RESEARCH ARTICLE

Poor Catch-up Growth in Late Adolescent Boys with EatingDisorders, Weight Loss and Stunting of GrowthIngemar Swenne*

Department of Women’s and Children’s Health, Uppsala University, Uppsala, Sweden

Abstract

Objective: The study aims to investigate the catch-up growth of boys presenting with an eating disorder (ED) and a stunting of growth.Method: Weight gain and growth of 46 boys with ED were followed up for 1–3 years.Results: A total of 13 boys who had not started their pubertal growth spurt at presentation started catch-up growth immediately follow-ing nutritional rehabilitation and weight gain. After 3 years, they had returned to their premorbid growth curve. Thirty-three boys whohad started their pubertal growth spurt prior to presentation never caught up in height but continued to catch down despite weight gain.After 3 years, they had lost 0.64� 0.55 height standard deviation scores corresponding to approximately 4.5 cm of potential height.Conclusions: In prepubertal boys with EDs, catch-up growth is possible. Pubertal boys are at a disadvantage in that catch-up growth maynot occur despite weight gain. Copyright © 2013 John Wiley & Sons, Ltd and Eating Disorders Association.

Keywords

eating disorder; boys; linear growth; catch-up growth

*Correspondence

Ingemar Swenne, Department of Women’s and Children’s Health, Uppsala University, S-75185 Uppsala, Sweden. Tel: +46 18 6115892; Fax: +46 18 6115853.

Email: ingemar.swenne@akademiska.se

Published online 4 June 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/erv.2237

Introduction

Growth and pubertal development require satisfactory nutrition.Malnutrition and/or intervening diseases impede growth and devel-opment. With treatment of disease and improvement of nutritionalstatus, these effects are reversed. Catch-up growth can then beachieved, and previous growth channels are recovered (Prader,1978; Tanner, 1981; Boersma & Wit, 1997). However, if skeletalmaturation is advanced, linear growth may be limited, and it maynot be possible to reach potential adult height (Tanner, 1981).

In eating disorders (EDs), there is weight loss due to undernu-trition usually in the absence of other diseases. In most femaleadolescents with ED, weight loss starts at a postmenarcheal agewhen growth is almost concluded and final height almost reached.On the other hand, in female patients who have not completedpuberty and reached menarche, there may be significant stuntingof growth and delay of pubertal development at presentation(Root & Powers, 1983; Russel, 1985; Pfeiffer, Lucas, & Ilstrup,1986; Gowers, Crisp, Joughin, & Bhat, 1991; Danziger, Mukamel,Zeharia, Dinari, & Mimouni, 1994; de Monléon et al., 1998;Modan-Moses et al., 2003; Swenne & Thurfjell, 2003; Abadie, deTournemire, & Alvin, 2003). Stunting of growth has also beenobserved in small numbers of male adolescents with ED (Root& Powers, 1983; Pfeiffer et al., 1986; Danziger et al., 1994;Modan-Moses et al., 2003) and may in these be an importantpresenting feature (Modan-Moses et al., 2003).

Adolescents who present with ED and stunting of growth maybe at risk of never reaching their potential adult height (Russel,1985; Lantzouni, Frank, Golden, & Shenker, 2002). However,

Eur. Eat. Disorders Rev. 21 (2013) 395–398 © 2013 John Wiley & Sons, Ltd and Eating Disorder

during nutritional rehabilitation, adolescent girls resume growthand pubertal development (Swenne, 2005). They can recoverprevious growth channels and reach their potential adult heightprovided that nutrition is adequate and uninterrupted (Pfeifferet al., 1986; Swenne, 2005; Rosé et al., 2007; Prabhakaran et al.,2008). Because of the later onset of male puberty, boys have manyremaining years of potential growth at the adolescent onset of anED. They would therefore be at a greater risk of significantstunting if they do not recover weight. On the other hand, ifweight gain is achieved, it could be argued that they have a longerperiod available to catch up in stature. Because of the relativescarcity of boys with ED, this has not been investigated in asizeable sample. We have now analysed the catch-up growth ofa group of adolescent boys with ED who presented with weightloss and retardation of growth.

Methods

Patients

Case records of patients referred to Uppsala University Children’sHospital for medical assessment of a diagnosed or suspected EDwere reviewed. During the period 1997–2011, 52 adolescent boyswith an ED had been assessed. Included in the present analysis are46 boys born in 1982–1998 who had been followed up for at least1 year, five of these were still in treatment. Copies of growth chartsfrom the school health services spanning age 6–18 years wereavailable for all. On the basis of the history obtained at presenta-tion, the diagnoses by the Diagnostic and Statistical Manual ofMental Disorders, fourth edition (American Psychiatric

395s Association.

Growth of Boys with Eating Disorders I. Swenne

Association, 1994) criteria were either anorexia nervosa or an EDnot otherwise specified of the restrictive subtype. No patientreported vomiting or binge eating at presentation, and none even-tually received a diagnosis of bulimia nervosa. Somatic comorbid-ity included inflammatory bowel disease (1), transplantation ofteeth (1), hypothyreosis (1) and Klinefelter’s syndrome (1). Ninepatients had psychiatric comorbidity, which included attentiondeficit-hyperactivity disorder (6), obsessive–compulsive disorder(2), Aspergers’s syndrome (1) and Tourette’s syndrome (1).

Analysis of weight change and growth

For all patients, there were measurements of weight and stature ata maximal recorded weight, at presentation and during the courseof treatment. Prepubertal measurements of weight and staturewere available from the growth charts of the school healthservices. When there were several measurements between ages 6and 10 years, the earliest was used. In most cases, this was ameasurement obtained at a general health examination duringthe first year in school. Weights not recorded in case files or ongrowth charts or weights obtained at home were not used in thecalculations. Body mass index (BMI) was calculated as weight/height2 (kg/m2) for all observations of weight and stature.Measures of weight, stature and BMI were recalculated into stan-dard deviation scores (SDS) (Lindgren, Strandell, Cole, Healy, &Tanner, 1995). The growth curves were also used to determinewhether the boys at presentation grew along their prepubertaltrajectory or whether the growth curve indicated that the pubertalgrowth spurt had started.

Measurements of weight and stature were used to calculate therates of change for the period from the maximal recorded weightto the first assessment and for 1-year periods following the start oftreatment. Weight velocity denotes weight change (kg/year),growth velocity denotes growth in stature (cm/year) and catch-up growth denotes change in height SDS (SDS/year).

Statistics

Values are given as means� SD if not otherwise indicated. Differ-ences in anthropometric indices between patients and generalpopulation were analysed by Student’s two-tailed t-test for inde-pendent samples. Changes during treatment were analysed by

Table 1 Anthropometric measures during the course of an eating disorder in adole

No growth spurt at presentation

Prepubertal At top weight At presentation 3-year follow

n 13 13 13 10

Age (years) 7.5� 1.2 12.5� 1.3 13.4� 1.0 16.2� 1

Weight (kg) 28.6� 7.4 47.5� 9.9 41.5� 6.7 61.8� 9

Height (cm) 129� 9 155� 8 158� 8 175� 8

BMI (kg/m2) 17.1� 2.2 19.7� 2.9 16.2� 2.0 20.1� 2.

Weight (SDS) 0.74� 1.07 0.51� 1.15 �0.81� 0.94 �0.01� 0

Height (SDS) 0.22� 0.94 0.11� 0.99 �0.17� 0.95 �0.03� 0

BMI (SDS) 0.73� 1.17 0.63� 1.20 �1.18� 1.33 �0.16� 1

Values are means� standard deviation.

BMI, body mass index; SDS, standard deviation score.

396 Eur. Eat. Disorders

paired t-tests. To predict growth in stature and catch-up growth,a multiple stepwise linear regression analysis was used.

The investigation was reviewed and approved by an institu-tional review board.

Results

Before puberty and at their top weight, the boys had weight andBMI above the population mean as evidenced by weight SDSand BMI SDS above zero (p< 0.001) (Table 1). The 13 boyswho had not started their pubertal growth spurt were of averagestature for age, but the 33 pubertal boys had a height SDS abovepopulation average at their top weight (p< 0.01). Up to presenta-tion, boys who had not started their growth spurt lost 5.9� 5.3 kgover 347� 239 days. During this period, they grew only3.9� 3.3 cm and lost 0.28� 0.22 SDS in height. Boys who hadstarted their pubertal growth spurt lost 10.3� 8.1 kg over296� 201 days. During the period of weight loss, they grew only2.6� 2.2 cm and lost 0.20� 0.27 SDS in height. None of the boyshad a complete arrest of growth.

Following the start of treatment, there was rapid weight gain inboth groups of boys (Figure 1(A)). The younger boys, who hadnot started their growth spurt at presentation, increased theirgrowth rate during the first year of treatment and maintainedthe higher rate for 2 years (Figure 1(B)), which resulted incatch-up growth (Figure 1(C)). The subgroup followed up for3 years thus did not differ in height SDS compared with theirprepubertal and top weight measurements (Table 1). At 3 years,they had reached the population mean (i.e. SDS= 0) for weightSDS and BMI SDS and had not returned to the higher levelsobserved before puberty (p< 0.05) or at their top weight(p< 0.01).

Boys who had started their pubertal growth spurt before assess-ment had a pattern of weight gain during treatment similar to thatof the younger boys (Figure 1(A)). They did, however, not in-crease their growth rate upon the start of treatment but showeda successive decline of linear growth (Figure 1(B)). Consequently,they did not catch up in height (Figure 1(C)). The subgroupfollowed up for 3 years had lost 0.64� 0.55 height SDS (p< 0.05)compared with their top weight measurements. At 3 years, theirweight SDS and BMI SDS did not differ from the population

scent boys

Growth spurt at presentation

-up Prepubertal At top weight At presentation 3-year follow-up

30 33 33 15

.2 7.7� 1.1 14.4� 1.4 15.2� 1.4 18.3� 1.4

.5 30.2� 9.1 62.9� 11.9 53.1� 7.4 67.3� 11.2

129� 8 170� 10 173� 9 176� 7

2 17.9� 3.7 21.7� 3.9 17.8� 1.7 21.7� 2.6

.92 0.82� 1.50 0.85� 1.11 �0.57� 1.14 �0.30� 1.39

.74 0.18� 1.16 0.50� 1.11 0.28� 1.19 �0.57� 1.03

.19 0.97� 1.51 0.73� 1.43 �1.03� 1.16 0.05� 1.18

Rev. 21 (2013) 395–398 © 2013 John Wiley & Sons, Ltd and Eating Disorders Association.

Figure 1. (A) Weight gain (kg/year), (B) growth velocity (cm/year) and (C) catch-

up growth (standard deviation scores (SDS)/year) before presentation and dur-

ing treatment of boys with eating disorders, weight loss and stunting of

growth. Values are means� standard error of the mean. Thirteen boys

presented before having started their pubertal growth spurt (●), and 33 boys

had started their growth spurt prior to presentation (○)

I. Swenne Growth of Boys with Eating Disorders

mean (i.e. SDS= 0) but were lower than at their top weight(p< 0.05).

Predictors of growth were analysed by entering ages, weights,and weight changes for all individual 1-year periods of treatment

Eur. Eat. Disorders Rev. 21 (2013) 395–398 © 2013 John Wiley & Sons, Ltd and Eating Disorder

as independent variables against either growth velocity or catch-up growth as the dependent variable in a stepwise multiple linearregression analysis (14). During a year of treatment, growth veloc-ity was predicted by age (R2 = 0.279, p< 0.001) and by the weightvelocity during that year (R2 = 0.047, p< 0.01). Catch-up during ayear of treatment was predicted by the weight velocity during thatyear (R2 = 0.051, p< 0.01) and also by the weight velocity of thepreceding year (R2 = 0.045, p< 0.05).

Discussion

The present investigation supports previous observations ofstunting of growth in adolescent boys with ED and weight loss(Root & Powers, 1983; Pfeiffer et al., 1986; Danziger et al.,1994; Modan-Moses et al., 2003). In boys who have not startedtheir pubertal growth spurt, due to their age and/or the effectsof starvation, catch-up growth to the premorbid growth trajec-tory appears possible. In older boys, who have startedtheir growth spurt prior to onset of weight loss, catch-upgrowth does not occur, and they will not reach their potentialtarget height.

At presentation, the younger boys resemble the premenarchealgirls previously studied in that there is stunting of growth anddelay of pubertal development (Swenne & Thurfjell, 2003;Swenne, 2005). They differ in that catch-up starts already duringthe first year of weight gain. This is likely to be related to thatthe boys had a relatively short period of undernutrition and adistinct onset of weight loss starting from a normal or in somecases an above-average weight. Girls with an insidious start ofinsufficient weight gain and stunting of growth have a longerduration of undernutrition, which does not reverse as rapidly(Swenne, 2005; Swenne, 2012). However, such girls have a delayin bone age (Root & Powers, 1983; Lacey, Crisp, Hart, & Kirk-wood, 1979), growth potential is thus conserved and catch-upcan be achieved by extending the growth period once nutritionis improved (Swenne, 2005). The young boys presently studiedare also likely to have a delay in bone age (Root & Powers,1983) but respond to weight recovery by rapid growth and returnto their previous growth trajectory. It is notable that this catch-upis achieved by returning to an average BMI (i.e. BMI SDS= 0) andthat the higher premorbid BMI SDS trajectory is not necessary forcatch-up. In view of that an (perceived) overweight may havebeen a cue to start weight-losing practices it can be reassuring thatit is adequate to set a target of average weight for height. Notably,these young boys were not followed up to their final height butwere growing significantly during the third year of treatment.To be able to conclude with certainty that catch-up and growthto their full potential height are achieved, a longer follow-upwould be necessary.

The older boys who had started their growth spurt beforelosing weight did not catch up to their previous growth trajectorybut continued to catch down. Their loss of height SDS during3 years of treatment would at an age of 18 years correspond to aloss of approximately 4.5 cm of potential height. They were notfollowed up to their final height, but their rapidly declininggrowth rate indicates that they would not catch up by extendingthe growth period but rather conclude growth at a stunted height.

397s Association.

Growth of Boys with Eating Disorders I. Swenne

This would suggest that although energy restriction and weightloss are sufficient to impede growth, bone maturation is not de-layed in proportion, and the epiphyseal growth plates continueto close (Tanner, 1981; Boersma & Wit, 1997). These boys hada BMI SDS above population average prior to onset of weightloss and recovered a BMI SDS close to population average(BMI SDS= 0) during treatment but did not return to theirpremorbid trajectory. In view of the fact that linear growth andcatch-up during treatment were related to weight gain, it maybe that these boys have a constitutionally high BMI and wouldhave benefited from further weight gain to achieve greater heightbefore closure of the epiphyseal growth plates. The observationthat weight loss after onset of the pubertal growth spurt can causepermanent stunting of growth is an observation pertinent also forthe younger boys who at the onset of treatment have not yetstarted their growth spurt. Should they successfully gain weightand enter puberty but relapse into weight loss, they would be atrisk of not reaching their potential adult height.

The present investigation shows that weight loss has a pro-found effect on the linear growth of boys. There may also be ef-fects on the quality of bone. In late adolescence, when boys areclose to concluding linear growth, there is still considerable

398 Eur. Eat. Disorders

calcium accretion in skeletal tissue, especially in the axial skeleton(Bonjour, Theintz, Buchs, Slosman, & Rizzoli, 1991; Bradneyet al., 2000; Nguyen et al., 2001). Bone mineral accretion is relatedto growth in size of the skeleton (Bonjour et al., 1991; Nguyenet al., 2001). An ED with weight loss and concurrent endocrinederangements during this period would thus not only stuntgrowth but also reduce peak bone mineral density and predisposeto future osteoporosis and fractures (Mehler, Sabel, Watson, &Andersen, 2008; Chevalley, Bonjour, van Rietbergen, Ferrari, &Rizzoli, 2011; Drake et al., 2012). Early detection of ED in adoles-cent boys and rapid start of treatment would thus be importantfor their adult stature and bone health.

Acknowledgements

This work was supported by the HRH Crown Princess LovisasFund for Child Health Care, the Gillbergska Foundation, the Firstof May Flower Annual Campaign, the Sven Jerring Foundationand Uppsala University.

The author reports no biomedical finance interests or otherpotential conflicts of interest.

REFERENCES

Abadie, I. B., de Tournemire, R., & Alvin, P. (2003). Anorexie

mentale: Conséquences sur la croissance et la minéralisation

osseuse. Archives de Pédiatrie, 10, 836–840.

American Psychiatric Association. (1994). Diagnostic and statistical

manual of mental disorders, 4th edn. (DSM-IV).Washington,

DC: American Psychiatric Association, pp. 539–550.

Boersma, B., & Wit, J. M. (1997). Catch-up growth. Endocrine Re-

views, 18, 646–661.

Bonjour, J.-P., Theintz, G., Buchs, B., Slosman, D., & Rizzoli, R.

(1991). Critical years of puberty for spinal and femoral bone

mass accumulation during adolescence. Journal of Clinical Endo-

crinology and Metabolism, 73, 555–563.

Bradney, M., Karlsson, M. K., Duan, Y., Stuckey, S., Bass, S., &

Seeman, E. (2000). Heterogeneity in the growth of the axial

and appendicular skeleton in boys: Implications for the patho-

genesis of bone fragility in men. Journal of Bone and Mineral Re-

search, 15, 1871–1878.

Chevalley, T., Bonjour, J. P., van Rietbergen, B., Ferrari, S., & Rizzoli,

R. (2011). Fractures during childhood and adolescence in boys:

Relation with bone mass, microstructure, and strength. Journal

of Clinical Endocrinology and Metabolism, 96, 3134–3142.

Danziger, Y., Mukamel, M., Zeharia, A., Dinari, G., & Mimouni, M.

(1994). Stunting of growth in anorexia nervosa during the prepubertal

and pubertal period. Israeli Journal of Medical Sciences, 30, 581–584.

Drake, M. T., Murad, M. H., Mauck, K. F., Lane, M. A., Undavalli,

C., Elraiyah, T., et al. (2012). Risk factors for low bone mass-

related fractures in men: A systematic review and meta-analysis.

Journal of Clinical Endocrinology and Metabolism, 97, 1861–1870.

Gowers, S. G., Crisp, A. H., Joughin, N., & Bhat, A. (1991).

Premenarcheal anorexia nervosa, Journal of Child Psycholology

and Psychiatry, 32, 515–524.

Lacey, J. H., Crisp, A. H., Hart, G., & Kirkwood, B. A. (1979). Weight

and skeletal maturation—A study of radiological and chronolog-

ical age in an anorexia nervosa population. Postgraduate Medical

Journal, 55, 381–385.

Lantzouni, E., Frank, G. R., Golden, N. H., & Shenker, R. I.

(2002). Reversibility of growth stunting in early onset

anorexia nervosa: A prospective study. Journal of Adolescent

Health, 31, 162–165.

Lindgren, G., Strandell, A., Cole, T., Healy, M., & Tanner, J. (1995).

Swedish population standards for height, weight and body mass

index attained at 6 to 16 years (girls) and or 19 years (boys). Acta

Paediatrica, 84, 1019–1028.

Mehler, P. S., Sabel, A. L., Watson, T., & Andersen, A. E. (2008).

High risk of osteoporosis in male patients with eating disorders.

International Journal of Eating Disorders, 41, 666–672.

Modan-Moses, D., Yaroslavski, A., Novikov, I., Segev, S., Toledano,

A., Miterany, E., et al. (2003). Stunting of growth as a major fea-

ture of anorexia nervosa in male adolescents. Pediatrics, 111,

270–276.

de Monléon, J. V., Simonin, G., Giraud, P., Achache, F., Sokolowsky,

M., Battista, M., et al. (1998). Déficit statural dû à l’anorexie

mentale. A propos de deux jumelles monozygotes. Annales de

Pédiatrie, 45, 702–706.

Nguyen, T. V., Maynard, L. M., Towne, B., Roche, A. F.,

Wisemandle, W., Li, J., et al. (2001). Sex differences in bone mass

acquisition during growth. Journal of Clinical Densitometry, 4,

147–157.

Rev. 21 (2013) 395–398 © 20

Pfeiffer, R. J., Lucas, A. R., & Ilstrup, D. M. (1986). Effect of anorexia

nervosa on linear growth. Clinical Pediatrics, 25, 7–12.

Prabhakaran, R., Misra, M., Miller, K. K., Kruczek, K.,

Sundaralingam, S., Herzog, D. B., et al. (2008). Determinants

of height in adolescent girls with anorexia nervosa. Pediatrics,

121, e1517–e1523.

Prader, A. (1978). Catch-up growth. Postgraduate Medical Journal, 54

(suppl 1), 133–146.

Root, A. W., & Powers, P. S. (1983). Anorexia nervosa presenting as

growth retardation in adolescents. Journal of Adolescent Health

Care, 4, 25–30.

Rosé, C., Doyen, C., Le Heuzey, M.-F., Armoogum, P., Mouren, M.-

C., & Léger, J. (2007). Predictors of late menarche and adult

height in children with anorexia nervosa. Clinical Endocrinology,

67, 462–467.

Russel, G. F. M. (1985). Premenarcheal anorexia nervosa and its se-

quelae. Journal of Psychiatric Research, 19, 363–369.

Swenne, I. (2005). Weight requirements for catch-up growth in girls

with eating disorders and onset of weight loss before menarche.

International Journal of Eating Disorders, 38, 340–345.

Swenne, I. (2012) Weight requirements for catch-up growth in ado-

lescent girls with eating disorders. In V. R. Preedy (Ed.), Hand-

book of growth and growth monitoring in health and disease.

London, UK: Springer Verlag, pp. 1015–1027.

Swenne, I., & Thurfjell, B. (2003). Clinical onset and diagnosis

of eating disorders in premenarcheal girls is preceded by

inadequate weight gain and growth retardation. Acta

Paediatrica, 92, 1133–1137.

Tanner, J. M. (1981). Catch-up growth in man. British Medical

Bulletin, 37, 233–238.

13 John Wiley & Sons, Ltd and Eating Disorders Association.