Statistics from Altmetric.com
Alemzadeh et al (
) reported that adult height, compared with the general Dutch population, was reduced by a mean of −0.9 SDS (95% confidence interval −1.55 to −0.28) in 15 Crohn’s patients with prepubertal onset of symptoms. However, the calculated deviation from “target” height (based on parental height) did not reach statistical significance and the authors have speculated that familial short stature, and not Crohn’s disease, may be a factor in this group. Furthermore, no height deficit was found in those with postpubertal onset of symptoms.
We are currently undertaking a review of those with childhood onset diseases attending our paediatric and adult IBD clinics. In the majority of cases parental height was measured by trained auxologists, although in some, details were not available in the case notes to discern the method of measurement and may therefore have included self reported parental heights. We calculated SDS scores from the revised British Longitudinal standards1 using the method described by Alemzadeh et al (mean British male adult height of 176.0 cm (SD 6.3) and female adult height of 163.6 (SD 5.7)). “Target height” was calculated for male patients by (paternal height+(maternal height+13))/2 and for female patients by (maternal height+(paternal height–13))/2 (cm). “Prepubertal” children were defined as males and females with onset of symptoms at <13 and <11 years, respectively. We defined the upper limit of “postpuberty” as 16 years, in contrast with Alemzadeh et al who used 22 years. The population we serve is ethnically diverse and therefore we have confined this analysis to Caucasians, as per the Dutch study.
There was no deficit in height of the parents (48 mothers and 46 fathers) of children with childhood onset Crohn’s disease compared with the general population. Furthermore, there was no deficit in height when examined by onset of their child’s Crohn’s disease: mean “prepubertal” parental SDS 0.00 (SD 1.11) and mean “postpubertal” parental SDS 0.08 (SD 1.12) (n=70 and n=24, respectively; p=0.34). In addition, we found no significant sex difference: mean paternal SDS −0.20 (SD 0.98) and mean maternal SDS 0.03 (SD 1.22) (n=46 and n=48, respectively; p=0.94).
In 27 cases (18 males, nine females) children of these parents are now aged more than 16 years (mean age 19.3 (SD 2.6) years). An analysis of final height is presented in table 1. A separate analysis was also carried out using the nomogram of the Child Growth Foundation2 which corrects height until the age of 22, but this did not alter our findings and the data are therefore not presented.
In the majority of patients (85% (23/27); p=0.008, χ2 test) final height was less than “target height”, and in 22% (6/27) the final height deficit was more than 10 cm. For those aged over 18 years the values were 88% ((14/16); p=0.022, χ2 test) and 25% (4/16), respectively.
In this sample of patients with childhood onset Crohn’s disease we found no evidence of a familial basis for short stature. Our data confirm the findings of others that mean adult height of patients with onset of symptoms before the age of 16 is reduced.3,4 Using age of onset of symptoms as a proxy for puberty, we found no significant difference in final height between those with pre- and postpubertal onset of symptoms. This is in contrast with the findings of Alemzadeh et al and may be because of differences in the upper age limit of “postpuberty” (16 versus 22 years).
Growth failure remains a concern to our British Crohn’s patients and although the mean deficit of 5–6 cm from target height may be considered by some to be clinically inconsequential this includes a subset with much more significant growth impairment. A better understanding of the mechanisms underlying growth failure is required to determine whether there is an identifiable group of children that may benefit from early and more intensive immunosupression and/or nutritional therapy.5,6
We agree with Alemzadeh et al that only larger (population based) studies will have the power to determine the effect of factors such as site of disease activity and therapeutic intervention.
I think that this study is clear. We agree with the authors that the differences are possibly caused by the other “end point” of puberty and the small population in both studies. Another difference could be that the authors did not calculate the corrected height SDS (height SDS target height SDS); this may be lower for the prepubertal group compared with the postpubertal group. Furthermore, they used another formula for target height. This formula does not include a correction for secular trend which will underestimate the deficit. Also, there is no information on the effect of corticosteroid use.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.