Introduction
The human immunodeficiency virus seems to exert a direct effect on protein metabolism and energy consumption, probably via proinflammatory cytokines[1,2]. This results in a disproportionate depletion of fat-free mass (FFM), the body‚s reservoir for protein, with relative sparing of fat. Although these changes in body composition can be observed relatively early in asymptomatic weight-stable patients[3], they are more pronounced during disease progression. They may eventually result in wasting, which is an AIDS-defining condition [4] and a predictor of disease progression[5].
Longitudinal studies [5,6] did not demonstrate a spontaneous restoration of weight in untreated patients with HIV infection or AIDS, despite apparent clinical stability. Several clinical trials [7-9] showed that combination regimens including a PI increase absolute CD4 lymphocyte counts, reduce plasma viral load and improve survival. More recently, it has been shown that PI-containing regimens improve the nutritional status of patients with HIV-related wasting syndrome[10,11]. However, some authors suggest that weight gain is mainly derived from an increase in fat mass[12].
Furthermore, metabolic studies on the effects of highly active antiretroviral therapy (HAART) [10,13]revealed inconsistent alterations in resting energy expenditure (REE).
We therefore investigated the changes in body composition and REE during the first 6 months of a newly initiated PI-containing regimen in a homogenous group of clinically stable HIV-infected and AIDS patients without wasting syndrome.
Materials and methods
Study design
A prospective longitudinal cohort study with individually matched healthy controls at a tertiary care centre of a University Hospital. Untreated stable patients have previously been followed for more than 6 months without exhibiting changes in body composition or REE[14-17]. However, as it is currently considered inappropriate to withhold HAART from HIV-infected individuals, we did not assign a placebo-controlled group of patients to single out the effects of PI therapy on body composition and REE.
Ethical approval
All patients and healthy controls gave written informed consent, and the study was approved by the Ethics Committee of the Vienna General Hospital.
Study population
In 1998, 37 clinically stable HIV-infected outpatients of either sex (25 men, 12 women; mean age: 35 years; range: 22-53 years) were recruited over 6 months from the HIV outpatient clinic of the Vienna General Hospital.
As a control group, 37 HIV-seronegative, healthy, weight-stable volunteers, who were matched for age, sex and body mass index (BMI), were investigated. They underwent an analysis of body composition, REE, and 7 day dietary intake.
All patients included were seropositive for HIV, as shown using enzyme-linked immunosorbent assay and a confirmatory immunoassay. The Centers for Disease Control and Prevention (CDC) status was recorded for all subjects[4]. AIDS diagnosis was not an exclusion criterion. However, patients with HIV-associated wasting were not recruited. Patients who had previously been taking nucleoside analogues had been free of this treatment for at least one year, and all of them were naive to protease inhibitors.
To isolate the direct or indirect effects of HIV infection during HAART on parameters of energy balance and body composition, the following inclusion criteria were defined: stable body weight over the past year (< 5% changes in weight); no clinical signs of diarrhoea (over four stools per day); no opportunistic infection within 3 months of screening; no persisting fever (> 37.8°C); no night sweats or elevation of C-reactive protein (CRP; > 1 mg/dl) 6 weeks before the beginning of the study[18-21].
Exclusion criteria were significant cardiovascular disease, malignancy, and endocrinological disease; the use of testosterone, anabolic steroids, growth hormone, systemic steroid therapy, or any drug abuse, because opioids[22], cocaine[23], or marijuana [24] may influence the parameters of energy expenditure. Moreover, wasting in association with heroin and cocaine abuse has been reported previously[25], and conversely, marijuana has been recommended for the treatment of HIV-associated wasting because of its appetite-stimulating effect.
Patients who developed an opportunistic infection during the course of the study or who did not adhere to the study protocol were excluded from the final evaluation.
All patients agreed to start a triple combination therapy consisting of nelfinavir 750 mg three times a day, didanosine 400 mg a day (250 mg for body weight < 60 kg), and stavudine 40 mg twice a day (30 mg for body weight < 60 kg), according to the current guidelines for antiretroviral treatment[26]. To preserve future treatment options in case the original regimen failed[26], nelfinavir was chosen for primary protease inhibitor therapy because its HIV-1 resistance profile differs from that of other protease inhibitors[27]. The two nucleoside analogues, didanosine and stavudine, were added because most of our pretreated patients had received zidovudine alone, or in combination with lamivudine or zalcitabine.
Furthermore, the choice of this triple regimen was made on the basis of practical issues such as regimen complexity, drug interactions and tolerance, and maintaining the current concept of maximizing the therapeutic benefit over time.
Experimental protocol
Patients were evaluated before the initiation of HAART and 24 weeks thereafter. The following measurements were performed at both visits: height, weight, tetrapolar bioelectrical impedance analysis (BIA) and anthropometric measurements (upper arm circumference) to determine the parameters of body composition; indirect calorimetry to determine REE; a 7 day food record; fasting glucose and serum triglycerides, cholesterol, albumin, CRP, thyroid stimulating hormone (TSH), triiodothyronine, testosterone; peripheral blood CD4 lymphocyte subset counts; viral load assessment; and stool samples for analysis of gastrointestinal pathogens. As shown in Fig. 1, additional assessments of viral load, CD4 lymphocyte counts, and REE were taken at weeks 8 and 12. All patients were routinely seen at 4 weeks intervals as outpatients to control the clinical course, and their adherence to HAART.
Body height was measured to the nearest 1 cm, and body weight to 0.1 kg, using a statiometer and digital electronic scales (Seca, Hamburg, Germany). Body composition was assessed after an overnight fast, and after voiding, within the first 5 min of resting in the supine position, with a bioimpedance apparatus at 50 kHz (Bioelectrical Impedance Analyser model BIA-101; RJL Systems, Florence, Italy) using standardized procedures[28,29]. Body composition parameters were evaluated using the software BODYGRAM2 version 2.2 (Akern-Florence, Italy). Coefficients of variation (CV) for repeated measurements of resistance and reactance on the same day were 0.7 and 1.0%, respectively; and the CV for the day-to-day variability averaged 2.0 and 3.9%, respectively. Fat-free mass (FFMKo) was calculated by Kotler‚s corrected sex-specific predictive equation for FFM, using parallel impedance, height, and weight[30], and fat mass (FM) was calculated as follows: FM = kg body weight - FFMKo.
For indirect calorimetry, patients and healthy volunteers were studied under standardized conditions in a quiet, temperature-controlled (22°C) room after an overnight fast. Readings were taken over 30 min, after a resting period of 30 min and an additional adaptation period of 10 min to allow for equilibration. Oxygen uptake and carbon dioxide excretion were measured using the Datex Deltatrac Metabolic Monitor (Datex Instruments, Helsinki, Finland). Absolute resting energy expenditure (REEabs) was calculated from respiratory gas analysis using the equation of Weir[31]. The CV for repeated measurements of oxygen consumption and carbon dioxide consumption on the same day were 6.5 and 9.2%. The CV for day-to-day variation of REE averaged 1.8%.
Patients and volunteers documented their daily food and beverage intake for seven consecutive days. All participants were instructed how to estimate their food and beverage consumption. Calculations of the intake of all major and micronutrients were evaluated by using a nutrition assessment software (ewp 3.0 dato denkwerkzeuge, Vienna, Austria). Subjects completed the diary within 10 days of the measurement of REE. All measurements were performed by one experienced investigator (K.S.) from the Institute of Nutritional Sciences.
Biochemistry
Fasting glucose and serum triglyceride levels, cholesterol, albumin and CRP were measured using a Hitachi 747 analyser; and TSH, triiodothyronine and testosterone were measured using an Elecsys 2010 analyser (both Roche Diagnostic Systems, Inc., Branchburg, NJ, USA).
Plasma HIV-RNA quantification
Plasma HIV-RNA levels were determined using an in-vitro nucleic acid amplification test (lower limit of detection: 50 copies/ml; Roche Ultrasensitive Assay; Roche Diagnostic Systems).
CD4 cell counts
Peripheral blood CD4 lymphocyte subset counts were determined by FACScan flowcytometer using monoclonal antibodies, and the Cellquest software (Becton-Dickinson, Mountain View, CA, USA)[32].
Data analysis
Data are expressed as means, and the range, unless otherwise indicated. Statistical calculations were carried out by the SAS Cary NC software package (SAS Institute, Cary, NC, USA), using non-parametric methods because data were non-normally distributed. Comparison between groups was performed using the Kruskal-Wallis analysis of variance and the Mann-Whitney U test. Changes within the treatment groups were tested using the Wilcoxon signed ranks test.
Results
Thirty-seven HIV-seropositive patients were included in the study (25 men, 12 women; differences in sex between groups: P < 0.02 at baseline, and P < 0.05 at final evaluation). Seventeen of these patients (46%) had AIDS in accordance with the CDC classification[4]. Previous treatment included zidovudine alone (n = 3), or in combination with lamivudine (n = 2) or zalcitabine (n = 1).
On an intent-to-treat-basis, 70% of all patients (n = 26) responded to PI therapy, as defined by a decrease of HIV-RNA plasma levels below 200 copies per ml after 24 weeks (‚responders‚). Fourteen of those ‚responders‚ even reached the lower level of detectability, i.e. 50 copies per ml after 8 weeks (n = 2), after 12 weeks (n = 4), and after 24 weeks (n = 8) of PI treatment.
During the 6 months of observation, 11 participants (i.e. eight AIDS and three HIV patients) had to be excluded from final evaluation (‚drop-outs‚). However, those 11 patients of the ‚drop-out group‚ also responded to HAART, as defined by a decrease in viral load by one log in 8 weeks[33]. In fact, the ‚drop-outs‚ showed a decrease in plasma viral load of 2.0 (± 0.4) log versus 2.7 (± 0.3) log in the ‚responders‚ after 8 weeks of HAART (P < 0.05 between groups). This difference could easily be explained by the non-compliance of the ‚drop-outs‚, rather than a true inefficacy of the HAART regimen. This notion is also supported by the finding that viral load decreased further and similarly in both groups after 12 weeks of treatment, i.e. by 3.4 (± 0.3) log in the remaining ‚drop-outs‚, and by 3.1 (± 0.2) log in the ‚responders‚ (P > 0.05).
The reasons for exclusion and follow-up data on the ‚drop-outs‚ are depicted in Fig. 2. The demographics of patients and controls are given in Table 1, and final outcome values for body composition and metabolic parameters are presented in Table 2.
Body weight and body composition
Baseline body weight, BMI, and FFMKo were lower in the AIDS than in the HIV group (P < 0.02 between groups), whereas FM was similar in the HIV and in the AIDS group (P > 0.05) (Table 1). Body weight, BMI, and FFM increased compared with baseline in both groups (P < 0.05). These changes tended to be more pronounced in the AIDS group. FM increased only in AIDS-patients, but remained stable in HIV patients (Table 2).
Resting energy expenditure
In order to single out abnormal metabolic baseline parameters we compared HIV and AIDS patients with their corresponding controls: REE was approximately 12% higher, and REE adjusted for weight and for fat-free mass (REE/FFM) were more than 15% higher in HIV and AIDS patients than in their corresponding controls (P values < 0.05 versus control).
Baseline REE, REE/weight and REE/FFM (Table 1) were not different between HIV and AIDS patients (P > 0.05 between groups), and decreased in both groups in a similar way during therapy (P < 0.05) (Table 2). After therapy, there was no difference between patients and healthy controls (P > 0.05 versus control).
Energy intake
Average total caloric intake calculated from diet records was not significantly different between patients and controls (P > 0.05) (Table 1). Likewise, no differences were observed in the total caloric intake of HIV and AIDS patients. The energy, protein, lipid and carbohydrate content of the diet remained stable in both groups during PI therapy (P > 0.05) (data not shown). No patient suffered from substantial diarrhoea associated with antiretroviral treatment.
Biochemical assays
Before the initiation of HAART, fasting glucose, serum triglycerides, cholesterol, albumin, TSH, triiodothyronine, and testosterone were within the normal range in the HIV and in the AIDS group. During PI therapy, fasting triglycerides increased in the AIDS group and increased significantly in the HIV group (P < 0.05 in the HIV group, P > 0.05 between groups) (Table 2). Total serum cholesterol and HDL increased in both groups (P < 0.05 within groups, P > 0.05 between groups) (Table 2).
Discussion
It has recently been shown that PI therapy results in weight gain in HIV-infected patients with a history of weight loss or wasting syndrome[10,12]. In our study, we demonstrate that weight increases in clinically stable HIV-infected as well as in AIDS patients during HAART.
We further report for the first time a decrease in REE during PI therapy, which may partly account for the weight gain. During PI therapy, patients in both groups gained weight, BMI, and FFM. The extent of these changes appears to be clinically relevant, and tended to be more pronounced in the AIDS than in the HIV group. The preferential gain in fat mass in the AIDS group of our trial is in line with previous findings[10,12]. Fat gain might reflect a physiological mechanism to reconstitute high-energy storage for periods of physical distress[34]. However, both groups in our study also gained FFM. This is noteworthy, because AIDS patients, similar to stress or trauma patients[5], exhibit complex patterns of metabolic and endocrine alterations, with a trend towards poor FFM retention[34].
It is possible to assume that our patients had preserved the ability to gain weight in quite a physiological manner, which comprises the increase of FM and FFM[35]. This notion is supported by the fact that, similar to healthy subjects, more than 70% of our patients were going to work regularly, and 14 patients reported regular physical training twice a week.
It is obvious that body weight, BMI and FFM were lower at baseline in the AIDS group. This effect was conceivably related to a preponderance of female patients in the AIDS group: because recruitment was restricted to 6 months and was performed in a consecutive manner by two study nurses who were blinded to the stage of infection of the patients, sex was finally not equally distributed. However, comparisons of the baseline BMI between groups, after adjustment for age and sex, showed that 85 and 71% of the HIV patients and AIDS patients, respectively, had a BMI less than the 50th percentile. This indicates that BMI values in both groups were below average, but were in fact not different after adjustment for sex.
Under physiological circumstances, body weight gain is most frequently caused by increased food intake. In our population, energy intake remained stable during PI therapy, which is in contrast to previous observations of increased food intake during HAART[10,12]. A possible explanation for this finding might be that our patients were not wasted, and exhibited normal appetite throughout the study. Therefore the stimulus to increase caloric intake during therapy was less pronounced than in weight-losing patients. Moreover, in our patients, weight gain occurred in the absence of additional anabolic stimuli such as androgen therapy, appetite stimulants, or human growth hormone[36-38]. Changes in body composition were therefore not attributable to a greater food intake.
Of note is the fact that BMI increased as early as viral load started to fall and CD4 lymphocyte counts started to recover (Fig. 1). Our data are, thus, consistent with a recently observed correlation between changes in body weight and changes in CD4 lymphocyte counts[10].
Interestingly, changes in REE were observed later than the increases in BMI. Primary weight gain did therefore not seem to be directly related to changes in REE. The reduction in REE could, however, have contributed to weight gain on a long-term basis, but we can only speculate about the mechanisms leading to a reduction in REE during antiretroviral treatment.
First, the reduction of viral load under HAART, and thus the reduced virion- and CD4 cell turnover preceded the decline in REE, which represents a prerequisite for a cause-effect relationship (Fig. 1). Second, Mulligan et al. [39] have demonstrated a positive correlation between viral load and REE in clinically stable patients. On the basis of these findings, they postulated that sustained reductions in viral load with potent antiretroviral therapy would decrease REE and blunt the response to immune activation. Indeed, we now demonstrate for the first time a reduction in viral load under HAART, which is followed by a uniform decrease in REE in stable HIV-infected individuals as well as in AIDS patients (Fig. 1).
As a limitation of the study, we do not have any direct proof that the decline in viral load under HAART has reduced REE. However, we postulate that the failure to suppress viral load, i.e. persisting viral replication, would be paralleled by the persisting elevation of REE. This concept is supported by data showing no spontaneous changes during serial assessments of REE in HIV-infected individuals with ongoing viral replication[14,17]. Therefore, it seems unlikely that the decline in REE and the subsequent weight gain in our patients under HAART were solely a matter of chance.
Of note is the fact that the reduction of REE in our study is at variance with previous findings of unchanged[10,12], or even elevated [13] REE under HAART. Unfortunately, the latter abstract did not provide information as to whether elevated REE under HAART was a short-term rather than a long-term effect. It is conceivable that the medication itself exerted a metabolic effect during the very early treatment phase, and was responsible for the observed increase in REE[40].
The limited sample size of the study population at the final evaluation may be a shortcoming of our study. However, the initial calculation of the sample size was based on previous body composition studies in wasting patients[10,12], and changes in our asymptomatic patients proved to be 50% smaller. A power analysis revealed that we would have needed approximately 46 HIV-infected and 85 AIDS patients, respectively, considering comparable drop-out rates, to demonstrate statistical significant differences between the HIV group and the AIDS group in weight and BMI changes. This would probably have required a multicentre study, to apply the stringent exclusion criteria and to warrant a relatively short recruitment period.
Conclusion
Protease inhibitor therapy is beneficial in terms of body composition and metabolic parameters, to both HIV-infected and AIDS patients, even in the absence of wasting. The decrease in HIV replication and a consecutive decrease in the turnover of newly infected CD4 lymphocytes under HAART may contribute to a slightly delayed reduction of REE in otherwise stable patients, irrespective of the stage of infection. Decreased REE could, thus, also contribute to weight gain on a long-term basis.
Acknowledgements
The authors wish to thank Thomas Pernerstorfer, MD, for helpful discussions and for sharing his scientific experience. The authors also thank Martin Stepan, PhD, from Roche Austria GmbH, for continued support, the Department of Internal Medicine III, Division of Nephrology, for providing the Datex Deltatrac Metabolic Monitor, and Claudia Vilmetti, MD, for technical assistance.
References
1. Kotler D, Heymsfield SB. HIV infection: a model chronic illness for studying wasting diseases. Am J Clin Nutr 1998, 68:519-520.
2. Macallan DC, Noble C, Baldwin C, et al. Energy expenditure and wasting in human immunodeficiency virus infection. N Engl J Med 1995, 333:83-88.
3. Ott M, Lembcke B, Fischer H, et al. Early changes of body composition in human immunodeficiency virus-infected patients: tetrapolar body impedance analysis indicates significant malnutrition. Am J Clin Nutr 1993, 57:15-19.
4. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep 1992, 41:1-19.
5. Kotler DP, Wang J, Pierson RN. Body composition studies in patients with the acquired immunodeficiency syndrome. Am J Clin Nutr 1985, 42:1255-765.
6. Chlebowski RT, Grosvenor MB, Bernhard NH, Morales LS, Bulcavage LM. Nutritional status, gastrointestinal dysfunction, and survival in patients with AIDS. Am J Gastroenterol 1989, 84:1288-793.
7. Gulick RM. Current antiretroviral therapy: an overview. Qual Life Res 1997, 6:471-474.
8. Hammer SM, Squires KE, Hughes MD, et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team. N Engl J Med 1997, 337:725-733.
9. Mouton Y, Alfandari S, Valette M, et al. Impact of protease inhibitors on AIDS-defining events and hospitalizations in 10 French AIDS reference centres. Federation National des Centres de Lutte contre le SIDA. AIDS 1997, 11:F101-105.
10. Carbonnel F, Maslo C, Beaugerie L, et al. Effect of indinavir on HIV-related wasting. AIDS 1998, 12:1777-1784.
11. Sherer R. Current antiretroviral therapy and its impact on human immunodeficiency virus-related wasting. Semin Oncol 1998, 25:92-97.
12. Silva M, Skolnik PR, Gorbach SL, et al. The effect of protease inhibitors on weight and body composition in HIV-infected patients. AIDS 1998, 12:1645-1651.
13. Shevitz A, Spiegelman D, French J, et al. Increased resting energy expenditure (REE) is associated with increased plasma HIV RNA. 12th World AIDS Conference. Geneva, 1998.
14. Grinspoon S, Corcoran C, Askari H, et al. Effects of androgen administration in men with the AIDS wasting syndrome. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1998, 129:18-26.
15. Oster MH, Enders SR, Samuels SJ, et al. Megestrol acetate in patients with AIDS and cachexia. Ann Intern Med 1994, 121:400-408.
16. Schambelan M, Mulligan K, Grunfeld C, et al. Recombinant human growth hormone in patients with HIV-associated wasting. A randomized, placebo-controlled trial. Serostim Study Group. Ann Intern Med 1996, 125:873-882.
17. Suttmann U, Ockenga J, Hoogestraat L, et al. Resting energy expenditure and weight loss in human immunodeficiency virus-infected patients. Metabolism 1993, 42:1173-1179.
18. Sharpstone DR, Murray CP, Ross HM, et al. Energy balance in asymptomatic HIV infection. AIDS 1996, 10:1377-1384.
19. Sharpstone DR, Ross HM, Gazzard BG. The metabolic response to opportunistic infections in AIDS. AIDS 1996, 10:1529-1533.
20. Melchior JC, Salmon D, Rigaud D, et al. Resting energy expenditure is increased in stable, malnourished HIV-infected patients. Am J Clin Nutr 1991, 53:437-441.
21. Melchior JC, Raguin G, Boulier A, et al. Resting energy expenditure in human immunodeficiency virus-infected patients: comparison between patients with and without secondary infections. Am J Clin Nutr 1993, 57:614-619.
22. Hoffman WE, Berkowitz R, McDonald T, Hass F. Ultra-rapid opioid detoxification increases spontaneous ventilation. J Clin Anesth 1998, 10:372-376.
23. Favier R, Caceres E, Koubi H, Sempore B, Sauvain M, Spielvogel H. Effects of coca chewing on hormonal and metabolic responses during prolonged submaximal exercise. J Appl Physiol 1996, 80:650-655.
24. Renaud AM, Cormier Y. Acute effects of marihuana smoking on maximal exercise performance. Med Sci Sports Exerc 1986, 18:685-689.
25. Hoy JF, Flanigan TP. Wasting associated with cocaine and heroin use in patients infected with human immunodeficiency virus. Clin Infect Dis 1994, 19:209-210.
26. Carpenter CC, Fischl MA, Hammer SM, et al. Antiretroviral therapy for HIV infection in 1998: updated recommendations of the International AIDS Society - USA Panel. JAMA 1998, 280:78-86.
27. Patick AK, Mo H, Markowitz M, et al. Antiviral and resistance studies of AG1343, an orally bioavailable inhibitor of human immunodeficiency virus protease. Antimicrob Agents Chemother 1996, 40:292-297.
28. Lukaski HC, Johnson PE. A simple, inexpensive method of determining total body water using a tracer dose of D2O and infrared absorption of biological fluids. Am J Clin Nutr 1985, 41:363-370.
29. Shizgal HM. Validation of the measurement of body composition from whole body bioelectric impedance. Infusionstherapie 1990, 17 (Suppl. 3):67-74.
30. Kotler DP, Burastero S, Wang J, Pierson RN Jr. Prediction of body cell mass, fat-free mass, and total body water with bioelectrical impedance analysis: effects of race, sex, and disease. Am J Clin Nutr 1996, 64:489S-497S.
31. Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. Nutrition 1990, 6:213-221.
32. Landay AL, Muirhead KA. Procedural guidelines for performing immunophenotyping by flow cytometry. Clin Immunol Immunopathol 1989, 52:48-60.
33. Centers for Disease Control and Prevention. 1998 guidelines for treatment of sexually transmitted diseases. MMWR Morb Mortal Wkly Rep 1998, 47:1-111.
34. Strawford A, Hellerstein M. The etiology of wasting in the human immunodeficiency virus and acquired immunodeficiency syndrome. Semin Oncol 1998, 25:76-81.
35. Dulloo AG, Jacquet J, Girardier L. Autoregulation of body composition during weight recovery in human: the Minnesota Experiment revisited. Int J Obes Related Metab Disord 1996, 20:393-405.
36. Von Roenn JH, Armstrong D, Kotler DP, et al. Megestrol acetate in patients with AIDS-related cachexia. Ann Intern Med 1994, 121:393-399.
37. Waters D, Danska J, Hardy K, et al. Recombinant human growth hormone, insulin-like growth factor 1, and combination therapy in AIDS-associated wasting. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1996, 125:865-872.
38. Gold J, High HA, Li Y, et al. Safety and efficacy of nandrolone decanoate for treatment of wasting in patients with HIV infection. AIDS 1996, 10:745-752.
39. Mulligan K, Tai VW, Schambelan M. Energy expenditure in human immunodeficiency virus infection. N Engl J Med 1997, 336:70-71.
40. Brinkman K, ter Hofstede HJ, Burger DM, Smeitink JA, Koopmans PP. Adverse effects of reverse transcriptase inhibitors: mitochondrial toxicity as common pathway. AIDS 1998, 12:1735-1744.
© 1999 Lippincott Williams & Wilkins, Inc.