Mid-Childhood Bone Mass After Exposure to Repeat Doses of Antenatal Glucocorticoids: A Randomized Trial
BACKGROUND AND OBJECTIVE: Treatment of women at risk for preterm birth with repeat doses of glucocorticoids reduces neonatal morbidity, but could have adverse effects on skeletal development. We assessed whether exposure to repeat antenatal betamethasone alters bone mass in children whose mothers participated in the Australasian Collaborative Trial of Repeat Doses of Corticosteroids.
METHODS: Women were randomized to a single dose of betamethasone or placebo, ≥7 days after an initial course of glucocorticoids, repeated each week that they remained at risk for preterm birth at <32 weeks’ gestation. In this follow-up study, children underwent whole-body dual-energy radiograph absorptiometry at 6 to 8 years’ corrected age.
RESULTS: Of 212 eligible childhood survivors, 185 were studied (87%; 91 repeat betamethasone group; 94 placebo [single course] group). Children exposed to repeat antenatal betamethasone and those exposed to placebo had similar whole-body bone mineral content (median repeat betamethasone: 553 g, interquartile range: 442–712 g; placebo: 567 g, interquartile range: 447–750 g; geometric mean ratio: 0.99; 95% confidence interval: 0.94–1.03, P = .55) and bone area (median repeat betamethasone 832 cm2, interquartile range: 693–963 cm2; placebo: 822 cm2, interquartile range: 710–1020 cm2; geometric mean ratio: 0.99, 95% confidence interval: 0.92–1.07, P = .75).
CONCLUSIONS: Exposure to repeat doses of antenatal betamethasone compared with a single course of glucocorticoids does not alter bone mass in mid-childhood.
- ACTORDS —
- Australasian Collaborative Trial of Repeat Doses of Corticosteroids
- DXA —
- dual-energy radiograph absorptiometry
What’s Known on This Subject:
Repeat antenatal glucocorticoids for preterm birth reduce neonatal morbidity, but preclinical studies suggest that increased fetal glucocorticoid exposure may adversely affect skeletal development. There are no randomized trial data on the effects of repeat antenatal glucocorticoids on later bone mass.
What This Study Adds:
Exposure to repeat doses of antenatal betamethasone compared with a single course of antenatal glucocorticoids does not alter bone mass at 6 to 8 years’ corrected age and is unlikely to increase the risk of later osteoporosis in offspring.
Antenatal glucocorticoid therapy is a mainstay in the management of preterm birth, but major neonatal complications are prevented in only approximately half of infants exposed to a single course of antenatal glucocorticoids.1 Preclinical studies and randomized trials suggest that optimal fetal maturation requires repeat exposure to glucocorticoids.2 We showed in the Australasian Collaborative Trial of Repeat Doses of Corticosteroids (ACTORDS) that administration of repeat doses of betamethasone in women at risk for preterm birth at <32 weeks’ gestation, ≥7 days after an initial course of glucocorticoids, reduces the incidence and severity of neonatal lung disease, clinically significant patent ductus arteriosus, and combined serious infant morbidity.3 However, there is concern that increased fetal glucocorticoid exposure could adversely affect long-term growth and function of a number of tissues, including the skeleton. Although peak bone mass is not affected by exposure to a single course of antenatal glucocorticoids,4 there are currently no data from randomized trials on the long-term effects of repeat doses on skeletal development.
Animal studies have suggested that repeat doses of glucocorticoids may impair fetal calcium uptake5 and mineralization6 and alter postnatal appendicular growth,7–9 and the only available clinical studies reported decreased fetal bone turnover.10,11 Glucocorticoids influence the concentrations of several hormones involved in the regulation of musculoskeletal growth, including leptin,12–14 growth hormone,15 and the insulin-like growth factors,7,15,16 and may increase cortisol secretion into early adulthood,17 which could accelerate bone loss.18 Furthermore, there is growing evidence that the risk of osteoporosis in later life is increased by adverse in utero conditions, particularly fetal undernutrition and growth restriction,19 which in turn have been associated with fetal overexposure to glucocorticoids.
Therefore, we investigated if exposure to repeat doses of antenatal betamethasone permanently alters bone mass, including bone size and mineralization, by performing whole-body dual-energy radiograph absorptiometry (DXA) in New Zealand school-aged children whose mothers participated in the ACTORDS trial.
Study Population and Intervention
The ACTORDS trial (ISRCTN identifier 48656428) has been reported previously.3 Briefly, women with single, twin, or triplet pregnancy at <32 weeks’ gestation were recruited if they had received a course of antenatal glucocorticoids 7 or more days previously and were judged to have ongoing risk of preterm birth. A central telephone randomization service was used to assign women to an intramuscular injection of either Celestone Chronodose, containing 7.8 mg betamethasone sodium phosphate and 6 mg betamethasone acetate (Schering-Plough, Sydney, New South Wales, Australia), or a saline placebo. The allocated treatment was repeated each week that a woman remained undelivered at <32 weeks’ gestation if she was considered by her responsible clinician to have ongoing risk of preterm birth. All subjects, clinicians, and investigators were blinded to treatment allocation. A total of 982 women (1146 live fetuses) were enrolled across 23 participating centers in Australia and New Zealand between 1998 and 2004 (290 women and 352 fetuses were recruited in New Zealand). Infants exposed to repeat betamethasone compared with placebo had clinically significant reductions in the incidence of respiratory distress syndrome, severe neonatal lung disease, and combined serious non-respiratory neonatal morbidity.
The current study was conducted as part of a wider follow-up of the entire ACTORDS cohort at 6 to 8 years’ corrected age to assess longer-term effects of repeat antenatal betamethasone treatment on health and development. Cognitive function, behavior, body size, blood pressure, spirometry, health-related quality of life, and use of health services have been reported and were similar between groups.20 All children who were followed up in New Zealand were at the same time invited to participate in additional physiologic studies, including risk factors for cardiovascular disease21 and bone mass assessed by DXA.
To minimize instrumental variation, DXA was restricted to children who were assessed in the 2 main study centers, Auckland and Christchurch, where there was access to an identical machine (Lunar Prodigy; GE Healthcare, Madison, WI). A single technician at each site performed all scans according to a common protocol by using the pediatric module of Encore software (GE Healthcare). Images for the first 24 children were acquired by using version 8 (8.1 or 8.8), after which both centers upgraded to version 11.4, which was used for all data processing. The scan mode and depth were set automatically by the computer, depending on the child’s weight and height. Body regions were determined by using computer-generated and manually confirmed default lines on an anterior view planogram. Bone mineral content and area for whole body, head, and spine were recorded. A daily quality assurance check was performed by using the manufacturer’s block phantom, with stable calibration being maintained throughout the study period. Anthropometric measurements and clinical determination of pubertal status were made by a single investigator (C.J.D.M.), as previously described.21
Written informed consent was obtained from caregivers, and assent was sought from the children. Ethical approval was provided by the Ministry of Health, Wellington, and the National Radiation Laboratory, Christchurch, New Zealand.
The primary outcomes for assessment of bone status were whole-body bone mineral content and bone area, less the head component. Secondary outcomes included body segment proportions, spinal mineral content and area by using regional analysis from the whole-body scan, and fracture incidence by parental report.
Treatment groups were compared by mixed linear regression with SAS JMP software (version 8.0.2; SAS Institute, Inc, Cary, NC). Models were adjusted for the following factors, decided a priori: gestational age at trial entry; ethnicity (European versus non-European); preterm prelabor rupture of membranes and antepartum hemorrhage, due to a small imbalance between groups in the original trial cohort3; and clustering of fetuses in multiple pregnancy (random effect). To account for body size, additional regression models adjusted for height; analysis of bone mineral content was also separately adjusted for bone area and lean mass. Treatment effects are reported as mean difference for normally distributed data or ratio of geometric means for positively skewed data, both with a 95% confidence interval. A 2-tailed P value <.05 was considered statistically significant.
A total of 328 children from the ACTORDS trial were presumed to be alive and residing in New Zealand at 6 to 8 years’ corrected age, of whom 308 (94%) were recruited to the main follow-up study (Fig 1). Of the 212 children who were assessed in the 2 centers with DXA facilities, 185 (87%) successfully completed whole-body scans (repeat betamethasone group n = 91, placebo group n = 94).
The mean gestational age at birth in both groups was 31 weeks, with 93% of children born preterm (Table 1). Women received a median of 2 doses of study drug, in addition to the prerandomization course of antenatal glucocorticoids (Table 1). There were no significant differences between groups in sex, weight, length, and head circumference at birth or primary hospital discharge, nor in the corresponding sex- and gestation-specific z scores (birth weight data shown in Table 1). The effect of repeat antenatal betamethasone treatment on primary neonatal respiratory outcomes was similar to that seen in the main trial (Table 1). Mean age (SD) at follow-up was similar between groups (repeat betamethasone: 7.2 [0.9] years, placebo: 7.2 [1.0], P = .99) as was BMI (repeat betamethasone: 16.5 [3.0], placebo: 16.6 [2.7], P = .81).
At early school age, children exposed to repeat antenatal betamethasone and those exposed to placebo had similar whole-body bone mineral content and bone area, including when adjusted for height (Table 2). Bone mineral content also did not differ between groups after adjustment for bone area or lean mass. There were no differences between treatment groups in other secondary outcomes, including spinal bone mineral content and area and body segment proportions (Table 2). The incidence of fractures was similar, occurring in 12 (13%) children exposed to repeat antenatal betamethasone and 10 (11%) children exposed to placebo (P = .65). Most fractures occurred in the upper limbs (11 in the repeat betamethasone group, 7 in the placebo group).
There was a nonsignificant imbalance between groups in ethnicity and rates of maternal preeclampsia (Table 1). However, adjustment for these factors did not alter the results.
Seven children in the repeat betamethasone group had medical conditions that could potentially affect bone mass, including 2 with celiac disease, 2 with cerebral palsy, 1 with lymphoblastic lymphoma, and 1 with congenital multicystic kidney disease requiring renal transplant. In the placebo group, 2 children had short stature, 2 had cerebral palsy, and 1 had steroid-resistant nephrotic syndrome. Three children in the repeat betamethasone group had entered puberty, but were only Tanner stage 2. In a sensitivity analysis, exclusion of these children did not alter the results.
We found no evidence that exposure to repeat doses of antenatal betamethasone compared with a single course of antenatal glucocorticoids adversely affects bone mass in children at 6 to 8 years’ corrected age. These findings suggest that treatment with repeat doses of antenatal glucocorticoids, as used in the ACTORDS trial, does not alter the trajectory of bone growth and is unlikely to affect peak bone mass and density at skeletal maturity.
Bone mass is a product of bone size, the amount of bone tissue contained within the periosteal envelope (cortex and trabecular plates), and the mineral density of that tissue. These elements, along with the structural organization of the organic matrix and the spatial geometry of bone tissue, determine bone strength.22 We used whole-body DXA to assess 2 aspects of total skeletal mass, namely, bone mineral content and bone size (anteroposterior bone area). These measures, when normalized for height, have been shown to be good predictors of compartmental mineral density and bone cross-sectional area, respectively, and both are correlated with mechanical strength.23 In contrast, areal density, which is commonly used to assess bone mass in adults, is susceptible to size-related artifacts during growth,24 and the biomechanical significance of bone mineral content relative to bone area across the whole skeleton is uncertain.23
Adequate accrual of bone mass is important not only for the prevention of childhood fractures,25 but also for the attainment of peak bone mass at the end of skeletal growth, which is a major determinant of the risk of osteoporosis and fragility fractures in later life.26 Although there is considerable variation in attained peak bone mass, in a given individual, traits such as bone size and mineral content show high tracking stability from mid-childhood through to the end of adolescence, when skeletal consolidation is largely completed.27,28 Therefore, given our findings that, at early school age, point estimates for the effect of treatment on whole-body bone mineral content and bone area were close to 1 and confidence limits were relatively narrow, it seems unlikely that exposure to repeat doses of antenatal betamethasone compared with a single course influences long-term accrual of bone mass. The similar incidence of fractures between groups also supports this conclusion.
Randomized trial data on the long-term effects of antenatal glucocorticoids on bone mass are limited to 1 other study, the Auckland Steroid Trial, in which a subgroup of 174 subjects underwent DXA at a mean age of 31 years. There were no differences between those exposed to a single course of antenatal betamethasone and those exposed to placebo in whole-body, lumbar, or femoral bone mass, and femoral geometry.4 Importantly, this trial tested 2 different doses of glucocorticoids and there was no evidence of a dose-response effect on peak bone mass. Similarly, a small observational study found that tibial bone mass in infants, as assessed by ultrasound, was unaffected by single or repeat antenatal glucocorticoid treatment.29
Despite these reassuring findings, we were concerned that exposure to repeat doses of antenatal glucocorticoids could adversely affect skeletal development, as has been suggested in several animal studies. Lambs exposed to weekly compared with a single dose of betamethasone had decreased plasma calcium concentrations, possibly due to impaired placental mineral transfer,5 which in humans has been associated with long-term reductions in bone mass.30 In rat pups, exposure to repeat doses of antenatal glucocorticoids reduced bone calcium content, especially of the axial skeleton.8 In another rodent study, repeat doses of dexamethasone did not alter volumetric or compartmental bone mineral density in mature offspring, but females had altered femoral geometry with reduced cortical thickness.9
This apparent discrepancy between animal data and the results of our study may be due to the use of relatively higher doses of glucocorticoids in small animals or interspecies differences in skeletal development. Although 1 human trial demonstrated suppression of cord blood indices of bone turnover in infants exposed to repeat compared with single courses of betamethasone,6 we found no adverse effects on actual bone mass in childhood, highlighting the importance of studying outcomes of long-term clinical relevance.
One limitation of using whole-body DXA to assess total skeletal mass is that results primarily reflect cortical rather than trabecular bone, because cortical bone accounts for ∼80% of total skeletal mass,23 even though tracking of both bone types is similar.28 Therefore, we also compared groups for spinal bone mass using regional vertebral analysis of the whole-body scan. Although this may be less precise than dedicated lumbar anteroposterior measurements,31 we found no evidence of altered mineralization of spinal trabecular bone, a site commonly affected in osteoporosis.22 Another limitation of our study was that radiologic evidence of fractures was not obtained, although large cohort studies have shown that parental recall of children’s fractures is reliable.25 In addition, all children in this cohort were exposed to a single course of antenatal glucocorticoids, so it was not possible in this study to compare outcomes with unexposed children. However, long-term follow-up of subjects from the Auckland Steroid Trial has clearly shown that a single course of betamethasone does not affect peak bone mass.4
In the Auckland Steroid Trial, adults exposed to a single course of antenatal betamethasone compared with those exposed to placebo had a greater proportion of stature in the lower body segment, even though height did not differ between groups.4 Altered appendicular growth has also been shown in animals exposed to antenatal glucocorticoids, although results have been conflicting with reports of both increased9,10 and decreased11 long bone length. One possible explanation for these findings is a slight alteration in the timing of puberty, because appendicular growth is more rapid than axial growth before puberty and decelerates at puberty when axial growth accelerates. Indeed, in 1 randomized trial, there was a trend for boys exposed to a single course of antenatal betamethasone to have slightly later pubarche compared with those exposed to placebo.32 We found that exposure to repeat doses of antenatal betamethasone did not affect body segment proportions, although our study was designed to assess children before pubarche. Future follow-up studies should assess the effect of antenatal glucocorticoids on the timing and duration of puberty.
Osteoporosis is one of a number of adult conditions, including diabetes, hypertension, stroke, and ischemic heart disease, in which the incidence is increased in individuals exposed to adverse early developmental conditions. Although variation in osteoporosis-related traits, such as peak bone mass and rate of bone loss, has an important genetic basis,33 there is increasing evidence that fetal nutrition and growth also influence long-term bone health. In several animal studies, both maternal total nutrient34 and isolated protein restriction35 decreased fetal bone mass accrual and ossification, with reductions in bone size, mineral content, and strength persisting into adulthood.36,37 Likewise, in humans, lower birth weight and impaired fetal growth have been associated with long-term decreases in bone mass,19 especially bone size,4 possibly due to altered activity of the somatotrophic38 and hypothalamic-pituitary-adrenal axes.18,39
Exposure to excess glucocorticoids in utero has been proposed as a key mechanism underlying the association between decreased fetal growth and health outcomes. Animal studies have shown that pharmacologic manipulations that increase placental transfer of maternal glucocorticoids result in decreased fetal growth and cardiovascular and metabolic dysfunction in offspring,40,41 whereas the adverse cardiovascular effects of maternal undernutrition can be prevented by the blockade of maternal glucocorticoid synthesis.42 However, due to the limited capacity for experimental investigation, the extent to which glucocorticoids contribute to developmental programming of disease in humans is unclear.43 Thus, clinical trials of glucocorticoid treatment provide a unique opportunity in which to test the hypothesis that developmental programming effects are mediated by fetal glucocorticoid exposure. In our cohort, approximately half of the children in the intervention arm were exposed to elevated glucocorticoid levels for at least 2 to 3 weeks. Although exposure to repeat antenatal betamethasone did not affect birth size in this cohort, nonhuman primate studies have shown that the effect of antenatal glucocorticoids on metabolic function is not dependent on effects on fetal growth.11 The fact that children in the repeat group showed no evidence of altered bone mass suggests that either glucocorticoids have a limited role in the programming of osteoporosis in humans or that a much longer period of exposure or higher doses are required.
Exposure to repeat doses of antenatal betamethasone compared with a single course of antenatal glucocorticoids does not alter bone mass at early school age. Given the high degree of tracking of bone mass from mid-childhood, it is likely that treatment groups will attain similar peak bone mass. Therefore, clinicians who wish to use this treatment because of the established short-term neonatal benefits can be reassured that administration of repeat doses of antenatal glucocorticoids to women at risk for preterm birth is unlikely to increase the risk of osteoporosis and fragility fractures in offspring.
We thank the mothers and children who participated in the ACTORDS trial and this follow-up study. We also thank Ann Mansfield from St Georges Radiology, Christchurch Radiology Group, for performing DXA scans on children in Christchurch and Coila Bevan for assistance with tracing and recruitment of subjects. We thank the ACTORDS study group for their support, including the coordinating committee (Caroline Crowther, Ross Haslam, Lex Doyle, Peter Anderson, Janet Hiller, Jane Harding, and Jeffrey Robinson); Pat Ashwood, ACTORDS clinical trial coordinator; and Kristyn Willson, trial statistician.
- Accepted February 10, 2017.
- Address correspondence to Jane E. Harding, DPhil, Liggins Institute, University of Auckland, Private Bag 92019, Victoria St West, Auckland 1142, New Zealand. E-mail:
This trial has been registered with the Australian New Zealand Clinical Trials Registry (www.anzctr.org.au/) (identifier ACTRN12606000318583).
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: This follow-up study was supported through project grants from the Australian National Health and Medical Research Council (NHMRC:453633), the New Zealand Health Research Council (07/204), and the Auckland Medical Research Foundation (81537).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Crowther CA,
- Haslam RR,
- Hiller JE,
- Doyle LW,
- Robinson JS; Australasian Collaborative Trial of Repeat Doses of Steroids (ACTORDS) Study Group
- Swolin-Eide D,
- Dahlgren J,
- Nilsson C,
- Albertsson Wikland K,
- Holmäng A,
- Ohlsson C
- Hauser J,
- Knapman A,
- Zürcher NR, et al
- Fonseca L,
- Ramin SM,
- Mele L, et al; Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Maternal Fetal Medicine Units Network (MFMU)
- Kajantie E,
- Hytinantti T,
- Koistinen R, et al
- Sugden MC,
- Langdown ML,
- Munns MJ,
- Holness MJ
- Crowther CA,
- Anderson PJ,
- McKinlay CJ, et al; ACTORDS Follow-up Group
- Boyanov M
- Smolders-de Haas H,
- Neuvel J,
- Schmand B,
- Treffers PE,
- Koppe JG,
- Hoeks J
- Lindsay RS,
- Lindsay RM,
- Edwards CR,
- Seckl JR
- Bloomfield FH,
- Harding JE
- Copyright © 2017 by the American Academy of Pediatrics