia in the first year of life required careful follow-up for EOS.

High Prevalence of Scoliosis in a Large Cohort of Patients with Prader-Willi Syndrome

“…Obesity, bone dysplasia and hypotonia of paravertebral muscles have been proposed as risk factors for the development of scoliosis [ 6 , 7 ]. In addition, a recent study has suggested that the elevated ghrelin levels may play a role in the pathophysiology of early-onset scoliosis in subjects with PWS [ 8 ]. The age of diagnosis of scoliosis is reported to follow a bimodal distribution pattern, with the first peak before the age of four and the second one in the adolescent period [ 9 , 10 ].…”