Delayed puberty and poor growth often complicate the clinical course of children diagnosed with IBD, especially children diagnosed with Crohn disease (CD). As progression through puberty and increased growth velocity are intricately linked, most studies that look at the effects of IBD on puberty examine both growth and pubertal progress. Normal prepubertal growth velocity after 3 years of age averages about 5–6.5 cm/year. The pubertal growth spurt provides an additional 15–25 cm of growth [9, 12–14]. Delayed puberty is often associated with lower peak height velocity. Midparental target heights can be calculated with the following formulas: for boys, add 13 cm to the mother’s height and average it with the father’s height; for girls, subtract 13 cm from the father’s height and average it with the mother’s height. Most children will fall within ±8.5 cm of this prediction.

From the viewpoint of the authors, one of the most interesting studies which assessed the effect of IBD on puberty is that of Hildebrand et al. [23]. This study obtained height and weight data collected from birth through final adult height in 46 patients with childhood-onset Crohn disease (CD) and 60 patients with childhood-onset ulcerative colitis (UC) from a defined area in Sweden. In this study, the age at peak height velocity (PHV) was stated to represent the middle of puberty. Individual values for height were converted into standard deviation scores (SDS) using the infancy–childhood–puberty growth standard of Karlberg et al. [24]. The PHV for healthy children in Sweden was reported to be 12.05 ± 0.88 years for girls and 14.15 ± 0.98 for boys. Delayed puberty was defined as a delayed age at PHV of >2.0 SDS. No significant delay was noted in children with UC with age at PHV 11.9 ± 1.1 years for girls and 14.0 ± 1.2 years for boys. However, mean age at PHV was later in patients with CD: 12.7 ± 1.4 years for girls and 14.9 ± 1.2 years for boys, and 23% of these children with CD had a delayed age of PHV of >2.0 SDS.

Brain et al. also observed several alterations in the pattern of puberty among pediatric patients with IBD [11]. The mean age of onset of puberty was delayed for both female and male patients when compared to healthy controls: 12.6 years versus 11.1 years in girls and 13.2 years versus 12.4 years in boys. In addition, the duration of puberty was prolonged, especially in adolescents with frequent relapses during puberty [11]. Some patients with IBD took up to 4 years to progress from Tanner stage II to stage IV. Peak height velocities during puberty reached rates >12 cm/year in patients who remained in remission in contrast to as little as 1–2 cm/year in those with relapsing disease. When surgical resection was performed in 11 prepubertal children with CD, puberty started within 1 year of resection. The authors postulated that if the onset of puberty was delayed beyond 14 years, then the final height may be “irreparably compromised.” Our data would confirm that statement, as we observed that there was a strong correlation between age at menarche and height gain [25]. When menarche occurred at <13 years of age, the mean increment in height was 10 cm compared with only 3.0 cm in those aged >15 years. Homer et al. also noted that catch-up growth, even in prepubertal patients, occurred only in those with sustained clinical remission [26].

Ferguson and Sedgwick described delayed puberty in IBD based on a retrospective survey of adults with a history of pediatric-onset UC and CD [27]. Their results were different from other published reports in several ways. Adult stature achieved by 67 of 70 patients was similar to normal adults, and no difference was seen whether the patients had CD or UC. Delayed puberty was based on patients’ recall many years later. Pubertal delay was reported as follows: Crohn disease, 11/28 (39%) of men and 13/22 (59%) of women compared with 2/9 (22%) of men and 3/11 (27%) of women with UC. These numbers were not statistically different. Age at menarche was reported to be >16 years in 8 of 11 (73%) women whose menarche occurred after the diagnosis of CD.

More recently, Gupta et al. compared the age at menarche in 34 patients with CD with that for 545 controls, using data from the National Health and Nutrition Examination Survey (NHANES) [28]. They found that the median chronological age at menarche (13.9 years) in CD was older than that in the NHANES sample (12.0 years). In CD patients, the cumulative incidence of menarche was 10% at chronological age 12 years, 51% at chronological age 14 years, and 100% at chronological age 16 years. Sixty-eight percent reached menarche by bone age 13.5 years and 100% by bone age >14 years. Menarche occurred earliest in South Asians, followed by East Asians, and then Caucasians. They suggested if menarche has not occurred by bone age >14 years, endocrinology referral should be considered.

Other studies also show that onset of IBD during the prepubertal period is frequently associated with subnormal growth. In the Hildebrand et al. study, growth velocity in children diagnosed with IBD during the prepubertal period was −2.0 SDS in 24% of children with UC and 40% of children with CD [23]. Kanoff et al. and Kirschner also reported impaired growth in 68–88% of prepubertal children with CD [29, 30]. In addition, within a group of prepubertal patients, Saha et al. noted the poorest growth in those with severe CD when compared with UC [31]. In this study, no difference was seen between patients with and without corticosteroid treatment. In contrast, Motil et al. and Sentongo et al. reported that the prevalence of growth failure was equal regardless of the stage of pubertal development [32, 33]. Sawczenko et al. studied the effect of CD on final height in 123 patients who were designated “prepubertal” based on age at onset of symptoms: <13 years for boys and <11 years for girls [34]. Nineteen percent had a final height 8.0 cm or more below the targeted or midparental height. Those children who received steroid therapy were not found to be significantly shorter than other children at final height, suggesting that the judicious use of systemic steroids should not lead to significant long-term growth delay. However, of the shorter children in their study, boys were overrepresented with an OR of 3.70.

Several additional studies have suggested that boys with CD are more likely to have abnormalities of growth, especially while their disease is active. In an earlier report, Griffiths et al. had also observed less catch-up growth in boys than girls [35]. More recently, Gupta et al. reported that serum IGF-1 levels were reduced in males with IBD for both chronological and skeletal age when compared with female patients with IBD, and this may explain, at least in part, why male children with CD achieve less catch-up growth and have lower ultimate height Z-scores than females [36]. Gupta et al. subsequently reported that mean bone age Z-scores were lower in females, perhaps providing an opportunity for greater catch-up growth for females, once they are in remission [37]. To determine if some of the newer therapies in use for CD lead to improvements in growth and normal advancement in puberty, Pfefferkorn et al. analyzed growth outcomes in children with newly diagnosed CD [38]. They found that despite improvements in disease activity, mean height SDS scores did not change significantly, and pubertal progression remained slow. Children diagnosed with CD prior to 9 years of age had a higher mean growth velocity 2 years after diagnosis, as compared to children diagnosed after 9 years of age. Children who required prolonged corticosteroid therapy (longer than 6 months) had poorer growth outcomes. These data suggest that despite advances in nutritional and anti-inflammatory therapies for CD, growth and pubertal delays continue to persist in these children with CD.

In contrast, a study by Malik et al. suggested that children who had a clinical response to infliximab therapy had improvement in their linear growth that was independent of their pubertal progression [39]. In addition, children who had not been exposed to exogenous glucocorticoids also exhibited better growth with infliximab therapy, suggesting that the effect on growth was not simply related to a decrease in glucocorticoid use. In a more recent study, Mason et al. followed 63 adolescents with IBD (CD, n = 45 [23 males] and UC, n = 18 [12 males]), median age 13.4 years (range 10–16.6 years) over 12 months [40]. Interestingly, they reported no significant delay in puberty in their subjects. However, attenuation of the pubertal growth spurt was evident in these subjects. In the adolescents with IBD, the median IGF-1 SDS score was lower and IGFBP-3 higher than the control group, suggesting an abnormality of IGF-1 bioavailability.

Pubertal delay in IBD can have many etiologies, and poor nutritional status is often thought to be the major cause, as optimal nutrition is necessary for the initiation and maintenance of reproductive function. GnRH secretion is blunted in the malnourished state which leads to pubertal arrest, and secretion of GnRH normalizes with weight gain [41]. However, the delay of puberty in IBD presents a more complex issue, with weight not the sole independent variable. Stress and inflammation likely also have important roles. In addition to delays in the onset of puberty, slowing or cessation of sexual maturation may occur in patients with IBD. For example, secondary amenorrhea is a well-recognized complication of weight loss.

The complex interactions between severity of disease, fluctuations in inflammatory cytokines, and their effect on nutritional status and hormonal profile make it difficult to determine how individual factors influence the onset and progression of puberty in pediatric patients with IBD. As a consequence, while nutritional deficits are well described in patients, other aspects such as the potential role of inflammatory cytokines on puberty are often extrapolated from animal models [42].

In otherwise healthy children, undernutrition may cause a delay in sexual maturation and menarche. Important studies done by Frisch and colleagues demonstrated that the age of pubertal growth and menarche in girls correlated more closely to weight than to chronological age [43–45]. During the adolescent growth spurt prior to menarche, girls had a continuous decline in the percent body water and increase in body fat, resulting in a change in the ratio of lean body weight from 5:1 to 3:1 and a mean percent body fat at menarche of 22% [43–45]. The investigators noted that the mean weight at menarche in girls in the United States was 47.8 ± 0.5 kg [43–45]. A possible relationship between body fat and menarche was suggested by adipose tissue being a significant extragonadal site of estrogen production through conversion of androgen into estrogen. She postulated that the decrease in age at menarche (approximately 3–4 months each decade over the past 100 years) is due to girls reaching the “critical” weight earlier, secondary to improved nutrition. In girls with primary amenorrhea due to undernutrition, a minimal equivalent of 17% body fat may be necessary for menarche to occur [43–45]. For girls experiencing secondary amenorrhea, resumption of menses usually occurred when weight gain was 10% higher than the weight at menarche.

Dreizen et al. compared the age at menarche of 30 girls with “chronic undernutrition” with 30 “well-nourished” girls living in north central Alabama [46]. The average age at menarche was 14.5 years in the former group and 12.4 years in the latter group. Interestingly, standing heights that had differed by 9.2 cm at 12.5 years decreased to a difference of only 3.5 cm at 14.5 years and were not significantly different (1.1 cm) at 17 years. Similarly, skeletal age was delayed in the undernourished group, but at the time of menarche, the bone age in the undernourished girls was only 3.8 months more advanced than the well-nourished group. Complete fusion of the epiphyses was delayed in the malnourished group to 17.6 years versus 15.9 years for healthy controls. Therefore, although the timing of the adolescent growth spurt was delayed by undernutrition, final height (in the absence of underlying disease) was not significantly reduced. An earlier study by the same authors in undernourished boys also showed delayed epiphyseal fusion to 18.7 years versus 17.0 years and a mean difference in height between the groups of 2.68 inches at 16 years [47]. Unfortunately, final adult heights were not reported.

Similar delays in menarche (with onset averaging 15.1 ± 0.5 years) are seen in ballet dancers, swimmers, and runners whose training and low calorie intakes begin prior to menarche [44, 45]. Frisch postulated that these females have a raised lean/fat ratio. Both increased nutrition and reduction in the intensity of training may restore menses. Athletic amenorrhea is a hypothalamic reversion to a more immature pattern in GnRH response. Normalization may occur with reduction in exercise and/or other stress without the weight change estimated by Frisch.

Reduction in calorie intake has been documented in many studies of pediatric-onset IBD, especially CD [48–50]. Thus, undernutrition is likely to be one of the contributing factors leading to delay in the onset and progression of puberty. Similarly, secondary amenorrhea seen in female patients with IBD may be caused by weight loss, a frequent complication of IBD in adolescents.

Sentongo et al. used dual energy x-ray absorptiometry (DEXA) and anthropometric measures to compare fat mass (FM) and fat-free mass (FFM) in 132 pediatric patients with IBD and 66 healthy controls [33]. They found that patients had normal fat stores but reduced FFM, consistent with “inflammatory cachexia” [33]. They cited data suggesting that proinflammatory muscle-active cytokines may impair accretion of lean tissue.

Burnham et al. compared 104 North American patients with CD to 233 healthy control subjects and documented delayed sexual maturation in the CD group [51]. Patients within Tanner stages II–IV averaged 1.4–1.5 years older than control subjects at the same pubertal stages. Lean mass was reduced by 8% in the patient CD group. It is the opinion of the authors that the role of undernutrition in both growth failure and sexual maturation may be underestimated if these complications are compared only with documented weight loss. Failure to gain weight (without a history of weight loss) may also adversely affect the timing of menarche and the progression of puberty.

Advancement in puberty may also be related to excess weight gain [12, 52]. Early adrenarche appears to be related to excess weight gain and may be accompanied by skeletal advancement and possibly earlier true puberty. This may be related to peripheral aromatization of adrenal androgens to estrogens in fat.

Most studies of endocrine function in children and adolescents with IBD have been performed to investigate the causes of growth failure rather than the onset and progression of puberty [48–50, 53–57]. An intact growth hormone/insulin-like growth factor I (IGF-I) axis is necessary for normal postnatal growth. Thyroid hormone and cortisol are also important, as are the sex steroids at the time of puberty. IGF-I is produced in the liver under the stimulation of GH, and is thought to be the key mediator of the growth-promoting effects of GH. Reports in growth-impaired patients with IBD have generally demonstrated normal GH secretion, thyroid function, cortisol response to hypoglycemia, and gonadotropin response to GnRH. What changes were observed such as reduced amplitude of the GH pulse or increase in reverse triiodothyronine (rT3) were not associated with reduced growth velocity [55]. We observed that weight loss could be associated with prepubertal levels of circulating sex hormones despite previous physical signs of pubertal progression [56]. IGF-1 levels have been shown to be reduced in children and adolescents with IBD [42, 56, 58, 59]. This usually occurs despite the presence of adequate circulating levels of GH and is known as “growth hormone resistance.” Since IGF-1 is modulated by both GH and nutritional status, it is not clear whether the reduction of IGF-1 seen in this population is secondary to active disease or to the decrease in calorie intake (or both) [42, 56]. An increase in IGF-1 occurs following nutritional restitution in children with IBD. Some have suggested that the IGF-1 rise following enteral nutrition or surgical resection in children with active IBD precedes improvement in nutritional status (based on anthropometric measures); however, more rapid indices of nutritional restitution such as prealbumin were not measured in those studies. Corkins et al. noted the major binding protein for IGF-1 (IGFBP-3) was also reduced at diagnosis in children with IBD which would result in a reduced half-life for circulating IGF-1 [58]. Evidence using a knockout mouse lacking only liver-derived IGF-1 demonstrated normal growth and development, suggesting an important role for paracrine or autocrine production of IGF-1 by nonhepatic tissues [60]. The use of IGF-1 as a potential therapeutic agent to enhance growth in childhood IBD is hampered by concerns regarding a potential increased risk for colon cancer and other malignancies in this population [61].