To characterize the body composition of human immunodeficiency virus (HIV)–infected children, especially those with growth failure (GF), using laboratory-based methods.
A cross-sectional study of body composition measurements.
Urban, hospital-based body composition laboratory.
Thirty-four prepubertal children with HIV infection, aged 4 to 11 years, recruited from a pediatric HIV clinic. Eighteen HIV-infected children with GF, 16 HIV-infected children with normal rates of growth, and 52 healthy children were studied.
Main Outcome Measurements
Anthropometrics, body cell mass (BCM) by total body potassium counting, body fat percent, fat mass, and fat-free mass (FFM) by dual-energy x-ray absorptiometry were determined.
Both groups of boys with HIV infection had significantly lower FFM/height ratios compared with healthy boys. The mean BCM/height ratio was also lower in HIV-infected boys with GF compared with healthy boys. Measures of fat of the HIV-infected boys with GF did not differ from healthy controls, but a statistical trend suggesting decreased body fat percent and fat mass/height ratio was observed in HIV-infected boys without GF (P=.06 and .07, respectively). Mean height-for-age, weight-for-age, and weight-for-height percentiles were significantly decreased in HIV-infected boys regardless of growth status as compared with healthy boys. The mean FFM/height and BCM/height ratios were decreased in HIV-infected girls with GF compared with healthy girls. Body fat percentage and fat mass/height ratio did not differ among the 3 groups of girls. The mean weight-for-height percentiles were not different among the 3 groups of girls. The HIV-infected girls with GF had significantly lower mean height-for-age and weight-for-age percentiles than HIV-infected girls without GF and healthy girls. The mean height-for-age percentiles of the HIV-infected girls with GF did not differ from the healthy girls.
Boys and girls with HIV-associated GF had diminished FFM and BCM. The decrease in FFM and BCM was in striking contrast to the fat compartment, which was normal. Decreased FFM was also detected in boys with HIV infection and normal growth but not in girls with HIV infection and normal growth, suggesting that HIV infection may affect boys differently than girls. The preferential decrease in FFM and BCM over fat observed in these children is similar to findings reported in adults with acquired immunodeficiency syndrome wasting.
DISTURBANCES in growth, including growth failure (GF) and wasting, are common complications of childhood human immunodeficiency virus (HIV) infection and contribute to morbidity and mortality.1 Children with HIV infection and poor growth have as much as a 5-fold increase in the risk of death.2
Studies of children with HIV infection reveal early compromise of both height and weight gain.3- 5 Various patterns such as stunting, decreased weight-for-height, and weight loss have been reported. Few studies in children with HIV infection have included evaluations of body composition.6,7 These studies, which may not be relevant to older, perinatally infected children who now comprise the majority of cases, used anthropometric measurements such as weight, height, skin-fold thickness, and arm circumference in younger perinatally infected children or older children who acquired HIV through contaminated blood products. Assessment of body composition in children with chronic diseases is valuable for understanding the pattern of growth and development. Measurement of fat mass (FM) reflects the adequacy of energy stores while fat-free mass (FFM) reflects protein and muscle reserves. The amount of metabolically active tissue can be assessed by measuring the body cell mass (BCM), a component of FFM. Depletion of normal FM and FFM or BCM, which occurs in several chronic diseases, may influence survival and the quality of life.
Laboratory-based methods of body composition measurement have proven useful in characterizing the nature of acquired immunodeficiency syndrome (AIDS)–associated weight loss in adults, as well as in monitoring disease progression. In a report of clinically ill patients with AIDS studied prior to the modern era of treatment, adults with AIDS-associated weight loss preferentially lost BCM over fat, a pattern consistent with cachexia rather than starvation. Depletion of total body potassium, an index of BCM, predicts the time of death in adults with AIDS wasting.8 In adult AIDS patients with fewer than 200 CD4 cells per deciliter, loss of BCM is a more powerful prognostic measure for survival than CD4 cell count.9 Evaluation of changes in body composition has also been used to monitor response to therapies for AIDS wasting.10,11
The purpose of this study was to characterize the body composition of children with HIV infection using laboratory methods. In particular, we compared the body composition of HIV-infected children with and without GF to add to our understanding of its causes.
Thirty-four prepubertal children with HIV infection, aged 4 to 11 years, were recruited from the St Luke's–Roosevelt Hospital Center Pediatric HIV/AIDS program in New York City from January 1994 through November 1996. Eighteen HIV-infected children with GF and 16 HIV-infected children with normal rates of growth were enrolled. Growth failure was defined as a 12-month height velocity of the fifth percentile or lower for age using standard reference norms.12 One child acquired HIV as a result of transfusion of contaminated blood during the neonatal period. All others were HIV-infected as a result of perinatal transmission. Fifty-two healthy children of similar age and ethnicity, volunteers in an ongoing study of body composition in normal children, were used as controls. Healthy volunteers were recruited by notices in local newspapers, at schools, in after-school activity centers, and from children receiving routine health maintenance in a hospital-based clinic. A medical history obtained from a parent or guardian and a physical examination performed at the time of body composition evaluation confirmed normal health status. Informed consent was obtained from the parent or guardian for each child prior to participation; assent was also obtained whenever possible. The study was approved by the study site institutional review board.
Human immunodeficiency virus infection was diagnosed and disease stage was classified using Centers for Disease Control and Prevention criteria.13 Pubertal classification was performed according to Marshall and Tanner.14,15 Height-for-age, weight-for-age, and weight-for-height percentiles were calculated using EpiInfo software.16 Information concerning prior illnesses or other HIV-related conditions, treatments, and medications and results of lymphocyte phenotype analyses were obtained from medical records. No children treated with megestrol acetate or corticosteroids were enrolled. None of the subjects had clinically apparent renal or cardiac disease or known or suspected active intercurrent illnesses at the time of evaluation of body composition. Eighty-five percent of HIV-infected subjects were receiving antiretroviral medications.
Weight was measured with children wearing light clothing to the nearest 0.1 kg using a beam balance. Height was measured without shoes to the nearest 0.5 cm using a fixed wall-mounted stadiometer.
All body composition measurements were performed at the Body Composition Unit during a study visit at St Luke's–Roosevelt Hospital Center. Fat mass, body fat percent (BF%), and FFM were determined by dual-energy x-ray absorptiometry (Lunar DPX model, Lunar Radiation, Madison, Wis). Body cell mass was ascertained by the measurement of total body potassium in a 4-π whole-body liquid scintillation counter as previously described.17 Arm muscle area was calculated using tricep skin-fold thickness and midarm muscle circumference determined by a single researcher (J.W.) using a standardized technique.18 Fat-free mass, FM, and BCM were normalized for height to allow comparisons between groups.19
Descriptive statistics of the demographic variables were calculated. Means and SDs of the continuous variables were calculated by HIV infection status, growth category (ie, GF and normal growth), and sex. There were 3 groups each for boys and girls as follows: HIV-infected with GF, HIV-infected without GF, and healthy. Frequency distributions of the discrete variables were calculated in the same fashion.
The dependent variables were initially analyzed using 2-factor analysis of covariance (ANCOVA) for unbalanced designs, including HIV-infected with GF, HIV-infected without GF, healthy, and sex.
Because there were significant interactions between the HIV/GF status and sex, these data were reanalyzed using a single-factor ANCOVA for boys and girls. Age was used as the covariate and age-adjusted means of the dependent variables are reported for each group by sex.
All statistical calculations were performed using the SAS statistical software package for personal computers (SAS Institute, Cary, NC). The level of significance was .05 for all statistical tests.
The characteristics of boys and girls in the study are presented separately in Table 1. Consistent with our classification criteria, the mean growth rate (centimeters per year) and growth velocity percentile were lower in the HIV-infected children with GF. The mean height-for-age percentiles were lower among the boys and girls in the HIV-infected group with GF compared with the HIV-infected without GF and healthy groups.
No significant differences were observed between the boys and girls in corresponding growth pattern groups with respect to disease stage, degree of immune suppression as measured by CD4 cell count and CD4%, or the number of antiretroviral drugs prescribed. None of the subjects were receiving protease inhibitors at the time of evaluation.
The results of the analysis of anthropometric and age-adjusted means of body composition measurements for boys are presented in Table 2 and Figure 1. Mean height-for-age, weight-for-age, and weight-for-height percentiles were significantly decreased in the HIV-infected boys regardless of growth status, compared with healthy boys. Both groups of boys with HIV infection had significantly lower mean FFM compared with healthy children. The differences in FFM were also observed when height-standardized means were compared. Both groups of HIV-infected boys had lower mean BCM compared with healthy boys. However, only the HIV-infected boys with GF had lower mean BCM/height ratios compared with healthy boys. In addition, HIV-infected boys with GF had diminished mean arm muscle area compared with healthy boys. The HIV-infected boys without GF also had decreased mean arm muscle area compared with healthy boys, but the difference was not statistically significant (P=.09). Body fat percentage and FM/height ratio of the HIV-infected boys with GF did not differ from healthy controls. A statistical trend suggesting lower mean BF% and FM/height ratio in HIV-infected boys without GF compared with healthy boys was observed (P=.07 and .08, respectively).
Age-adjusted fat mass (FM), fat-free mass (FFM), and body cell mass (BCM), normalized for height, in human immunodeficiency virus (HIV)–infected boys with and without growth failure (GF) and healthy boys (top), and HIV-infected girls with and without GF and healthy girls (bottom).
A different pattern was observed among the girls. These results are presented in Table 2 and the bottom part of Figure 1. The HIV-infected girls with GF had significantly lower mean height-for-age and weight-for-age percentiles than HIV-infected girls without GF and healthy girls. In contrast to the boys, the mean height-for-age percentile of the HIV-infected girls without GF did not differ from the healthy controls and the mean weight-for-height percentiles were not different among the 3 groups of girls. The mean FFM and FFM/height ratio was lower in HIV-infected girls with GF compared with HIV-infected girls without GF and healthy girls. The mean BCM and BCM/height ratio was lower in the HIV-infected girls with GF compared with HIV-infected girls without GF and healthy girls. Arm muscle area was smaller in HIV-infected girls with GF. In addition, there was a statistical trend suggesting smaller arm muscle area in HIV-infected girls without GF compared with healthy controls (P=.06). Body fat percentage and FM/height ratios did not differ among the 3 groups.
To evaluate the effect of the inclusion of children with weight loss in the study results, a second analysis was performed with removal of 2 boys with weight loss of 10% of body weight or more. No significant changes in the intergroup differences in FM and FFM patterns were observed (results not shown).
In our sample, boys and girls with HIV-associated GF had diminished FFM. Our data indicate that the lower FFM is largely due to reductions in BCM, the major constituent of the FFM. The decrease in FFM is in striking contrast to the fat compartment, which was normal in children with HIV-associated GF. Decreases in FFM were also detectable in boys with HIV infection and normal growth, suggesting that alterations in FFM are measurable even in the absence of GF.
The decrease in FFM in childen with HIV-associated GF was demonstrated by arm muscle area calculated from anthropometric data as well as by dual-energy x-ray absorptiometry. These findings are similar to previous observations by anthropometrics reported in younger children. Miller et al6 reported decreased muscle mass measured by midarm circumference, with preservation of fat mass by age 2 years in children with perinatally acquired HIV. In contrast, a study of older hemophiliac boys with transfusion-acquired HIV and significantly impaired linear growth found no alteration in lean body mass by skinfold methods.7
The pattern of preferential decrease in FFM over fat observed in boys and girls with HIV-associated GF has been described in the cachexia of cancer as well as in adults with early HIV infection and with AIDS wasting,20,21 and is distinct from the body composition alterations observed in starvation and in children with nutritionally based stunting.22,23
The causes of GF in HIV infection are not well understood but are likely multifactorial. While endocrine abnormalities are occasionally encountered, including cases of classic growth hormone deficiency,24 overall endocrine function is usually intact.25,26 Gastrointestinal tract dysfunction, including infection and malabsorption, has also been reported, but no clear relationship to GF has been documented.27,28 Our data suggest that for some children with HIV infection, metabolic abnormalities leading to preferential adipogenesis over anabolism and growth may be involved. Disturbances in the growth hormone–insulinlike growth factor axis that have been described in HIV infection could play a role in the growth and body composition abnormalities we observed.29 Physical activity and exercise might also influence body composition and should be evaluated in future studies. The effect of antiretroviral medications on body composition also warrants investigation in this population, especially in light of recent reports of abnormal fat distribution in HIV-infected adults receiving protease inhibitors.30
Studies of AIDS-related weight loss and wasting in adults indicate that both decreased dietary intake and altered metabolism are factors. Abnormalities in metabolism, including increased basal energy requirements and altered fat metabolism, have been reported in some adults with HIV infection and AIDS wasting.31- 35 In adults these abnormalities seem to be less significant to actual weight loss than episodic decreases in dietary intake associated with secondary illnesses.36 There are no published studies measuring metabolic rates of children infected with HIV. Studies of dietary intake indicate that decreased nutrient intake is not encountered in children with HIV infection. The dietary intake among children with symptomatic HIV infection is reported to be equal to or greater than that in noninfected children.37 Also, increasing the nutritional intake of children with HIV-associated GF with use of enteral tube feedings improves weight but has no effect on linear growth deficits and lean body mass (eg, arm muscle area), suggesting that additional variables, such as metabolic abnormalities, are involved.38
Despite prior documentation of the high prevalence of abnormalities in linear growth, at present there is no height- or height velocity–defined diagnosis included in the CDC HIV classification system for children.12 The AIDS wasting syndrome, which involves weight loss or weight gain decelerations, is the only specifically defined growth abnormality included for children and is an AIDS-defining condition. Results from our study revealing compromises in the quantity of FFM in children with HIV-associated growth failure are an additional indication of the importance of abnormalities in linear growth. The prognostic significance of body composition abnormalities in children with HIV infection must be addressed in future studies. It is possible that changes in body composition may be identified as an additional risk factor for disease progression, as demonstrated in adults with AIDS.9
An unexpected finding in our study was the striking difference in anthropometric and body composition patterns between boys and girls with HIV infection. Boys with HIV infection without GF had significantly lower weight-for height and FFM/height ratios than the other boys, while the HIV-infected girls without GF did not. Our study was not designed to evaluate this specifically. In addition, our sample is not population-based; thus, it is not possible to determine whether HIV truly affects growth and body composition differently in boys and girls. Prior studies, including population-based studies, have not evaluated for sex-specific differences in growth characteristics.2- 4 Sex-specific effects in body composition changes in adults with AIDS have been reported. In contrast to men, women exhibit a disproportionate decrease in fat in the early stages of AIDS wasting, losing significant lean body mass in later stages.39 Sex differences in growth and body composition in children in response to environmental insults such as malnutrition and infection, with boys being more sensitive, have been described in developing countries.40 Future population-based studies are required to confirm sex differences in growth and body composition changes among children infected with HIV.
Children with HIV, especially those with GF, have decreased FFM and BCM. These functionally important compartments potentially affect both the quality of life and survival in these children. Additional studies evaluating the factors that influence growth and body composition, including the role of HIV replication on energy balance and metabolism, are required. Evaluating the effect of HIV therapies on body composition will also be important in the future.
Accepted for publication March 10, 1998.
This study was supported in part by grant 50641 from the Pediatric AIDS Foundation, Santa Monica, Calif, and grant DK37352 from the National Institutes of Health, Bethesda, Md.
A preliminary report of this study was presented at the annual meeting of the Society for Pediatric Research, Washington, DC, May 6, 1997. The abstract was published as: Arpadi S, Horlick M, Wang J, et al. Changes in body composition in HIV-infected children with growth failure. Pediatr Res. 1997;41:80A. Abstract 464.
We thank John C. Thornton, PhD, and Jason Barbour, MPH, for assistance in this study and Lenore S. Levine, MD, and Richard N. Pierson, MD, for their thoughtful review of the manuscript.
Editor's Note: More bad news about HIV infections—they bypass the fat and go straight to the good stuff.—Catherine D. DeAngelis, MD
Corresponding author: Stephen M. Arpadi, MD, MS, St Luke's–Roosevelt Hospital Center, 1111 Amsterdam Ave, New York, NY 10025 (e-mail: firstname.lastname@example.org).
Arpadi SM, Horlick MNB, Wang J, Cuff P, Bamji M, Kotler DP. Body Composition in Prepubertal Children With Human Immunodeficiency Virus Type 1 Infection. Arch Pediatr Adolesc Med. 1998;152(7):688-693. doi:10.1001/archpedi.152.7.688