HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kim, J. H Articles by Gorbach, S. L Search for Related Content PubMed PubMed Citation Articles by Kim, J. H Articles by Gorbach, S. L Agricola Articles by Kim, J. H Articles by Gorbach, S. L

The correlates of dietary intake among HIV-positive adults^1,2,3

¹ From the Departments of Population and International Health, of Biostatistics, of Epidemiology, and of Nutrition, Harvard School of Public Health, Boston, and the Department of Family Medicine and Community Health, Tufts University School of Medicine, Boston.

² Supported by NIDDK grant no. DK4 5734-03. The General Clinic Research Center of the New England Medical Center, Boston, was supported by the Division of Research Resources of the NIH, grant no. MO1-RR00054.

³ Address reprint requests to SL Gorbach, Department of Family Medicine and Community Health, Tufts University School of Medicine, 136 Harrison Avenue, Arnold 205, Boston, MA 02111. E-mail: sherwood.gorbach{at}tufts.edu.

	ABSTRACT

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Background: Dietary adequacy, as distinct from weight loss, has not been examined thoroughly in a diverse cohort of HIV-infected individuals.

Objective: An analysis was undertaken to determine the correlates of inadequate dietary intake among HIV-infected adults.

Design: In a cross-sectional study of 463 men and 170 women (aged 21–70 y) with HIV infection, dietary adequacy was evaluated by using 3-d diet records.

Results: Among nondieting males, whites had higher energy intakes than did nonwhites. Injection drug users consumed less energy than did nonusers. Among nondieting females, only the absence of nausea and vomiting was marginally associated with higher energy intakes. Inadequate energy intake, which occurred in 38% of this population, was independently associated with female sex among nondieters. A significant proportion of the study cohort (52%) was consuming less than the recommended dietary allowance of vitamin A. Inadequate protein intake, found in 11% of the study population, occurred more often in females, those without a caregiving adult in the household, and individuals with reduced appetite. A considerable proportion of the participants (23%) reported that they were dieting to lose weight.

Conclusions: Dietary inadequacy was strongly correlated with being in the sociodemographic groups that are at heightened risk of adverse clinical outcomes. It may be worthwhile to study dietary intake as a potential determinant of the clinical outcomes of HIV infection.

Key Words: Diet • sociodemographics • AIDS • HIV • human immunodeficiency virus • wasting • nutrition • socioeconomics

	INTRODUCTION

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Since the AIDS epidemic began, the prevalence of HIV infection and AIDS has increased steadily in all demographic groups, yet the epidemiologic data indicate that minorities and lower-income populations are disproportionately affected by HIV disease (1–5). Lower-income populations have had higher rates of opportunistic infections and admissions to intensive care units and decreased survival times (6, 7). HIV wasting syndrome was found to occur more frequently in nonwhites and injection drug users than in whites and nonusers (8). Although unequal access to health care was considered the primary cause of the poorer prognosis of HIV disease in these populations, other causes related to sociodemographic attributes may be responsible for differential clinical outcomes (9). The social and biological mechanisms underlying the unfavorable epidemiologic profile in these communities have not been studied extensively.

Nutritional intake is an often-overlooked factor in the progression of HIV disease, although the relation between nutrition and immune function is well established. Even in populations not infected with HIV, poor nutritional status is known to impair immune response (10, 11). In HIV-infected individuals, poor nutritional status is a strong predictor of survival, even after controlling for CD4⁺ cell counts (12, 13): a weight loss of >66% of ideal body weight was linked to the timing of death in AIDS patients (14). It was noted that HIV wasting syndrome was associated with decreased serum concentrations of vitamin A, folate, and carotene (11). Nutritional supplementation was shown to improve some of the immunologic symptoms associated with HIV disease (15, 16) and to increase immune function in healthy populations (17, 18). Given the evidence that malnutrition is a determinant of the development of AIDS and other adverse health outcomes (19), it is worthwhile to examine dietary adequacy in HIV-infected adults.

One objective of this study was to determine the correlates of reduced and inadequate nutrient intakes in a large, clinically and socioeconomically diverse cohort of HIV-infected adults. Previous nutritional studies focused on patients in the late stages of AIDS (14, 20, 21) or gay men who were socioeconomically advantaged relative to other HIV-infected groups (19, 22). Because a large number of individuals in this study cohort were in the early stages of their disease, we were able to examine the relations between diet and sociodemographic factors at a stage when eating habits are less affected by HIV-related symptoms.

We sought to determine the extent to which dietary inadequacy is correlated with clinical symptoms, economic inability to procure food, and lack of health awareness. We were also interested in exploring the effects of support for food acquisition and preparation as correlates of nutritional adequacy in HIV-infected adults. The findings may help health care providers develop nutritional guidelines and intervention strategies for patients at high risk of developing AIDS.

	SUBJECTS AND METHODS

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Subjects and study design
Nutrition for Healthy Living (NFHL) is a longitudinal study of wasting in HIV disease. From its inception in February 1995, this study has enrolled 679 HIV-positive participants from the greater Boston and Providence, RI, areas. To improve the generalizability of the study results to the HIV-infected population of the United States, the project sought to enroll large numbers of women, injection drug users, and people of lower socioeconomic status (SES) through recruitment practices and retention strategies that were sensitive to the needs of these individuals.

The study excluded potential participants at screening if they reported diabetes mellitus, current pregnancy, thyroid disease, active malignancy not related to HIV, an intent to move out of the area within 3 y, age <18 y, or inability to be reached by telephone. Potential participants were also excluded if the staff determined that their English language proficiency was insufficient.

Data in this report were obtained from 2 baseline clinic visits per participant. These visits consisted of a physical examination, anthropometric measurements, and administration of several questionnaires on clinical status, physical activity, alternative health treatments, and quality of life. The details of the study design are outlined elsewhere (23, 24).

Measurement of dietary intake
At the first baseline visit, each participant was given a 3-d diet record that was to be returned at the second baseline visit. Three days of diet records were requested to obtain intake data for multiple days while minimizing the noncompliance associated with asking participants to keep diet records for longer periods of time. The study staff carefully instructed each participant on how to keep accurate diet records. The NFHL dietitian checked these diet records when they were collected at the second visit. Specially trained personnel entered the data from the diet records into the database. Nutrient calculations were performed by using the NUTRIENT DATA SYSTEM FOR RESEARCH software (version 4.0) developed by the Nutrition Coordinating Center, University of Minnesota, Minneapolis (Food and Nutrient Database 28) (25). If an analytic value was not available for a nutrient in a certain food, the value was calculated on the basis of the nutrient contents of the ingredients in the same food or the nutrient contents of similar foods (25).

We measured dietary intakes as absolute intakes (in kJ, g, or µg) and also as percentages of the US recommended dietary allowances (RDAs) (26). The RDAs published in 1989 rather than the updated RDAs were used to allow comparisons with the nutritional adequacy of other groups studied during the same time period. We included supplement use in all nutrient calculations because total nutrient intake was the variable of interest. We also determined the adequacy of energy intake by comparing each participant's energy intake with the energy needed to sustain minimal physical activity. Each participant's resting energy expenditure (REE) was measured with indirect calorimetry; we determined the participant's oxygen consumption and carbon dioxide production (27) for a duration of 20 min after 5 h of fasting. The indirect calorimetry module of the Sensormedic Vmax series spirometer (model 2130; Sensormedic Corp, Yorba Linda, CA) was used. REE was calculated by using the abbreviated Weir equation below (28):

(1)

where O₂ (oxygen uptake) and CO₂ (carbon dioxide production) were measured in mL/min and REE was subsequently converted to kJ and MJ for statistical analysis.

We multiplied each participant's REE by an activity factor of 1.3 to represent a minimal (not bedridden) amount of physical activity (26). We chose this activity factor, which represents 10 h/d of rest and 14 h/d of very light activity (such as playing cards or watching television), to obtain conservative estimates of energy requirements, disregarding higher amounts of physical activity. This method of determining total energy expenditure was used instead of the conventional Harris-Benedict equation because the latter is derived from and thereby applicable to healthy persons with metabolisms that are not altered by underlying diseases such as HIV infection (29, 30).

Statistical analyses
To determine the independent correlates of absolute protein and energy intakes, we conducted multivariate linear regression analyses including all variables that were associated with a P < 0.20 in the unadjusted analyses (31). We examined the following variables in unadjusted analyses: age, CD4⁺ cell count, nonwhite race, gay male lifestyle, history of injection drug use (compared with never using injection drugs), AIDS diagnosis, presence of nausea or vomiting, presence of a caregiving adult, presence of dependent children in the household, presence of diarrhea, presence of decreased appetite, educational attainment (an ordinal variable), household income (an ordinal variable), use of protease inhibitors or antiretroviral drugs, and dieting status (currently trying to lose weight or not). Individuals who were dieting to lose weight were excluded from the multivariate analyses.

We controlled for body size by including height instead of weight in the regression model of energy intake. This was done because many of the participants suffered from HIV wasting syndrome, which itself could lead to changes in energy intake. Preliminary analyses indicated that height was strongly and significantly correlated with energy intake, even though weight and body mass index (BMI; in kg/m²) were not. We included alcohol consumption in all analyses of energy intake. The preliminary analysis of alcohol intake indicated that the median percentage of energy from alcohol was <1% for this cohort with an interquartile range of 0–0.3%. Alcohol contributed >15% of energy intake in only 2.5% of the participants. Moreover, alcohol intake was not significantly associated with the adequacy of energy intake; hence, we did not adjust for alcohol consumption in our analyses.

We present the results of the multivariate regression analyses along with the mean unadjusted intakes for each subgroup considered. We adjusted for energy intake in all regression analyses of protein intake to address the composition of the participants' diets, independent of absolute intake. We conducted influence diagnostics of the final results to ensure that the results were not unduly influenced by a few outlying measurements (32). The final regression coefficients were obtained after eliminating participants whose dfß coefficients (33) for the covariates lay beyond 3 interquartile ranges from the median. Removal of influential points yielded more conservative models with fewer parameters that were statistically significant.

Correlates of dietary inadequacy (ie, intakes less than the RDA) were identified with preliminary unadjusted analyses by using linear models with a log link function and binomial distribution for the prevalence response variable (34). All variables that showed a P value <0.20 in the unadjusted analyses for energy, protein, vitamin A, and zinc intakes were included in a multivariate model (full model) (35). Any parameters with P values <0.20 were remodeled in a reduced model. In this article, we report the prevalence ratios and 95% CIs for each of the independent correlates of dietary inadequacy on the basis of binomial regression models with the log link function (35). All statistical analyses were conducted with SAS software, version 6.12 (36). P < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 
The characteristics of the NFHL Study population at baseline are shown in Table 1. The cohort included relatively large numbers of nonwhites (37.9%), females (26.9%), and persons who had not been diagnosed with AIDS (44%). There were several statistically significant differences between the male and female participants. Men tended to be slightly older and had a higher household income and higher educational attainment than did women. The largest racial category was white (70%) for men but was African American (42.9%) for women. Female participants were more likely to be living with children in their household (46.8% compared with 9.3% of men). Men tended to have more advanced disease, as shown by their lower mean CD4⁺ cell counts and larger proportions of individuals with AIDS, diarrhea, and HIV wasting syndrome.

Certain clinical symptoms of HIV disease may interfere with adequate food consumption. Of these symptoms, reduced appetite and nausea and vomiting were relatively prevalent in this cohort, whereas diarrhea was reported by only 7.1% of the participants at baseline. Approximately 1 of 7 participants had been diagnosed with HIV wasting syndrome. A questionnaire item asked whether the participant had been clinically diagnosed with wasting related to HIV infection. When classifying participants as having or not having HIV wasting syndrome, we also considered self-reported weight changes in the 3 y before study enrollment. An involuntary weight loss of >10% of initial body weight or a sustained loss of >5% of body weight within the previous 6 mo was used as the definition of HIV wasting syndrome. About 25% of the participants reported a lack of assistance with food shopping and a similar proportion reported lack of assistance with food preparation. Almost 72% of the participants were using either a protease inhibitor or an antiretroviral drug at enrollment.

The variables that were correlated with energy and protein intakes at a significance of P < 0.20 in the unadjusted analysis are shown in Table 2. For each of these variables, the mean energy and protein intakes are shown. Because dieting was strongly associated with energy and protein intakes in the preliminary analyses, we eliminated all dieters from the multivariate analyses to uncover other relations. The P values shown for energy intake were obtained from models that adjusted for height to control for body size, and the P values for protein intake reflect adjustment for energy consumption to reduce the extraneous variation resulting from differences in energy intake. We also give the full multivariate adjusted results for all variables that had a P < 0.20 in the bivariate analyses.

Multivariate analyses of energy-adjusted fat and carbohydrate intakes indicated that sociodemographic parameters were the primary correlates of these macronutrient intakes. For males, history of injection drug use, low CD4⁺ cell count, and the presence of children in the household were significantly associated with higher fat intakes. For females, the presence of children in the household was the only significant correlate of higher fat intake. For both men and women, the presence of children in the household was the only significant correlate of lower carbohydrate intake.

The correlates of inadequate intakes of energy, protein, niacin, folate, iron, zinc, and vitamins A, B-6, B-12, C, and E are shown in Tables 3 and 4. Intakes were considered inadequate if they were below the RDA. Energy intake was inadequate in 37.5% of the participants. Inadequate intakes of zinc, protein, and vitamin A were found in 41.6%, 11.0%, and 51.6% of participants, respectively.

Participants who were dieting to lose weight had consistently lower nutrient intakes than did nondieters. Of particular concern, dieting was significantly associated with a higher percentage of individuals not reaching the RDAs for vitamin A (64.2% compared with 47.7% of nondieters) and zinc (54.1% compared with 39.3% of nondieters).

We also compared the NFHL cohort to the general population of the northeastern United States regarding adequacy of nutrient intakes. For all the nutrients examined, a lower percentage of the NFHL cohort had inadequate intakes compared with the general population. This was found despite the fact that lower-SES participants were less successful at obtaining the RDAs for micronutrients than were higher-SES participants in the NFHL Study.

Of the clinical symptoms examined in this analysis (Table 4), only reduced appetite was consistently and significantly associated with inadequate dietary intakes in high percentages of participants. Neither CD4⁺ cell count nor the presence of the other clinical symptoms showed significant associations with any of the nutrient intakes studied. Although the difference was not significant, participants with an AIDS diagnosis were more successful at meeting their RDAs than were their counterparts without AIDS. Use of protease inhibitors or antiretroviral drugs was associated with higher nutrient intakes. Nevertheless, the use of these drugs was not associated with dietary adequacy in any of the multivariate models, with the exception of vitamin C intake.

The multivariate analyses of the correlates of inadequate energy, protein, vitamin A, and zinc intakes are shown in Table 5. A preliminary analysis showed that the only independent correlate of inadequate energy intake was dieting to lose weight. Likewise, dieting was independently associated with inadequate vitamin A and zinc intakes. When the multivariate analysis was rerun after excluding all participants who were dieting to lose weight, female sex was the only variable that entered the regression model for inadequate energy intake (prevalence ratio = 1.6; 95% CI: 1.1, 2.2). Inadequate protein intake was associated with lower educational attainment, absence of a caregiving adult in the household, and reduced appetite. Nonwhite race and younger age were the only independent correlates of inadequate vitamin A intake. Inadequate zinc consumption was independently associated with reduced appetite, lower educational attainment, the absence of an adult caregiver, and the presence of dependent children in the household. Adjusting for supplement use did not appreciably alter the results of the multivariate analyses.

	DISCUSSION

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

Our analysis has several limitations. Because it was not feasible to conduct an exhaustive analysis of every nutrient in the human diet, we concentrated on nutrients known to affect immune function or known to be low in the serum of HIV-infected persons. Hence, we could have missed many important associations. The validity of self-reported dietary data, particularly data from less educated and obese individuals, has also been questioned (40). Nevertheless, diet records are the most accurate method of dietary analysis for studies that include many subjects of ethnic minorities; their intakes may not be reflected accurately by food-frequency questionnaires (41). Despite the day-to-day variations in food consumption, we believe that these data are an accurate representation of the group mean intakes because of the large sample size (41, 42) and because we eliminated influential outliers from the regression analyses. A correlation previously found between dietary inadequacy and the presence of hunger in this cohort also supports the general validity of the dietary record data (23).

Finally, we cannot fully rule out the possibility of differential dietary underreporting. Previous studies reported conflicting results regarding the relations between SES and dietary underreporting (43–47). However, dieting and obesity have been consistently associated with dietary underreporting in the literature (43–46, 48). Because dieters were eliminated from the multivariate analyses altogether, dieting behavior could not have contributed to differential reporting of food intake in the data used in the multivariate analyses. Of the remaining participants, only 8% were clinically obese.

In the analysis of macronutrient intakes, many variables showed opposite trends for men and women. We speculate that these differences are attributable to greater health-consciousness in the male participants, who were better educated than their female counterparts. As their illness progresses, male participants appear to engage in compensatory eating behaviors. A previously conducted study of this cohort indicated that 55% of the men used vitamin supplements, whereas only 29% of the women took vitamins regularly (24). The female participants may not have the financial means, health consciousness, or social support to eat well as their health declines.

Although more members of this HIV-infected cohort met the RDAs for the various nutrients studied than in the general population of the northeastern United States (39), we cannot state unequivocally that dietary needs were better met in our study cohort. The burden of the disease would presumably necessitate increased nutrient intakes (49, 50), and therefore a direct comparison with the general population's dietary intakes would be unwise. In addition, differences in methodology, population composition, or both may account for the lower percentages of individuals with dietary inadequacy in our cohort. Because an HIV-negative control group was not included in our study, it is difficult to draw definitive conclusions.

The results of the multivariate analysis suggest that inability to meet the RDAs is largely a socioeconomic rather than a clinical phenomenon, because the only clinical symptom consistently associated with inability to meet the RDA was reduced appetite. Individuals with an AIDS diagnosis did not fare worse than participants who were still asymptomatic. Minorities, participants with dependent children, individuals without an adult caregiver, and those without food shopping assistance consistently had less adequate dietary intakes. Despite the noted side effects of protease inhibitors and antiretroviral drugs, we found that participants using these drugs were more successful at meeting their RDAs. Nevertheless, the use of these drugs was not associated with dietary adequacy in any of the multivariate models, except in the case of vitamin C intake. Hence, it can be assumed that the use of these medications (which is strongly correlated with higher educational attainment and having private insurance) is an indicator of SES that does not have independent effects on nutrient intakes once SES is accounted for. Likewise, the use of supplements, which is strongly associated with micronutrient adequacy, was also shown in a preliminary analysis to be significantly correlated with educational attainment.

Our data provide evidence that BMI can vary greatly in cohorts of individuals with HIV disease: >20% of the women in this study were obese, and 50% were overweight. We found that dieting behavior is correlated with high BMI. Of the obese participants, 47% said they were dieting, whereas 34% of individuals with a BMI between 25 and 30 were dieting. Only 13% of those with a BMI between 20 and 25 were voluntarily restricting their food intake. It is possible that overweight HIV-infected individuals are not aware of the potentially detrimental effects of dieting on their long-term prognosis, or that the social pressure to be thin prevails over their health concerns. No participant who was wasting reported attempts to lose weight, suggesting that dieting behavior is curtailed once the health effects of caloric restriction become observable.

BMI was inversely associated with risk of death on the basis of preliminary analyses of data from the 679 participants, 39 of whom have died since enrollment. Individuals with a baseline BMI >25 had a much lower risk of death than did those with a BMI <25 (OR = 0.4; 95% CI: 0.2, 0.8).

The behavioral and attitudinal aspects of dietary intake among HIV-infected adults have been largely overlooked by researchers. It may have been presumed that inadequate dietary intake is primarily involuntary, resulting from limited resources or clinical symptoms. Our results suggest that in HIV disease, dietary intake is a complex sociobehavioral phenomenon that reflects the confluence of attitudinal, economic, and lifestyle factors. Hence, clinicians and health care professionals should consider sociobehavioral models of dietary modification when designing intervention strategies for HIV-infected individuals. Public health practitioners should strive to understand the reasons underlying food choices, food procurement, and eating practices in this population, which includes disproportionate numbers of individuals who are drug abusers, indigent, disabled, and homeless. Given their limited resources, it is unclear how these individuals prioritize food procurement amid competing costs such as medical treatment, childcare, illegal drugs, and housing-related expenses. Future research should focus on longitudinal changes in food consumption patterns and the effects of dietary counseling and programmatic aid on nutritional intake.

	REFERENCES

TOP ABSTRACT INTRODUCTION SUBJECTS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Friedman SR, Chapman TF, Perlis TE, et al. Similarities and differences by race/ethnicity in changes of HIV seroprevalence and related behaviors among drug injectors in New York City. J Acquir Immune Defic Syndr 1999;22:88–91.
2. Murrain M, Barker T. Investigating the relationship between economic status and HIV risk. J Health Care Poor Underserved 1997; 8:416–23.[Medline]
3. Hankins C, Tran T, Hum L, et al. Socioeconomic geographical links to human immunodeficiency virus seroprevalence among childbearing women in Montreal, 1989–1993. Int J Epidemiol 1998;27:691–7.[Abstract/Free Full Text]
4. Centers for Disease Control (CDC). Advance report of final mortality statistics, 1993. Mon Vital Stat Rep 1995;43:1–9, 26–29.
5. Centers for Disease Control (CDC). Update: acquired immunodeficiency syndrome—United States, 1994. MMWR Morb Mortal Wkly Rep 1995;44:64–7.[Medline]
6. Hogg RS, Strathdee SA, Craib KJ, O'Shaughnessy MV, Montaner JSG, Schechter MT. Lower socioeconomic status and shorter survival following HIV infection. Lancet 1994;344:1120–4.[Medline]
7. Easterbrook PJ, Keruly JC, Creagh-Kirk T, Richman DD, Chaisson RE, Moore RD. Racial and ethnic differences in outcome in zidovudine-treated patients with advanced HIV disease. JAMA 1991;266: 2713–8.[Abstract]
8. Nahlen BL, Chu SY, Nwanyanwu OG, Berkelman RL, Martinez SA, Rullman JV. HIV wasting syndrome in the United States. AIDS 1993;7:183–8.[Medline]
9. Schechter MT, Hogg RS, Aylward B, Craib KJ, Le TN, Montaner JS. Higher socioeconomic status is associated with slower progression of HIV infection independent of access to health care. J Clin Epidemiol 1994;47:59–67.[Medline]
10. Cunningham-Rundles S. Effects of nutritional status on immunological function. Am J Clin Nutr 1982;35:1202–10.[Free Full Text]
11. Coodley GO, Coodley MK, Nelson HD, Loveless MO. Micronutrient concentrations in the HIV wasting syndrome. AIDS 1993;7:1595–600.[Medline]
12. Suttman U, Ockenga J, Selberg O, Hoogestraat L, Helmuth D, Muller MJ. Incidence and prognostic value of malnutrition and wasting in human immunodefiency virus-infected outpatients. J Acquir Immune Defic Syndr Hum Retrovirol 1995;8:239–46.[Medline]
13. Semba RD, Caiaffa WT, Graham NM, Cohn S, Vlahov D. Vitamin A deficiency and wasting as predictors of mortality in human immunodeficiency virus-infected injection drug users. J Infect Dis 1995;171:1196–202.[Medline]
14. Kotler DP, Tierney AR, Wang J, Pierson RN Jr. Magnitude of body-cell-mass depletion and the timing of death from wasting in AIDS. Am J Clin Nutr 1989;50:444–7.[Abstract/Free Full Text]
15. Baum MK, Shor-Posner G, Lu Y, et al. Micronutrients and HIV-1 disease progression. AIDS 1995;9:1051–6.[Medline]
16. Tang AM, Graham NM, Kirby AJ, McCall LD, Willett WC, Saab AJ. Dietary micronutrient intake and risk of progression to AIDS in HIV-infected homosexual men. Am J Epidemiol 1993;138:937–51.[Abstract/Free Full Text]
17. Meydani SN, Meydani M, Blumberg JB, et al. Vitamin E supplementation and in vivo immune response in healthy elderly subjects. A randomized controlled trial. JAMA 1997;277:1380–6.[Abstract]
18. Beharka A, Redica S, Leka L, Meydani SN. Vitamin E status and immune function. Methods Enzymol 1997;282:247–63.[Medline]
19. Abrams B, Duncan D, Hertz-Picciotto I. A prospective study of dietary intake and acquired immune deficiency syndrome in HIV-seropositive homosexual men. J Acquir Immune Defic Syndr 1993;6:949–58.
20. Singer P, Rothkopf MM, Kvetan V, Kirevela O, Gaare J, Askanazi J. Risks and benefits of home parenteral nutrition in the acquired immunodeficiency syndrome. JPEN J Parenter Enteral Nutr 1991;15:75–9.[Abstract]
21. Kotler DP, Tierney AR, Culpepper-Morgan JA, Wang J, Pierson RN Jr. Effects of home total parenteral nutrition upon body composition in patients with acquired immunodeficiency syndrome. JPEN J Parenter Enteral Nutr 1990;14:454–8.[Abstract]
22. Hogg RS, Zadra JN, Chan-Yan C, et al. Analysis of nutritional intake in a cohort of homosexual men. J Acquir Immune Defic Syndr Hum Retrovirol 1995;9:162–7.[Medline]
23. Kim JH. The correlates of hunger and reduced dietary intake in a cohort of HIV-positive adults of the Boston area: a cross-sectional study with analyses of data quality and post-recruitment retention rates. PhD dissertation. Harvard School of Public Health, Boston, 2001.
24. Woods MN, Knox TA, Speigelman D, et al. Dietary intake in a large HIV cohort that includes women and minorities. J Am Diet Assoc (in press).
25. Regents of the University of Minnesota, Nutrition Coordinating Center. NDS-R software 8, version 4.0. Minneapolis: University of Minnesota, 1998.
26. National Research Council. Recommended dietary allowances. 10th ed. Washington, DC: National Academy Press, 1989.
27. Matarese L. Indirect calorimetry: technical aspect. J Am Diet Assoc 1997;97(suppl):S154–60.[Medline]
28. Feurer I, Mullen JL. Bedside measurement of resting energy expenditure and respiratory quotient via indirect calorimetry. Nutr Clin Pract 1986;1:43–9.
29. Harris JA, Benedict FG. Standard basal metabolism constants for physiologists and clinicians: a biometric study of basal metabolism in man. Washington, DC: Carnegie Institute of Washington, 1919.
30. Frankenfield DC, Muth ER, Rowe WA. The Harris-Benedict studies of human basal metabolism: history and limitations. J Am Diet Assoc 1998;98:439–45.[Medline]
31. Dales LG, Ury HK. An improper use of statistical significance testing in studying covariables. Int J Epidemiol 1978;4:373–5.
32. Belsley DA, Kuh E, Welsch RE. Regression diagnostics. New York: John Wiley & Sons, 1980.
33. Neter J, Wasserman W, Kutner MH. Applied linear statistical models: regression, analysis of variance, and experimental designs. 3rd ed. Homewood, IL: Richard D Irwin Inc, 1990.
34. Wacholder S. Binomial regression in GLIM: estimating risk ratios and risk difference. Am J Epidemiol 1986;123:174–84.[Abstract/Free Full Text]
35. Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol 1989;129:125–37.[Abstract/Free Full Text]
36. SAS Institute Inc. SAS/STAT7 software: changes and enhancements through release 6.12. Cary, NC: SAS Institute Inc, 1997.
37. Committee on Diet and Health, National Research Council. Diet and health: implications for reducing chronic disease risk. Washington, DC: National Academy Press, 1989:114.
38. Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL. Overweight and obesity in the United States: prevalence and trends, 1960–1994. Int J Obes Relat Metab Disord 1998;22:39–47.[Medline]
39. Beltsville Agricultural Research Center. Food and nutrient intakes by region, 1994–96, Table set 13, Table 3A. Nutrient intakes: percentages of individuals with diets below selected levels of the 1989 recommended dietary allowances (RDAs), by region, 2-day average, 1994–96. Internet: http://www.barc.usda.gov/bhnrc/foodsurvey/pdf/region.pdf (accessed 25 December 2000).
40. Klesges RC, Eck LH, Ray JW. Who underreports dietary intake in a dietary recall? Evidence from the second National Health and Nutrition Examination Survey. J Consult Clin Psychol 1995;63:438–44.[Medline]
41. Willett W. Nutritional epidemiology. New York: Oxford University Press, 1990.
42. Howat PM, Mohan R, Champagne C, Monlezun C, Wozniak P, Bray GA. Validity and reliability of reported dietary intake data. J Am Diet Assoc 1994;94:169–73.[Medline]
43. Price GM, Paul AA, Cole TJ, Wadsworth ME. Characteristics of the low-energy reporters in a longitudinal national dietary survey. Br J Nutr 1997;77:833–51.[Medline]
44. Voss S, Kroke A, Klipstein-Grobusch K, Boeing H. Is macronutrient composition of dietary intake data affected by underreporting? Results from the EPIC-Potsdam Study. European Prospective Investigation into Cancer and Nutrition. Eur J Clin Nutr 1998;52:119–26.[Medline]
45. Johnson RK, Soultanakis RP, Matthews DE. Literacy and body fatness are associated with underreporting of energy intake in US low-income women using the multiple-pass 24 hour recall: a doubly labeled water study. J Am Diet Assoc 1998;98:1136–40.[Medline]
46. Lafay L, Basdevant A, Charles MA, et al. Determinants and nature of dietary underreporting in a free-living population: the Fleurbaix Laventie Ville Sante (FLVS) Study. Int J Obes Relat Metab Disord 1997;21:567–73.[Medline]
47. Hirvonen T, Mannisto S, Roos E, Pietinen P. Increasing prevalence of underreporting does not necessarily distort dietary surveys. Eur J Clin Nutr 1997;51:297–301.[Medline]
48. Little P, Barnett J, Margetts B, et al. The validity of dietary assessment in general practice. J Epidemiol Community Health 1999;53:165–7.[Abstract]
49. Coodley GO, Loveless MO, Merrill T. The HIV wasting syndrome: a review. J Acquir Immune Defic Syndr 1994;7:681–94.
50. Gorbach S, Knox T. Weight loss and human immunodeficiency virus infection: cachexia versus malnutrition. Infect Dis Clin Pract 1992;1:224–9.

This article has been cited by other articles:

				K. M Hendricks, D M. Mwamburi, P. Newby, and C. A Wanke Dietary patterns and health and nutrition outcomes in men living with HIV infection Am. J. Clinical Nutrition, December 1, 2008; 88(6): 1584 - 1592. [Abstract] [Full Text] [PDF]

				L. Normen, K. Chan, P. Braitstein, A. Anema, G. Bondy, J. S. G. Montaner, and R. S. Hogg Food Insecurity and Hunger Are Prevalent among HIV-Positive Individuals in British Columbia, Canada J. Nutr., April 1, 2005; 135(4): 820 - 825. [Abstract] [Full Text] [PDF]

				K. M Hendricks, K. R Dong, A. M Tang, B. Ding, D. Spiegelman, M. N Woods, and C. A Wanke High-fiber diet in HIV-positive men is associated with lower risk of developing fat deposition Am. J. Clinical Nutrition, October 1, 2003; 78(4): 790 - 795. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kim, J. H Articles by Gorbach, S. L Search for Related Content PubMed PubMed Citation Articles by Kim, J. H Articles by Gorbach, S. L Agricola Articles by Kim, J. H Articles by Gorbach, S. L