Statistics from Altmetric.com
Editor,—Children with Perthes' disease have aseptic necrosis of the hip and a selective impairment of growth. Trace metal deficiency in chickens leads to a similar condition, and in 1989 Hall et al reported low blood manganese levels in 27 children with this disease.1 With consent from the ethical committee of the same hospital as in their study, we have measured blood Mn of a further group of Perthes' patients, and control patients with minor musculoskeletal trauma matched for age and sex. Blood was drawn through stainless steel needles into plastic syringes and transferred into acid washed tubes containing EDTA. They were frozen upright at −20°C. Whole blood analysis was undertaken as reported previously1 and statistical analysis was by conditional logistic regression, adjusting for age, sex, and socioeconomic status.
A number of the frozen tubes cracked. Mean Mn levels in the tubes from control children were higher in the cracked tubes (197 (93) nmol/l) than in the uncracked tubes (157 (45) nmol/l; p = 0.02). We attribute the raised levels to dehydration or contamination and recommend freezing at an angle to obviate this complication. After exclusion of cracked tubes, results from 21 patients and matched controls remained. Mean blood Mn levels were 179 (77) and 157 (45) nmol/l respectively. Whole blood iron did not differ significantly in the two groups.
These findings do not support the earlier results from this hospital, although the entry criteria to the study and methods of analyses were similar. The difference between the studies may be due to chance, or to an inherent bias in one of them.
We also undertook a concurrent randomised intervention trial in 25 patients with Perthes' disease, all of whom took 3.4 g of flavoured maltodextrin daily, supplemented in 11 patients with 1 mg of manganese chloride. The outcome was determined over a two year period by regular assessment of symptoms, by sequential anthropometry, and measurements of femoral head morphology and subluxation. The patients did not like the oral supplement and there was no improvement in the measured parameters in those taking the active supplement. Their mean blood Mn levels at baseline and six months were 177 and 195 nmol/l respectively, whereas for those taking the inert powder the mean levels were 187 and 214 nmol/l. However, the fact that ingestion of 1 mg Mn daily had no effect on blood levels may not be surprising as this is about 25% of the normal daily intake, and manganese has a short biological half life. We found no benefit from supplementary Mn, but assessment was difficult due to the variable presentation and natural history of Perthes' disease.