Objective High-dose steroid therapy in children is known to impair growth. What is unknown is the level of steroid therapy at which children continue to grow normally. This study was designed to deduce a dosage of prednisolone compatible with normal growth.
Patients and design The growth of 41 children (age 1.92–13.2 years) with steroid-dependent nephrotic syndrome (SDNS) was studied using recordings from clinic visits over the course of their follow-up at Great Ormond Street Hospital (study period range 1.38–8.43 years, mean 4.2 years, total 172 years). The height standard deviation score (SDS) and the SDS velocity between clinics were calculated, and compared to the contemporary dose of prednisolone (converted to an equivalent daily dose when on an alternate day regime).
Results The mean dose of prednisolone was 0.44 mg/ kg/day (range 0.06–1.45 mg/kg/day). The mean change in height SDS velocity over the course of recording was −0.02 SDS/year (boys −0.14 SDS/year, girls +0.16 SDS/year). Overall, there was no negative effect on growth seen at doses of prednisolone of less than 0.75 mg/kg/day. At doses higher than 0.75 mg/kg/day, there was a small decline in height SDS velocity (−0.14 SDS/ year).
Conclusions Overall, prednisolone treatment in these children was not shown to adversely affect their height SDS. This was true even at doses of prednisolone up to 0.5–0.75 mg/kg/day. There was some decline in height SDS seen during periods of higher steroid use (over 0.75 mg/kg/day), but periods on lower doses allowed for adequate catch up growth.
Statistics from Altmetric.com
Steroids were introduced as a medical treatment over 65 years ago.1 Prednisolone was first used in the 1950s,2 and has since been used in a wide spectrum of diseases, becoming a cornerstone in the treatment of all inflammatory diseases and other conditions such as nephrotic syndrome.3 4 Despite the development of many new and effective drugs during the last 50 years, steroids still hold a place as first-line therapy in many of these diseases. It was, however, recognised early on that steroids produced a large number of unwanted side effects,5 which occur in virtually every patient treated with high doses. Lower doses give less pronounced side effects, but still have the potential to cause severe problems, especially if used for long durations.
In children, one very important side-effect is stunting of growth.6 7 No child will grow on high dose steroids. Clinical experience does, however, suggest that children can grow on lower doses, and there have been many attempts to analyse the effects of prednisolone on growth, particularly in nephrotic syndrome.8,–,16 These studies have largely concluded that high dose, long-term treatment is associated with stunting of growth, and catch-up growth is possible during reduced dose or steroid-free periods (with or without steroid-sparing agents), but none have been able to deduce a dosage that would consistently not cause growth retardation. A study in cystic fibrosis sufferers showed growth retardation after 24 months on a prednisolone dose of 1 mg/kg on alternate days, and after only 6 months on 2 mg/kg alternate days,17 but steroid treatment (for instance, for steroid-dependent nephrotic syndrome (SDNS)) is often of a much longer duration.
What is already known about this topic
High-dose, long-term steroid treatment impairs growth in children. Even inhaled steroids have previously been shown to slow growth.
However, the relationship between growth and steroids is complex. A safe level of steroid therapy, at which children continue to grow, is not known. Alternatively, cumulative dose of steroids may be more influential than daily dose, and catch-up growth while not on steroids is important to ensure normal adult heights, the relationship between growth and steroids is complex. A safe level of steroid therapy, at which children continue to grow, is not known. Alternatively, cumulative dose of steroids may be more influential than daily dose, and catch-up growth while not on steroids is important to ensure normal adult heights.
Some patients are treated with steroids for diseases that themselves are known to impair growth. A child with, for example, systemic lupus erythematosus, may be exhibiting slow growth velocity, due either to overtreatment with steroids, or undertreatment and heightened disease activity. The clinical picture may be otherwise one of remission, whereas biochemical markers of disease activity may be slightly raised. This leaves the clinician with a difficult decision of whether to increase or decrease treatment.
What this study adds
Children treated for steroid-dependent nephrotic syndrome do not exhibit an overall loss in height compared to their peers during the treatment of their condition.
Some of their growth is due to periods of catch-up when on lower doses of steroids.
Periods of treatment with up to 0.5–0.75 mg/ kg/day of prednisolone do not seem to have a negative impact on growth.
The aims of the present study were to evaluate the influence of long-term treatment with steroids on growth, and to calculate an average dose of steroids that consistently does not retard growth.
Patients and design
The nephrology database was searched for patients treated at Great Ormond Street Hospital with a diagnosis of SDNS from 1993 to 2004. Exclusion criteria were children with other chronic illness or congenital syndrome, and follow-up of less than 1 year. The notes of 41 children (25 boys) were subsequently retrospectively studied. Their height, weight and prednisolone dose at every clinic visit was recorded. Standard deviation scores (SDS) for height and weight were calculated according to the British Growth Reference 199018 using Growth Program 1.12, 2004 (Institute of Child Health) and the ΔSDS and rate of change (ΔSDS/time in years), between every clinic visit and over the whole study period was calculated. The data were also analysed using sequential clinic visits at intervals of at least 6 months, thereby producing a more accurate trend in growth and minimising the effect of inconsistency in measurement.
The children were treated with daily or alternate day prednisolone; to make a statistical comparison possible all doses were converted to the equivalent daily dose. Relapses were treated with high-dose prednisolone (60 mg/m2), which was then tapered according to clinical parameters. At times, some children were on a ‘steroid holiday’, while taking either levamisole or ciclosporin A as replacement therapy.
Data were analysed with SPSS V.13.0. Paired samples were analysed by t test, and where there was more than two groups, a one-way analysis of variance was performed.
The studied was registered and approved by the research ethics committee of The Institute of Child Health.
Overall, data from 41 subjects (25 boys) were collected. A total of 637 clinic visits (mean 15.5) over 172 patient years (mean 4.2) were studied. During the study period, there were 43 relapses (0.25 per patient year), and 21 periods of steroid sparing treatment (13 with levamisole, 6 with ciclosporin A and 2 with cyclophosphamide).
The mean change (Δ) in height SDS over the whole study period per child was −0.02 (range −2.05 to +1.26) (figure 1). This change was −0.14 (range −2.05 to +1.09) for boys and +0.16 (−0.91 to +1.26) for girls (non-significant difference between sexes: p=0.206). Eighteen children showed an overall decrease in height SDS (13 boys) and 23 showed an increase during the follow-up period (12 boys). The mean prednisolone dose per child throughout the study was 0.38 mg/kg/day (boys=0.39, girls=0.37).
We found no significant correlation between growth velocity and mean daily prednisolone dose between each of the 637 clinic visits (data not shown). The short interval between many of the clinic visits did, however, affect the accuracy of ΔSDS calculation between each visit and the results were therefore recalculated for the 229 clinic visits that were more than 6 months apart. This showed a weak correlation between change in height SDS and steroid dose (Pearson coefficient (R)=0.155, p=0.019). There was no correlation between ΔSDS for height over the whole study period and mean prednisolone dose (data not shown, R=0.208, p=0.192), nor was there a correlation between cumulative exposure to steroids and growth.
The 6-month data were then grouped by mean prednisolone dose in order to examine trends within prednisolone doses (figure 2): group 0 (no steroids, including periods of steroidsparing treatment); group 1 (up to 0.25 mg/kg/day); group 2 (0.25–0.5 mg/kg/day); group 3 (0.5–0.75 mg/kg/day); group 4 (0.75–1 mg/kg/day); and group 5 (>1 mg/kg/day). When taken together, groups 4 and 5—comprising all measurements on more than 0.75 mg/kg/day—showed a significant overall decrease in height SDS/year (−0.14, n=44) compared with those children on less steroid (+0.0039, n=185, p=0.038). During the time when a child was not taking prednisolone due to levamisole or ciclosporin A treatment he or she showed a mean increase in growth velocity of +0.018 SDS/year.
The age of onset at treatment (range 1.92–13.2 years) did not influence the ΔSDS for height per year (data not shown). Anecdotally it can be noted that the two oldest children showed most rapid decline in height SDS (both boys). Insufficient pubertal staging was available from the clinic notes to make valid statements on the cohort.
Overall, children treated for SDNS were growing well, although sustained doses of more than 0.75 mg/kg/day were associated with a small but significant decline in growth velocity.
The results of our study confirm our clinical impression that low to moderate doses of prednisolone—even when given for extended periods—do not negatively effect the long-term growth of most children. Previous studies have had contradictory results on this point. These differing results may be due to different ages in the study population: for instance, it has been suggested that prednisolone treatment can delay puberty.11 This can result in delayed growth, even if the child’s final height might not be affected. Indeed, previous long-term follow-up of children with nephrotic syndrome has shown that they do achieve normal heights as adults.19 We were unable to ascertain the role that potentially delayed puberty had on the growth of our cohort, as there was insufficient information in the clinic notes. We were able to show, however, that age at onset of treatment did not significantly influence growth.
Our data are reassuring in that our present treatment of steroid-resistant nephrotic syndrome does not cause stunting of the majority of treated children. The data cannot, however, be interpreted that indiscriminate use of steroids in children is not harmful to final height. High doses of daily steroids do without any doubt cause severe side-effects and there is always a need to try use as little steroids as possible. Even the small change of 0.14 SDS per year, will, if prolonged, cause a significant drop-off in height. For instance, an 8-year-old boy on the 50th centile for height will be 128 cm tall; 3 years of −0.14 ΔSDS/year will result in him being 141 cm, rather than 143 cm—the height of an 11-year-old on the 50th centile. In addition, our work did not aim to uncover other potential unwanted effects, which may also be controlled in a dosedependent fashion. In SDNS the use of steroid-sparing drugs such as ciclosporin A20 21 and levamisole22 23 to minimise the dose of steroids remains important. The use of these drugs may well have allowed significant catch-up growth in our sample, reflected by the overall increase in height when not taking prednisolone (figure 2). Catch-up growth has previously been seen in asthmatic children who had slower growth velocities while taking inhaled fluticasone,24 and it is likely that this phenomenon has a significant bearing on our findings. It has also been suggested that cumulative steroid dose is important in final height,15 16 which may be one component of the positive effect on growth of a ‘steroid holiday’.
It is also important to draw attention to the small number of patients who showed poor growth during the study. Even though the growth velocity of these children did not correlate with prednisolone dose (indeed, slow growth velocity was seen at all steroid doses), it would be foolish to assume that steroids were not having a negative effect. It is therefore essential to consider each patient individually, and note growth velocity on different doses for that particular child.
This group also show a low relapse rate for nephrotic syndrome.25 this may be in part due to the fact that these steroid-dependent children, who had been referred to a tertiary centre for troubling relapses, often had very frequent follow-up visits which allowed very careful weaning of their steroids, or implementation of steroid sparing agents. It may also be influenced by the children who were followed-up for shorter periods of time.
Although there is no guarantee that the side-effects of steroids are modulated in similar ways in different diseases, we feel our findings may have important implications for the long-term use of steroids in children with autoimmune or inflammatory diseases. Typical examples are systemic lupus erythematosus (SLE) and inflammatory bowel disease. These children often have slow growth, and it can be difficult to identify whether it is uncontrolled disease activity or a therapeutic side-effect that is the main cause. This confusion is compounded by the fact that biochemical markers of disease and clinical signs and symptoms may be inconclusive regarding disease activity which delivers a tricky clinical conundrum to the physician: whether to increase or decrease prednisolone dose (while being mindful of other side effects of steroid treatment). Our present data support the clinical impression that doses of steroids that have been thought to be linked to growth retardation may actually be safe from this respect, and in certain circumstances are undertreating children with poor growth. There may be many children with, for example, SLE, that are not growing as well as they should due to grumbling low to moderate grade disease rather than overtreatment. This is also supported by our observation that children with lupus with very slow growth respond very well to improved immunosuppressive treatment with, for example, rituximab.26
This study is limited by its retrospective nature, and by assumptions about patient compliance. We do not, however, think that our findings can be explained by non-compliance, which is common among nephrotic children.27 This is supported by the normal distribution of the ΔSDS. Since these children have been discharged from clinic before they have reached their final height, we were not able to make statements about adult height in this group, although the fact that—as a group—these patients did not exhibit an overall decrease in height SDS during the study must be seen as encouraging. Another potentially enlightening feature would have been calculation of a predicted final height from the heights of an individual’s parents. We were also unable to review sufficient radiographs to examine bone age in this cohort.
For the purposes of analysis, we had to convert alternate day dosing to an equivalent daily dose. Although our analyses showed no difference in effects on heights between alternate and equivalent daily dosing, there may well be a different sideeffect burden sustained between the two methods.
Although 41 patients is not a large sample compared to many growth studies, we consider that our study is strengthened by the specific question asked in one specific disease state, and by the large number of clinic visits recorded. We are also reliant on the measurements made in clinic by several observers over a long period of time, who were trained in paediatric measurement, but were not informed beforehand of the study. Again, with the number of measurements analysed, it is unlikely that could have skewed our results in a significant way.
In conclusion, long-term treatment with prednisolone in doses up to 0.25 mg/kg/day did not negatively effect growth in children, and allowed for periods of catch-up growth. Doses of between 0.25 and 0.75 mg/kg/day had little effect on growth, whereas doses of more than 0.75 mg/kg/day caused some slowing of growth velocity.
Funding This study was funded by the Department of Nephrology, Great Ormond Street Hospital.
Competing interests None.
Ethics approval The study was registered and approved by the research ethics committee of The Institute of Child Health.
Provenance and peer reviewed Not commissioned; externally peer reviewed.
The studied was registered and approved by the research ethics committee of The Institute of Child Health.
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.