Springer 2005

Quality of Life Research (2005) 14: 981–990 DOI 10.1007/s11136-004-2580-2

The relationship between lipodystrophy-associated body changes and measures of quality of life and mental health for HIV-positive adults Robert Burgoyne1,2, Evan Collins1,2, Cheryl Wagner3, Susan Abbey1,2, Mark Halman2,4, Margaret Nur1 & Sharon Walmsley1,5 1 Toronto General Hospital (E-mail: [email protected]); 2Department of Psychiatry, University of Toronto; 3HIV primary care physician; 4St. Michael’s Hospital; 5Department of Medicine, University of Toronto, Toronto, Canada Accepted in revised form 26 August 2004

Abstract Objectives: To investigate the relationship between lipodystrophy-specific symptom severity and wellbeing. Methods: HIV-positive adult patients with body fat redistribution (lipodystrophy syndrome) associated with antiretroviral therapy reported their total non-lipodystrophy symptoms and side effects and completed measures assessing body fat changes (yielding Atrophy, Hypertrophy and Total Lipodystrophy scores), mental health and quality of life. Effects of total symptom complex and lipodystrophy severity on quality of life and mental health were analyzed using Spearman’s rho correlations. Logistic regression analyses were utilized to determine the relative-odds of depression produced by overall symptom count and lipodystrophy score increments. Results: Mean ratings for Hypertrophy and Atrophy corresponded to ‘very mild’ and ‘mild’ degrees of severity, respectively. The total symptom complex was associated with ratings for most of the mental health and quality of life measures. Patient-perceived body image scores were the sole study variable responsive to lipodystrophy severity ratings. In comparison to reference norms, a pronounced degree of body image impairment was evident. Conclusion: Although responsive to the total symptom profile, psychosocial measures typically utilized for evaluating quality of life and mental health status in HIV disease lacked sensitivity and specificity for measuring the consequences of lipodystrophy-associated fat distribution changes alone. Lipodystrophy severity did impact negatively on body image. Key words: Body image, HIV, Lipodystrophy, Mental health, Quality of life

Introduction HIV-related symptoms attributable to both the underlying disease and treatment-associated adverse events affect a wide range of components of wellbeing, as documented in previous research [1– 10]. The use of highly active antiretroviral therapy (HAART) has more recently been associated with body fat redistribution and metabolic abnormalities, a phenomenon known as the lipodystrophy syndrome [11, 12]. The morphologic changes of lipodystrophy include atrophy (localized fat loss typically in the face, buttocks and extremities) and/ or hypertrophy (localized fat gain typically in the

breasts, abdomen and nape of the neck). These changes have gained attention in the care of people living with HIV/AIDS (PHA) and in the literature, with increasing concerns raised about the potential negative psychological consequences and deterioration in quality of life posed by lipodystrophy. The distinctive body shape changes are particularly important given their disfiguring potential and the possibility of compromising HIV serostatus privacy [13], resulting in disappointment and skepticism counterbalancing the promise of HAART for improving quality and quantity of life [14, 15]. From the time lipodystrophy was initially recognized, clinicians worried it would lead to depression

982 and other psychosocial and mental health issues, compounding other HAART-related adverse effects potentially compromising its optimal use [16, 17]. Efforts were initiated to measure the effects of lipodystrophy on psychosocial wellbeing. Research employing open-ended survey techniques, singleitem questions thematically associated with body changes, or other grounded qualitative study methods corroborated the initial apprehensions concerning lipodystrophy [18–22]. The finding of one study suggested that many PHA would trade off length of life or accept greater risk of mortality in order to maintain health status free of lipodystrophy-related body alterations [23]. However, the one published study in which a standardized quality of life measure was employed did not find significant differences in ratings of patients with clinically defined lipodystrophy compared to those without lipodystrophy [24]. The latter finding raises questions about whether standardized instruments are capable of detecting the effects of lipodystrophyrelated body changes among PHA. Are quality of life and mental health measures already widely in use sufficiently sensitive and conceptually specific for capturing the impact of body changes? What can be learned about various measures’ discriminant validity, that is, ability to distinguish among different degrees of severity of body alterations thought to be attributable to lipodystrophy? The purpose of our study was to systematically examine the relationship of physically visible manifestations of lipodystrophy experienced by PHA to a broad range of standardized measures of quality of life and mental health.

Methods Study population and recruitment Adult HIV-infected patients on HAART regimens and who had self-identified in discussion with their treating physician their perception of fat wasting and/or fat accumulation, that is, at least one lipodystrophy-associated body fat change of any severity were recruited from a university hospital HIV specialty clinic or HIV primary care physician practice in an urban setting in Toronto, Ontario, Canada. The sample selection was based on both patient and physician willingness to participate.

Physicians were made aware of the study objectives through regular channels of local network communication in relation to HIV, were the source of referrals to the research project, and provided adjunct assessment of lipodystrophy severity. Patients with mild fat gain in the abdominal area reported as the sole body alteration were excluded from participation due to the potential confounding of normally occurring age-weight gain leading to a ‘false-positive’ of hypertrophy. Measurement of lipodystrophy and overall symptom profile The lipodystrophy measurement scale utilized for assessing body changes perceived by study patients to have occurred since starting HAART was adapted from a lipodystrophy questionnaire specific to morphologic alterations as conceptualized at the time of the initiation of our research [25], and corroborated in previous qualitative research lead by key members of our investigative team [18]. Additional methods incorporated to capture the relevant aspects of body changes and their consequences to wellbeing included in-depth qualitative clinician-investigator interview with each patient. The full extent of results of our triangulated methodology (e.g. qualitative findings) is not herein reported. The lipodystrophy severity scale consisted of eight single-focus items corresponding to four body fat alteration or redistribution features of lipoatrophy (i.e. fat loss, wasting) in the face, arms, legs and buttocks; and four features of lipohypertrophy (i.e. fat gain) in the breasts, abdomen, back of the neck, and under the chin. The latter feature was added due to our previous experience of neck enlargement prevalence far outweighing that of lipomata (discrete fat masses on the torso or elsewhere). Each of the eight items was rated by subjects on a 5-point scale of severity: 0 ¼ none; 1 ¼ very mild; 2 ¼ mild; 3 ¼ moderate; 4 ¼ severe. Patients were asked to rate each item according to their perception of the degree of change since commencing antiretroviral medication treatment, that is, the magnitude of body change in each of the eight body areas according to their view of what was believed by them to be lipodystrophy-related. They were advised not to incorporate perceptions of body changes they thought to be the product of other

983 factors such as those related to aging, lifestyle, or adjustments in diet or exercise. Irrespective of attempts on the part of study patients alone or in concert with physician consultation to devise strategic plans to reduce onset or otherwise mitigate the emergence and progressive development of lipodystrophic body alterations, they were asked to rate their current perception of lipodystrophy-related change at each body site in question. Adding the raw scores of the scale items yielded non-weighted domain scores for Hypertrophy and Atrophy and a Total Lipodystrophy score with a possible range from 0 to 32. Thus, the Total Lipodystrophy score reflected a composite of body site range and severity of alterations. Based on the limited research and considerations regarding the most cogent way to present our study results, we assumed that combining both the type and site of body changes would most likely reflect the cumulative effects of the varied manifestations of lipodystrophy on wellbeing, and be most likely to yield data meeting the assumptions of normal distribution. Following written informed consent and hospital/university research ethics board authorization, 77 participating patients completed the lipodystrophy instrument and a checklist, adapted from previously developed symptom inventory assessments [26], of 15 other commonly reported HIVrelated non-lipodystrophy physical symptoms and/ or treatment side effects experienced consistently during the previous month. The degree of severity of non-lipodystrophy symptoms was not rated by patients. The HIV treating physicians independently assessed the degree of body fat alterations using the same eight-item lipodystrophy scale. An Intraclass Correlation Coefficient was calculated to assess physician–patient inter–rater reliability for Total Lipodystrophy severity ratings. Health related quality of life and mental health measures A battery of standardized quality of life and mental health measures was administered to the patients during the same measurement and interview period in which lipodystrophy was rated. The three quality of life measures utilized were the 34item Medical Outcomes Study HIV Health Survey (MOS-HIV) [10, 27], with ten subscale ratings ranging from 0 to 100; the 34-item HIV/AIDS-

Targeted Quality of Life (HAT-QoL) Instrument [28], with nine subscale ratings from 0 to 100; and the HIV Overview of Problems – Evaluation System (HOPES) [29] Physical Summary scale, Psychosocial Summary scale, Sexual Summary scale, and a distinctive Body Image subscale within the Psychosocial domain, all ratings ranging from 0 to 4. Our selection of quality of life measures from the wide array of well-being instruments available was based on considerations of the value of a broad-based multidimensional profile enabling exploration of a full range of areas relevant to our research questions; the desire to examine the consistency of any effects of symptom profile and lipodystrophy across dimensions of physical, emotional and social wellbeing; and a mix of measurement strategies including adaptation from generic instruments (i.e. MOS-HIV), instrument adaptation based on the views of health professionals (i.e. HOPES, derived from cancer research and applied to HIV), and a patient-centered approach with items derived from PHA focus groups (i.e. HAT-QoL). The quality of life measures are included in a review of instruments used in HIVrelated research but were noted, with the exception of the HOPES, to make no direct reference to body appearance [30]. Therefore, an additional measure aimed specifically at self-reported body perceptions was included: the Body Image subscale of the Derogatis Sexual Function Inventory [31], with scores ranging from 0 to 60. The mental health instruments completed by study participants were the ‘state’ anxiety component of the Spielberger State-Trait Anxiety Inventory (STAI) [32], with scores ranging from 10 to 40; the Rosenberg Self-Esteem Scale (RSES) [33], with scores ranging from 0 to 40; and the Beck Depression Inventory, 2nd edition (BDI-II) [34], with scores ranging from 0 to 60. The psychological dimensions inherent in these three scales allowed for a more fully comprehensive coverage of domains already partially contained in the quality of life measures used in our study. Statistical analyses The relations between the two independent variables of symptom complex and Total Lipodystrophy severity ratings to quality of life and mental health dependent variables were first

984 significance was set conservatively (i.e. P £ 0.01) due to the number of tests potentially inflating alpha error. Logistic regression analyses were utilized to evaluate the relative odds produced by increments of overall symptom count and by increments of Total Lipodystrophy severity ratings for mood scores beyond the cut-off representing mild depression. Analyses were conducted using Statistical Products and Service Solutions, Version 11.0 software.

analyzed using Spearman rank correlation coefficients. Least-squares regression analyses were then to be used for testing associations in which assumptions of linearity were satisfied (i.e., variables demonstrating normal distribution with a range of greater than 10 possible scores, skewness not exceeding ±1.0, and absence of floor and/or ceiling effects exceeding 50%). Demographic and clinical variables studied were assessed for significance of univariate relations to the dependent variables, to determine their inclusion in multivariate analyses in order to control for their potential to confound the results of associations between the variables of predominant interest. Potential interaction between gender and effects of symptom count and Total Lipodystrophy severity on mental health and quality of life variables were also explored. Effects of the total non-lipodystrophy symptom profile on quality of life and mental health were compared to those of lipodystrophy severity alone. For statistically significant associations between Total Lipodystrophy severity and dependent variables, additional analyses were carried out for Atrophy and Hypertrophy domain severity ratings separately. For analyses of association between symptom count and lipodystrophy ratings and all dependent variables, the threshold of statistical

Results Demographic and clinical characteristics for the sample of 77 participants are presented in Table 1. It was not necessary to exclude any participants from the study, as none of the patients considered for participation reported mild abdominal fat gain as the sole body alteration. Symptom inventory and lipodystrophy ratings for the sample are depicted in Table 2. The mean number of total symptoms excluding lipodystrophy features was 8.7, with a normal (i.e. Gaussian) data distribution The mean number of lipodystrophy features reported was 5.3 (median, 5), with 42% indicating that at least six of eight possible features were present to some degree; point prevalence in descending order: fat loss in

Table 1. Sample demographic and clinical characteristics n = 77; 60 male, 17 female

Mean ± SD

Range

Age in years Years since HIV diagnosis Years on antiretroviral medications Log10 viral load (copies/ml) a ; 56% BLQ CD4 count (cells/lL) CD4 nadir (cells/lL)

44.0 9.9 7.1 2.4 523 177

23–65 1.7–19.0 1.3–16.4 1.4–5.0 25–1493 9–620

a

b

± ± ± ± ± ±

8.3 3.8 3.4 1.3 315 148

Absolute viral load: mean 9300 copies/mL (SD, 21600); Range, BLQ to 108000 copies/mL. BLQ: below limit of quantification, also known as ‘undetectable’ (i.e. <50 copies/mL).

b

Table 2. Sample ratings of overall symptoms and lipodystrophy n = 77

Mean ± SD

Range

Symptom count a Lipoatrophy score (4 ‘fat loss’ items) Lipohypertrophy score (4 ‘fat gain’ items) Total Lipodystrophy severity score

8.7 9.0 4.4 13.4

0–15 0–16 0–14 1–28

a

Excluding lipodystrophy features.

± ± ± ±

4.5 4.3 3.5 6.0

985 legs, 88%; fat loss in buttocks, 86%; fat loss in arms, 82%; fat loss in face, 81%; fat gain in abdomen, 80%; fat gain under chin, 42%; fat gain at back of neck, 36%; fat gain in breasts, 34%. The proportion of patients with body changes at multiple sites (i.e. ‡2 body areas) was 96%. Mean Hypertrophy and Atrophy domain ratings were not significantly correlated (rs ¼ 0.14; p ¼ 0.2), suggesting that the magnitude of overall fat gain was not associated with magnitude of overall fat loss. A simple calculation in which these mean ratings were divided by the total possible four items for each dimension revealed that average Hypertrophy and Atrophy scores corresponded, respectively, to ‘very mild’ and ‘mild’ degrees of severity. The proportions of participants with Atrophy and Hypertrophy exceeding the ‘moderate’ level were, respectively, 21% and 4%. Using the same approach, the central tendency of Total Lipodystrophy ratings was nearest to ‘mild’ degree of severity; approximately 36% scored on average at the ‘mild’ level or worse, and 8% at the ‘moderate’ level or worse. Total Lipodystrophy severity ratings were distributed in a manner consistent with the assumptions of normality enabling parametric statistical analysis. Total Lipodystrophy severity ratings were not associated with total non-lipodystrophy symptom count (rs ¼ 0.12; p ¼ 0.3). Patients’ mean ratings were higher than physician-assessed ratings (paired sample t-test) of Atrophy (p ¼ 0.02), Hypertrophy (p ¼ 0.005) and Total Lipodystrophy (p ¼ 0.000). In spite of these differences, inter–rater reliability assessment comparing patient and physician Total Lipodystrophy severity ratings suggested moderate agreement (Intra-Class Correlation Coefficient: 0.69) and the difference in mean Total Lipodystrophy ratings was equivalent to one-quarter of a point in Likert-scale terms. Details of ratings for dependent measure variables and their relations, based on associations for ranked data, to Total Lipodystrophy severity are reported in Table 3, as are logistic regression results for the Beck Depression Inventory. Mean MOSHIV ratings for Physical Function, Body Pain, Mental Health, Cognitive Function, Heath Distress and Overall Quality of Life were consistent with early AIDS-related Complex normative values and compared less strongly to norms for asymptomatic

HIV infection published several years before the introduction of HAART [10]. In contrast, mean ratings for Role Function, General Health perceptions, Vitality/Energy and Social Function were more consistent with norms for pre-HAART-era AIDS-defined illness as defined by CD4 counts lower than 100 cells/lL but without major opportunistic infection at that time [27]. The mean HATQoL subscale ratings for all dimensions: Overall Function, Life Satisfaction, Health Worries, Financial Worries, Medication Worries, HIV Mastery, Disclosure Worries, Provider Trust, and Sexual Function more closely resembled norms for ‘non-AIDS’ HIV-positive adults than norms for AIDS-defined disease progression [28]. As seen in Table 3, Spearman’s rho associations between Total Lipodystrophy severity ratings and MOS-HIV and HAT-QoL subscales were nonsignificant. In contrast (not shown) the total nonlipodystrophy symptom count was significantly related to all MOS-HIV subscales and five of the nine HAT-QoL subscales. Similar results occurred for the HOPES Physical, Psychosocial and Sexual Summaries, but total non-lipodystrophy symptom count was not associated with HOPES Body Image ratings. Total Lipodystrophy ratings were significantly associated with both of the body image measures, the HOPES Body Image subscale and the DSFI Body Image subscale. DSFI Body Image ratings reflected considerably poor selfconcept. Mean scores compared to gender-specific general population reference norms [31] were at approximately the 5th percentile for men in the sample and the 3rd percentile for women in the sample. Approximately 70% of males and 81% of females in the sample, respectively, had scores below the 10th percentile. The study sample’s STAI state-anxiety mean score resembled the normative values [32] for employed general population adults and ‘non-stressful’ conditions. The RSES mean sample rating reflected moderately intact self-worth and confidence [33], with 7% reporting scores consistent with marked self-image impairment. The mean BDI-II score corresponded to the lower cut-point of the ‘mild’ depression score spectrum [34]. Closer examination of these BDI-II ratings revealed that 77% of the sample was euthymic or had minimal depression, 23% of the sample evidenced at least mild or worse depression, and 8% reported at least

986 Table 3. Mean sample scores of quality of life and mental health variables; Spearman rank correlations (rs) assessing effects of Total Lipodystrophy (LD) severity ratings on quality of life and mental health variables, and logistic regression assessing effects of LD severity ratings on likelihood of depression n = 77

Total LD Sample scores

Severity

Mean ± SD

rs

p

77.3 73.1 60.7 52.2 55.3 78.4 66.8 76.8 75.1 67.9

± ± ± ± ± ± ± ± ± ±

23.9 25.9 46.8 24.9 22.0 25.9 19.6 20.3 21.4 20.6

).25 ).22 ).08 .01 ).11 ).18 ).04 ).02 ).22 ).05

0.03 0.05 0.47 0.96 0.34 0.11 0.74 0.84 0.06 0.64

74.0 67.3 81.7 69.4 73.1 85.8 67.1 91.1 64.4

± ± ± ± ± ± ± ± ±

20.6 23.6 18.9 33.0 22.7 22.7 27.3 18.3 34.5

).17 ).05 ).06 ).07 ).26 ).07 ).15 ).14 .06

0.13 0.68 0.63 0.53 0.03 0.56 0.20 0.22 0.64

0.92 1.01 1.53 2.2 24.4 36.8 31.5

± ± ± ± ± ± ±

0.6 0.7 1.0 1.2 7.9 14.0 7.1

.27 .18 .09 .53 .35 .09 .00

0.02 0.13 0.44 0.00 0.00 0.42 0.97

a

MOS-HIV Physical function Body pain Role function General health Vitality/energy Social function Mental health Cognitive function Health distress Overall QoL HAT-QoL a Overall function Life satisfaction Health worries Financial worries Medication worries HIV mastery Disclosure worries Provider trust Sexual function HOPESb Physical summary Psychosocial summary Sexual summary subscale HOPES Body Imageb DSFI Body Imageb STAI - State Anxietyb RSESa BDI-IIb

c

3.0 ± 9.5

RO 1.00 (p = 0.95)

a

higher scores are favorable. lower scores are favorable. MOS-HIV: Medical Outcomes Study HIV Health Survey; HAT-QoL: HIV/AIDS-Targeted Quality of Life Instrument; HOPES Psychosocial Summary: average of eight subscales of HIV Overview of Problems–Evaluation System; HOPES Sexual Summary scale: average of three subscales; HOPES Body Image: distinct HOPES subscale within the Psychosocial domain; DSFI Body Image: Derogatis Sexual Function Inventory body image subscale; STAI: Spielberger State-Trait Anxiety Inventory; RSES: Rosenberg SelfEsteem Scale; BDI-II: Beck Depression Inventory, 2nd ed. c logistic regression delineating relative odds of a score exceeding the ‘mild’ depression threshold (i.e. >18/60). Note: p £ 0.01 is set as threshold of statistical significance. b

moderate to severe depression. The relationship of total symptom profile to these three measures of mental health (not shown) was statistically significant. Each additional symptom produced approximately 20% increased odds of a depression score representing ‘mild’ or worse mood impairment (RO ¼ 1.2; 95% CI [1.07 ) 1.36]; p ¼ 0.002). Relations between Total Lipodystro-

phy severity ratings and the STAI, RSES, and BDI-II mental health measures (Table 3) were statistically non-significant. The ratings for symptom inventory and for lipodystrophy severity met the assumptions of normal distribution, as mentioned earlier, rendering them appropriate for entry as continuous variables in regression analyses. Several quality of

987 of wellbeing. In contrast, the associations found between total symptom complex and several of the dependent variables were consistent with previous findings [1–10], suggesting negative repercussions for various dimensions of mental health and quality of life. An examination of relations between lipodystrophy-related body changes and wellbeing yielded no evidence of a gradient of negative effects according to the continuum of lipodystrophy severity, except for self-perceived body image. Individually assessing constituent elements of the total symptom complex likely stretches the capacity for standardized measures to capture differences in the range of morbidity of any single symptom, and may account for the apparent lack of impact of lipodystrophy on quality of life reported previously [24]. Our study results are also consistent with other research in which patients with lipodystrophic changes were twice as likely to feel recognizable as being HIVpositive yet were not different in attitudes to health condition or self-reported general wellbeing compared to patients with no evidence of lipodystrophy [21]. The present study found that lipodystrophy severity produced a relatively dramatic negative effect on body image. This finding, along with body image ratings at the low end of reference norms, supports the view that fat redistribution or alterations experienced by PHA can impair body image. There appeared to be some capacity for non-lipodystrophy symptoms to also deleteriously affect body image, a finding consistent with previous research on disease-specific symptoms of HIV [35]. Thus, perceptions by PHA of their sense of bodily self-concept and body integrity may be influenced by a feeling of lack of control of symptoms in general and their effects on the body. Fortunately, the lack of association between lipodystrophy severity and overall non-lipodystrophy symptom count in our study suggests that the

life subscales (i.e., four of the ten MOS-HIV and six of the nine HAT-QoL) did not satisfy assumptions of linearity due to skewness, ceiling effects, or constricted range of scores. The results of multivariate regression analyses for effects of symptom count and Total Lipodystrophy severity ratings on quality of life and mental health variables that did satisfy the assumptions of relational linearity, however, are not reported due to the lack of univariate relations between the dependent variables and demographic and clinical variables including gender, age, years since diagnosis of HIV, years on antiretroviral medications, CD4 count, or Viral Load. For women in the study sample a univariate relation between Total Lipodystrophy severity ratings and HAT-QoL Medication Worries was found (R2 ¼ 0.37; p ¼ 0.01); however, this association was partly attributable to one outlier value in the context of a small subsample size and, furthermore, did not retain statistical significance when non-lipodystrophy symptom count was held constant in a multivariate regression analysis. Associations of Atrophy and Hypertrophy domain ratings with body image were conducted separately and are reported in Table 4. These results are depicted in simple linear regression format as assumptions of linearity were met. For these analyses, Atrophy and Hypertrophy domain ratings accounted for variance in the HOPES Body Image subscale and DSFI Body Image subscale ratings at a statistically significant level.

Discussion The results of this study suggest that the degree of morphologic changes reflecting lipoatrophy and/or lipohypertrophy as experienced by adult PHA do not appreciably affect quality of life and mental health when assessed using conventional measures

Table 4. Least-squares linear regression analyses assessing effects of Atrophy and Hypertrophy ratings on HIV Overview of Problems– Evaluation System and Derogatis Sexual Function Inventory body image ratings n = 77

Atrophy p

0.27 0.11

0.000 0.004

R HOPES Body Image DSFI Body Image

Hypertrophy

2

R2 0.12 0.09

p 0.002 0.008

988 potential for emergence of lipodystrophy features is not inflated by the amount of other types of symptoms and side-effects. Similarly, the study results suggest no unidirectional or reciprocal influence between the severity of lipoatrophy and lipohypertrophy. If either of these two principal features of lipodystrophy inflated the occurrence of the other, the psychological consequences might be seen to be stronger. The study results appear to be at odds with previous research that incorporated survey methods and anecdotal reports regarding the psychological damage of lipodystrophy. The themes consistently emerging, including forced disclosure of diagnosis, stigmatization, discrimination, isolation, anxiety, depression, impaired self-esteem, challenges to interpersonal relations and sexual function, and concerns about the adverse effects of pharmacological treatment [18–22] logically lead to the expectation that such issues could be readily captured by standardized instruments purporting to measure those same concepts. However, the effects of lipodystrophy alone seem to be poorly captured by existing instruments evaluating factors commonly viewed as central to illness adjustment [30]. Body image measures are the notable exception, but tend to be lacking in range and depth of psychosocial constructs. Therefore, such tools need to be combined with supplementary patient interviewing to more accurately determine patients’ perceptions of the effects of lipodystrophy. The cross-sectional design of this study, without case-control such as a comparison sample of PHA without lipodystrophy features, limits causal assumptions about the ways lipodystrophy might affect important aspects of physical, mental and social wellbeing. The effects of lipodystrophy could be more clearly identified in longitudinal designs comparing PHA for whom body fat changes occur to those without such alterations, while taking into account other symptoms and side effects as well as strategies undertaken by patients and their health care providers to mitigate the impact of lipodystrophy, e.g. diet and exercise regimens aimed at overall fitness, muscle tone, etc. Determining the timing and co-factors related to the emergence of morphologic changes impacting on wellbeing would help guide interventions aimed at averting psychosocial damage.

The measures chosen for this study, although included out of consideration for the importance of breadth of multidimensional conceptual frameworks and multiple sourcing of item content, likely have insufficient sensitivity and specificity for responding to lipodystrophy-related impairment. The distributions of ratings for one-half of the quality of life subscales were not amenable to parametric (i.e. multivariate regression) analyses. The study may also be limited by sampling bias in terms of factors that predispose individual PHA to be more accepting of the tradeoff between side effects and the clinical benefits conferred by HAART. The sample tended to be of older age, experienced in living with HIV and antiretroviral therapy and benefiting clinically from such treatment. The mean nadir CD4 count was less than 200 cells/lL, suggesting that a large proportion of the study sample had previously reached a level of disease consistent with AIDS or had been at risk for opportunistic infections prior to CD4 repletion. Although relations between the degree of immune system repletion and current wellbeing were not strong enough to reach statistical significance in our study, previous research has shown that immunologic restoration potentially mitigates the effects of symptoms on quality of life over the longterm, including those associated with HAART [36]. PHA with the most concerns about the negative effects of HAART and not availing themselves of treatment would less likely be represented in the study. Younger, asymptomatic PHA requiring HAART might experience greater negative repercussions and adjustment difficulties associated with the onset of lipodystrophy. It is also noteworthy that mean MOS-HIV study sample ratings for Role Function, General Health perceptions, Vitality/Energy and Social Function were similar to reference values for patients with AIDS-defined stage of disease, even though only 14% of participants had a CD4 count below 200 at the time of completing the study measures. These findings may be attributable to previous illness-based quality of life deterioration that was more difficult to recover and, perhaps, less likely to be further reduced by the eventual emergence of lipodystrophy because the ratings were already low. Compared to hypertrophy, atrophy tended to be rated as more severe. This trend suggests that in future research there is merit in differentiating

989 between the ‘fat loss’ and ‘fat gain’ morphologic constructs when examining the phenomena of body fat redistribution and its effects on wellbeing. In addition, body alterations in the opposite expected direction in terms of the fat gain and fat loss areas typically associated with commonly held views of lipodystrophic changes should not be overlooked. Future research should also take into account the challenges concerning maximum objectivity in quantifying the degree of lipodystrophy severity. Physicians in this study tended to rate atrophy and hypertrophy lower in severity relative to patients, a trend consistent with previous research findings suggesting patients’ perceptions of symptom complex exceed that of their health care providers [26, 37, 38]. The actual disparity in ratings in the present study was relatively small, suggesting that patients’ ratings are likely sufficient for capturing the degree of body fat redistribution that has occurred. However, multiple sources of lipodystrophy measurement would help to detect any major inconsistencies that might call into question the true magnitude of body changes. Standardized measures typically utilized for evaluating mental health and quality of life of PHA lacked sensitivity and specificity in relation to lipodystrophy syndrome as studied using a cross-sectional methodology. To the extent that influence on these variables was evident, the greater contribution to well-being ratings was attributable to the total aggregate of symptoms and adverse experiences comprising the effects of HIV disease and treatment side effects. The potential for lipodystrophy to more profoundly impact on body image, however, was confirmed.

Acknowledgements The investigators thank the Canadian Foundation for AIDS Research (CANFAR) for financial support of this project. References 1. Badia X, Podzamczer D, Garcia M, Lopez-David C, Consiglio E. A randomized study comparing instruments for measuring health-related quality of life in HIV-infected patients. AIDS 1999; 13: 1727–1735.

2. Burgoyne R, Renwick R. Social support and quality of life over time among adults living with HIV in the HAART era. Soc Sci Med 2004; 58: 1353–1366. 3. Burgoyne R, Saunders D. Quality of life among urban Canadian HIV/AIDS clinic outpatients. Int J STD AIDS 2001; 12: 505–512. 4. Carrieri P, Spire B, Duran S, Katlama C, Peyramond D, Francois C, et al. Health-related quality of life after 1 year of highly active antiretroviral therapy. J Acquir Immune Defic Syndr 2003; 32: 38–47. 5. Cunningham W, Shapiro M, Hays R, Dixon W, Visscher B, George L, et al. Constitutional symptoms and healthrelated quality of life in patients with symptomatic HIV disease. Am J Med 1998; 104: 129–136. 6. Murri R, Ammassari A, Fantoni M, Scoppettuolo G. Disease-related factors associated with health-related quality of life in people with nonadvanced HIV disease assessed using an Italian version of the MOS-HIV health survey. J AIDS Hum Retrovirol 1997; 16: 350– 356. 7. Saunders D, Burgoyne R. Evaluating health-related wellbeing outcomes among outpatient adults with human immunodeficiency virus infection in the HAART era. Int J STD AIDS 2002; 13: 683–690. 8. Vogl D, Rosenfeld B, Breitbart W, Thaler H, Passik S, McDonald M, et al. Symptom prevalence, characteristics, and distress in AIDS outpatients. J Pain Symptom Manage 1999; 18: 253–262. 9. Wachtel T, Piette J, Mor V, Stein M, Fleishman J, Carpenter C. Quality of life in persons with human immunodeficiency virus infection: Measurement by the medical outcomes study instrument. Ann Intern Med 1992; 116: 129–137. 10. Wu A, Rubin H, Mathews W, Ware J, Brysk L, Hardy W, et al. A health status questionnaire using 30 items from the medical outcomes study. Med Care 1991; 29: 786–798. 11. Carr A, Samaras K, Thorisdottir A, Kaufmann G, Chisholm D, Cooper D. Diagnosis, prediction, and natural course of HIV-1 protease-inhibitor-associated lipodystrophy, hyperlipidaemia and diabetes mellitus: A cohort study. Lancet 1999; 353: 2093–2099. 12. Heath K, Hogg R, Chan K, Harris M, Montessori V, O’Shaugnessy M, et al. Lipodystrophy-associated morphological, cholesterol and triglyceride abnormalities in a population-based HIV/AIDS treatment database. AIDS 2001; 15: 231–239. 13. Schambelan M, Benson C, Carr A, Currier J, Dube M, Gerber J, et al. Management of metabolic complications associated with antiretroviral therapy for HIV-1 infection: Recommendations of an International AIDS Society-USA panel. J AIDS 2002; 31: 257–275. 14. Hogg R, Heath K, Yip B, Craib K, O’Shaughnessy M, Schecter M, et al. Improved survival among HIV-infected individuals following initiation of antiviral therapy. JAMA 1998; 279: 450–454. 15. Palella F, Delaney K, Moorman A, Loveless M, Fuhrer J, Satten G, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338: 853–860.

990 16. Bassetti S, Battegay M, Furrer H, Rickenback M, Flepp M, Kaiser L, et al. Why is highly active antiretroviral therapy (HAART) not prescribed or discontinued? J AIDS 1999; 21: 114 –119. 17. Shafer R, Vuitton D. Highly active antiretroviral therapy (HAART) for the treatment of infection with human immunodeficiency virus type 1. Biomed Pharmacother 1999; 53: 73–86. 18. Collins E, Wagner C, Walmsley S. Psychosocial impact of the lipodystrophy syndrome in HIV infection. AIDS Reader 2000; 10: 546–550. 19. Dukers H, Stolte I, Albrecht N, Coultinho R, de Wit J. The impact of experiencing lipodystrophy on the sexual behaviour and well-being among HIV-infected homosexual men [Letter]. AIDS 2001; 15: 812–813. 20. Oette M, Juretzko P, Kroidl A, Sagir A, Wettstein M, Siegrist J, et al. Lipodystrophy syndrome and self-assessment of well-being and physical appearance in HIV-positive patients. AIDS Patient Care 2002; 16: 413–417. 21. Paton N, Earnest A, Ng Y, Karim F, Aboulhab J. Lipodystrophy in a cohort of human immunodeficiency virus-infected Asian patients: Prevalence, associated factors and psychological impact. Clin Infect Dis 2002; 35: 1244– 1249. 22. Power R, Tate H, McGill S, Taylor C. A qualitative study of the psychosocial implications of lipodystrophy syndrome on HIV positive individuals. Sex Transm Infect 2003; 79: 137–141. 23. Lenert L, Feddersen M, Sturley A, Lee D. Adverse effects of medications and trade-offs between length of life and quality of life in human immunodeficiency virus infection. Am J Med 2002; 113: 229–232. 24. Blanch J, Rousaud A, Martinez E, De Lazzari E, Peri J, Milinkovic A, et al. Impact of lipodystrophy on the quality of life of HIV-1-infected patients. J Acquir Immune Defic Syndr 2002; 31: 404–407. 25. Carr A, Samaras K, Burton S, Freund J, Chisholm D, Cooper D. A syndrome of peripheral lipodystrophy, hyperlipidaemia and insulin resistance in patients receiving HIV protease inhibitors. AIDS 1998; 12: F51– F58. 26. Justice A, Rabeneck L, Hays R, Wu A, Bozzette S. Sensitivity, specificity, reliability, and clinical validity of provider-reported symptoms: A comparison with self-reported symptoms. J AIDS 1999; 21: 126–133.

27. Wu A, Revicki D, Jacobson D, Malitz F. Evidence for reliability, validity and usefulness of the Medical Outcome Study HIV Health Survey (MOS-HIV). Qual Life Res 1997; 6: 481–493. 28. Holmes W, Shea J. A new HIV/AIDS-targeted quality of life (HAT-QOL) instrument: Development, reliability and validity. Med Care 1998; 36: 138–154. 29. Ganz P, Schag C, Kahn B, Petersen L, Hirji K. Describing the health-related quality of life impact of HIV infection: Findings from a study using the HIV Overview of Problems–Evaluation System (HOPES). Qual Life Res 1993; 2: 109–119. 30. Martinez E, Garcia-Viejo M, Blanch J, Gatell J. Lipodystrophy syndrome in patients with HIV infection: Quality of life issues. Drug Safety 2001; 24: 157–166. 31. Derogatis L, Melisaratos N. The DSFI: A multidimensional measure of sexual functioning. J Sex Marital Ther 1979; 5: 244–281. 32. Spielberger C, Gorsuch R, Lushene R, Vagg P, Jacobs G. State-Trait Anxiety Inventory for Adults. Palo Alto, CA: Mind Garden Inc., 1983. 33. Rosenberg M. Society and the Adolescent Self-Image. Princeton, NJ: Princeton University Press, 1965. 34. Beck A, Steer R, Brown G. BDI -II Manual for Beck Depression Inventory, 2nd edition, San Antonio, TX: The Psychological Corporation (Harcourt Brace & Co.), 1996. 35. Chapman L. Body image and HIV: Implications for support and care. AIDS Care 1998; 10: S179–S189. 36. Burgoyne R, Rourke S, Behrens D, Salit, I. Long-term quality of life outcomes among adults living with HIV in the HAART era: The interplay of changes in clinical factors and symptom profile. AIDS Behav 2004; 8: 151–163. 37. Fontaine A, Larue F, Lassauniere J. Physicians’ recognition of the symptoms experienced by HIV patients: How reliable? J Pain Symptom Manage 1999; 18: 263–270. 38. Justice A, Chang C, Rabeneck L, Zackin R. Clinical importance of provider-reported HIV symptoms compared with patient-report. Med Care 2001; 39: 397–408.

Address for correspondence: Robert Burgoyne, Toronto General Hospital Immunodeficiency Clinic, NCSB-5C430, 585 University Avenue, Toronto, Ontario, Canada M5G2N2 Phone: 416-340-4800, ext. 8609; Fax: 416-340-4890 E-mail: [email protected]