The Impact of Chronic Venous Insufficiency and Leg Function on the Quality of Life of HIV-Positive Persons
Human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) is a worldwide health problem. In addition to delaying HIV progression, a goal of treatment is to maintain optimal quality of life. Because wounds have the potential to negatively impact quality of life, their assessment and management in HIV-positive persons is essential.
Chronic venous insufficiency (CVI) — heralded by varicose veins, edema, stasis dermatitis, indurated discolored skin, ulcers, and leg pain — may impair lower extremity function. The condition has a high occurrence in persons who have injected drugs — use of injected drugs is also a causative factor for HIV.1 Little is known about the impact of CVI on the quality of life of HIV-positive persons. The purpose of this study was to examine 1) the extent to which CVI and leg function contribute to quality of life in HIV-positive persons and 2) other variables that contribute as mediators.
Health-related quality of life is increasingly recognized as an important measure for assessing the burden of chronic illness2; therefore, symptoms are often examined in terms of quality-of-life impact.
HIV/AIDS, symptoms, and quality of life. In a survey (n = 504), Vogl and colleagues3 assessed the physical and psychological symptoms experienced by ambulatory patients with AIDS. No association was found between symptoms and immune status but injection drug users reported more symptoms and higher overall physical symptom distress than persons who reported contracting HIV disease through homosexual or heterosexual contact. Injection drug users experienced more pain and arm or leg swelling (26.7% prevalence for arm or leg swelling) and 41.9% reporting a high distress rating. This study did not separate swelling in the arms versus the legs.
Hudson and colleagues4 examined symptoms in 118 women enrolled in an outpatient clinic, 27 of whom recently had been hospitalized with an AIDS diagnosis. The 10 most prevalent HIV-related symptoms were muscle aches, depression, thirst, weakness, fear/worries, fatigue, dry mouth, difficulty concentrating, gas/bloating, and painful joints. Quality-of-life scores determined using Medical Outcomes Short Form-36 (MOS SF-36) were lower than the normative sample data by 0.9 standard deviations or greater. Symptoms were significantly associated with lower role functioning (P = .003).
Over a 4-year period, Burgoyne and colleagues5 also examined symptoms associated with HIV/AIDS in 41 persons. The HIV symptom status of these participants was rated as asymptomatic (29%), symptomatic (32%), and AIDS (39%). Symptom increases were most commonly due to diarrhea (32%), muscle/abdominal pain and/or headache (37%), weakness/fatigue (20%), peripheral neuropathy (20%), dermatological problems (eg, rash/itch — 15%), and nausea (15%).5,6 Fleming and colleagues used a cross-sectional design to examine health-related quality of life using the SF-36. The study included 136 persons with co-infections of hepatitis C and HIV, 110 persons with hepatitis C only, and 53 persons with HIV only. The three groups were similar in quality-of-life scores. Age, foreign birth, unemployment, injection drug use, and depression were associated with diminished quality of life.2
In the 4-year prospective study, quality-of-life scores were reported over time as “stable” or “slightly improved” during highly active antiretroviral therapy (HAART) (1997–2000).5 The stability of quality-of-life scores was suggested as a complex interplay between clinical outcome and change in symptoms and/or side-effect profile over time. Quality-of-life scores were less sensitive to immunologic/virologic changes than symptom changes.5 Preau and colleagues7 longitudinally studied factors associated with better health-related quality of life during the first 3 years after starting HAART. Independent factors associated with high quality-of-life mental component scores were male gender, not more than one change in HIV treatment, few self-reported symptoms, and trust in the physician. Independent factors associated with high physical quality-of-life scores were employment, no children, few self-reported symptoms, and satisfaction with the information and explanation by the medical staff.7
Pain impacts quality of life and affects 25% to 80% of HIV-positive persons.8 In a descriptive study of 103 adults with AIDS, Norval9 identified the lower legs as the most prevalent site for pain. Pain may be due to the disease, a side effect of treatment, or unrelated to the disease or treatment. Two common types of pain for patients with HIV disease are musculoskeletal pain and peripheral neuropathy. Pain can impair functional mobility and limit physical activity. Pain severe enough to interfere with daily living has been associated with a lower level of functional quality of life.10
In summary, HIV/AIDS affects quality of life, often in relation to symptoms. Quality-of-life scores for HIV-positive persons are lower than the general population. Although not reported, CVI symptoms that may influence quality of life are pain, dermatologic problems, and leg swelling. Quality of life appears affected by a complex interplay between clinical outcome and change in symptom and/or side effect profile over time.
Leg function, CVI, and HIV/AIDS. Little is known about the effect of HIV/AIDS on the calf muscle pump. Ankle mobility and calf-muscle motion are important mechanisms of the lower leg. Venous return is facilitated by the relationship among intact venous valves, the ankle joint, and calf muscle pump.11 This pumping action empties the deep veins in a cephalic direction. Ankle or calf-muscle damage can increase venous pressure and damage veins, leading to venous reflux. In addition, increased severity of CVI can worsen range-of-motion and calf-muscle pump function but the effects vary widely.12-14 A higher prevalence (61%) of CVI has been reported in HIV-positive injection drug users than in HIV-positive persons who did not inject drugs in a cross-sectional stratified design with quota sampling.15 As measured by goniometry, HIV-positive injection drug users also had less ankle motion than those who had not injected drugs.16
Lower extremity function in persons with HIV/AIDS can be affected by distal symmetrical polyneuropathy that occurs as a result of HIV/AIDS, some antiretroviral therapy, disorders such as alcoholism, diabetes mellitus, and nutritional problems.17 Distal symmetrical polyneuropathy can cause non-painful paresthesias, abnormalities of pain and temperature perception, and reduced or absent ankle reflexes, affecting the person’s ability to walk, work, and participate in activities. Studies15,16 have shown the presence of peripheral neuropathies was not significantly related to CVI or less ankle mobility.
CVI, symptoms, and quality of life. Kurz and colleagues18 examined the impact of varicose veins on quality of life. Study participants (1,054 persons with varicose veins and 259 without varicose veins) represented the seven clinical categories of the clinical, etiologic, anatomic, and pathophysiologic (CEAP) classification for venous insufficiency, ranging from having no varicose veins to varicose veins with active ulcers. The physical composite score of the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) quality-of-life instrument decreased according to increasing severity of CVI; the lowest scores were found in persons with varicose veins and active ulcers. Physical quality-of-life scores in patients with varicose veins were associated with concomitant venous disease such as edema, aching, heaviness of the legs, and/or ulcers, not just the presence of varicose veins.
Kahn and colleagues19,20 examined CVI as classified by the CEAP in relation to the SF-36 quality-of-life instrument. The physical component quality-of-life score decreased significantly with increasing CEAP class; no such association was found with the mental component score. After adjusting for confounding variables, CEAP class was predictive of the disease-specific quality-of-life instrument score but was not predictive of the generic SF-36 quality of life. Van Korlaar and colleagues21 noted in an analysis of 25 articles — including 12 in which quality of life in patients with CVI was assessed — that quality of life of patients with CVI is affected in the physical domain mainly by pain, physical functioning, and mobility and that patients suffer from negative emotional reactions and social isolation.
Persons with venous ulcers experience many symptoms that can negatively impact quality of life. For example, several studies have shown that persons with venous ulcers experienced leg pain, less vitality, and restrictions on physical and social functioning.22-24 Leg pain associated with venous ulcers can be severe and has been rated as the worst component of a leg ulcer in qualitative and quantitative research studies.22-30 Persons with CVI often were forced to make significant life changes and to find satisfaction in new activities. The leg ulcer was perceived as the person’s most significant health problem even when other important medical conditions were present in one qualitative and one quantitative research study.26,27 Persons with venous ulcers often are depressed, fearful, socially isolated, and angry.23,27 In addition, they experience disturbed sleep, impaired mobility, and financial concerns.22,27 Lack of employment negatively affects quality of life.30 Other reported symptoms were pruritus, drainage, and edema.27
When Franks and colleagues31 compared quality of life in persons with venous ulcers (n = 118) to a normative sample, they found the former had lower role-emotional, social functioning, role-functioning, role-physical, and bodily pain quality-of-life scores. If leg ulcers healed, bodily pain, mental health, role-physical, role-emotional, and vitality improved.31 In contrast, persons with venous ulcers in a Brazilian study (n = 89)30 had good quality of life in terms of spirituality and family.
In a cross-sectional study of 32 persons who had venous ulcers, higher current pain (r = .42, P = .05) and higher worst pain (r = .41, P = .05) were significantly related to larger wound area.32 Leg pain also has been associated with greater difficulties in the home and poor quality of life.33 The most painful activities were working, walking outside, standing, and stair climbing.
Hence, CVI has many clinical presentations that depend on the severity of the condition. Leg pain has the ability to interfere with activities, compromising daily function.
Research question. Because of its many symptoms, CVI has the potential to negatively impact quality of life for HIV-positive persons. Research examining the impact of CVI on quality of life for HIV-positive persons is lacking. The purpose of this prospective, cross-sectional study was to examine 1) the extent to which CVI and leg function contribute to quality of life in HIV-positive persons and 2) other variables that contribute as mediators.
Design and procedure. The extent to which CVI and leg function contribute to quality of life in HIV-positive persons was part of a larger project that examined ankle mobility and CVI.15,16 A cross-sectional design was used. Because previous work1 indicated that injection drug use would have the greatest impact on the legs, quota sampling was used to obtain 46 injection drug users and 27 non-injection drug users who were HIV-positive. A sample of this size has good power (= .8) to detect a correlation of .33 or larger at a significance level (alpha) of .05.34 Thus, in eight out of 10 samples of the current size, population correlations of .33 or larger will be found significant. This calculation was determined from a table in which power was given as a function of sample size, the size of the population correlation coefficient, and alpha level. A correlation of .33 is a medium effect size.34 Participants were recruited from an infectious diseases clinic in a large, urban midwestern city. Participants had intact bilateral lower extremities, were able to walk, were 30 to 65 years of age, not pregnant, and able to understand and respond in English.
The study was approved by a Human Investigation Committee for the Protection of Human Rights. The infectious diseases clinic staff allowed information about the study to be posted in their examination rooms. Interested patients telephoned the primary author who determined eligibility and arranged a study appointment. In a clinic examination room, the study was described in detail and signed consent was obtained. Questionnaires were read to participants; their legs were examined for CVI status from the knees to the feet. Participation took 60 to 90 minutes. Participants were told that information obtained during the research study was not part of their medical record. All persons who arranged a research appointment completed the study and participants were compensated $20.
Instruments. The Demographic and Health History Questionnaire was used to obtain information to describe the sample. It included questions about age, sex, race, and other health conditions the person had in addition to HIV/AIDS. The self-reported health problems were tabulated for a comorbidity score. “Years diagnosed” was determined from the dates participants reported they learned their HIV-positive status. Most recent CD4 (T-helper cells) values and HIV viral load values (HIV-RNA: human immunodeficiency virus-ribonucleic acid) were taken from patient medical records; other laboratory studies were not recorded.
A Drug Use History Questionnaire provided information about the person’s use of illicit drugs, route of use, and years of use.35 Self-report of drug histories has been found to be reliable when compared with collateral interviews and biomarkers.36-39 “Years of injection drug use” was calculated by subtracting age of first injecting from age of last injecting across all injected drugs.
Quality of life was measured using the SF-36,40 a comprehensive, self-reported generic measure of health status that includes physical, emotional, and social components. The instrument has been tested in multiple healthy and ill adult populations. Thirty-six items address eight dimensions of quality of life: physical functioning, body pain, role limitations due to physical problems, general health perceptions, vitality/energy, social functioning, mental health, and role limitations. Scores on each dimension of the SF-36 range from 0 to 100; higher scores reflect better quality of life.40 An overall quality-of-life score was obtained by adding the scores from the eight dimensions. Reported reliability of the scales ranges from .63 to .94 for a random sample of a US population.40 The SF-36 has been used in studies related to HIV/AIDS and CVI/venous ulcers.
A Leg Function Scale score was a compilation of three instruments: the Leg Pain Questionnaire, Leg Interference with Daily Living Scale, and Difficulty in Using Legs Scale. The Leg Pain Questionnaire was adapted from an American Pain Society questionnaire.41 Participants were asked how much pain they were currently having in their legs, worst leg pain in 24 hours, and average leg pain in 24 hours. These three items were measured on 11-point scales, which ranged from no pain (score of 0) to worst possible pain (score of 10). The pain ratings were highly correlated (median r = .83). A composite leg pain score, developed from tabulating the three pain items, had a reliability coefficient alpha = .89 in this study and .92 in a previous study.42
The Leg Interference with Daily Living Scale asks how much pain in the legs interfered with nine items: general activity, mood, walking, relationships with others, sleep, work, stair use, drug treatment, and enjoyment in life. Scoring for the items ranged from 0 (does not interfere) to 10 (completely interferes). The nine items were totaled for a composite interference score; the coefficient alpha was .95.
The Difficulty in Using Legs Scale asks how legs were affected by walking, standing, stair use, and working. The scale for these items ranged from 0 (no problem) to 10 (great difficulty). The four items were totaled for a composite difficulty score; the coefficient alpha was .95. The Leg Function Scale was computed by summing the three component scales and then subtracting this total from a positive constant — higher Leg Function scores reflected better leg function. The Leg Function Scale with the components as items had a coefficient alpha of .67.
Participants’ legs were evaluated from the knee to the foot for evidence of CVI. Each leg was classified for venous disease according to the clinical component (C) of the CEAP classification.43 The seven classifications include: class 0, no visible or palpable signs of venous disease; class 1, telangiectasias or reticular veins; class 2, varicose veins; class 3, edema; class 4, skin changes due to venous disease; class 5, skin changes from healed ulcers; and class 6, skin changes with active ulcers. The highest classification of the worst leg was used in statistical analysis. The CVI clinical classification score has been used in other research studies.43-45
Participants. Of the 73 participants, 36 were men, 37 were women, mean age = 45.8 years (SD = 7.6); 93% were African American. The mean number of years of being HIV-positive was 9.55 (SD = 4.40 years). Other common medical conditions included depression (n = 35), hypertension (n = 32), arthritis (n = 26), hepatitis C (n = 25), bronchitis (n = 20), and peripheral neuropathies (n = 16). None of the participants had vascular lesions associated with hepatitis C.
The 46 persons who injected drugs did so for a mean of 16.57 years (SD = 10.21). Injected drugs included heroin (n = 44), cocaine (n = 32), amphetamines (n = 9), methadone (n = 5), and other opioids (n = 7); 41 injected in the veins of the arms and upper body and groin, legs, and feet (mean years = 8.15, SD = 9.42) and five injected only in the arms and upper body. Because the focus was on injecting drugs irrespective of site of injection, the five persons who injected only in the upper body remained in the study.
Statistical analysis. Standardized SF-36 scores were used to compare quality of life of study participants. Correlations and t-tests were used to examine the relationships between SF-36 scores and medical and sociodemographic variables. Correlations also were computed among SF-36 scores, Leg Function scores, and CVI severity score. Polynomial regression was used to examine the possibility of nonlinear relations between leg function and quality of life as dependent variables and linear and nonlinear components of CVI as predictor variables. In a polynomial regression, nonlinear relationships are captured by including additional variables that are products of the variable of interest. Thus, the regression equations included a variable that was CVI multiplied by itself (CVI squared) in addition to CVI as a predictor variable. This squared term represents the nonlinear aspect of the regression because every increment in CVI is associated with an increment raised to the second power and both of these increments contribute to the predicted score. To control for potential confounding and to highlight the relations of importance, a causal model including injection drug use, comorbidities, CVI, leg function, and quality of life was tested using structural equation modeling techniques.
Quality of life and CVI. The SF-36 scores are presented in Figure 1 with a comparison to standardized scores.40 All SF-36 scores were lower than 1 standard deviation, except for “vitality” which was 0.9 standard deviations, from the standard score. This large difference indicates that the average HIV-positive person in this study had a quality-of-life score lower than 84% of the normative population. The SF-36 scores were significantly correlated with each other, ranging from .37 to .71.
The CVI scores of the worst leg were rectangularly distributed: 18 participants in Class 0; six in Class 1; 14 in Class 2; four in Class 3; seven in Class 4; one in Class 5; and 13 in Class 6. The CVI classification of the left and right leg was highly correlated (r = .90, P <.01). None of the eight SF-36 scores was significantly correlated to CVI classification score. The largest correlation (r) was .13, P = .28. The SF-36 scores were examined in relation to other health-related and demographic variables in the study. Each of the SF-36 scores was inversely and significantly correlated with the number of comorbid health problems (r = -.34 to -.48, P = .01); the higher the quality of life, the fewer the comorbid health conditions. None of the SF-36 scores was significantly related to age, sex, race, history of injection drug use, CD4 (T-helper cells) values, HIV viral load values (HIV-RNA: human immunodeficiency virus-ribonucleic acid), or years diagnosed HIV.
Leg function, quality of life, and CVI. The correlations among the Leg Function component scores and SF-36 scales are presented in Table 1. The mean Leg Function score ranged from 3.87 to 4.74 on a 0 to 10 scale, indicating moderate levels of leg pain, interference, and difficulty in using the legs. Leg Function component scores were significantly correlated among themselves and with all SF-36 scales except “vitality”. In general, the higher the Leg Function Scale scores, the lower the quality-of-life scores. Among the SF-36 scales, “bodily pain” was most highly correlated with the Leg Function components. Chronic venous insufficiency was not significantly correlated with the Leg Function Scale or the SF-36. To investigate the possibility of nonlinear correlations between CVI and the measures of quality of life and leg function, polynomial regression was used. In these regressions, the criterion variable was either the total SF-36 quality-of-life score or the Leg Function Scale score; the predictors were CVI score and the CVI score squared. The regression equation with the SF-36 quality of life total as the dependent variable was not significant and only accounted for 2% of the variance in quality of life. The regression equation with the Leg Function Scale score as the dependent variable was highly significant (F[2,70] = 7.12, P = .002) and accounted for 17% of the variance in leg function. The linear and nonlinear terms were both significant (P <.01). The polynomial regression of leg function on CVI score is shown in Figure 2. The open circles show the actual joint distribution of leg function and CVI scores. Each circle is one participant’s pair of scores. Wide variation in function, etc. at each level of CVI were noted. For example, at CVI level 0, leg function scores ranged from a little less than two to 10. The general shape of this distribution is best approximated by a quadratic function that first decreases then increases. Leg function first decreased with increasing disease severity and then in the later stages of CVI, leg function increased.
Model development: CVI, injection drug use, leg function, and quality of life. A model was developed to examine the relationship among CVI, injection drug use, leg function, quality of life, age and comorbidities. The model is based on understanding of the causal relations among the variables examined (see Figure 3) and estimated using AMOS 5 structural equation modeling software.46 Parameter estimates and standard errors were obtained using the nonparametric bootstrap as recommended for mediation models with a small sample size.47 The overall fit of the model was excellent, Chi-square (11, N = 73) = 9.12, P = .61, comparative fit index (CFI) = 1.0, and root mean square error of approximation (RMSEA) <.01.
The path coefficients shown in the figure are similar to standardized regression coefficients and should be interpreted in the same way. One can interpret the model by starting with the rightmost variable (quality of life, QOL) as the most distal outcome. This is regressed on both variables that have arrows going to it — namely, leg function and comorbidity. This part of the model shows that the direct effect of leg function on QOL is greater than the direct effect of comorbidity on QOL (.47 versus -.25). Because these are standardized coefficients, the magnitude of the effect is in relation to a standard deviation change in the predictor. A one standard deviation change in leg function is expected to result in a .47 standard deviation change in QOL, controlling for level of comorbidity. The other parts of the model are interpreted similarly. Each variable with arrows pointing to it is a dependent variable in a regression in which the variables pointing to it are predictors; therefore, leg function has three predictors, two of which are significant.
The model accounted for 42% of the variance in quality of life and 41% of the variance in leg function. As noted above, leg function had a direct effect on quality of life, controlling for comorbidities. Several other significant direct effects were of interest: CVI-nonlinear had a direct effect on leg function, controlling for comorbidities, and injection drug use had a direct effect on CVI severity, controlling for age of participant.
Specific indirect effects also can be identified by following a chain of significant paths in the causal direction. Two indirect effects were significant: the effect of CVI-nonlinear on quality of life through leg function and the effect of comorbidity on quality of life through leg function. In other words, leg function mediated or accounted for the relations among quality of life, CVI-nonlinear, and comorbidity.
The extent to which CVI contributed to quality of life and leg function in HIV-positive persons, as well as other variables that contributed as mediators, was examined. Low quality-of -life scores were noted in the sample, which is similar to other studies that report low quality-of-life scores.2,4 In the general population, CVI is associated with restricted employment options, loss of functional independence, decreased quality of life, and increased pain and suffering. A significant nonlinear relation between CVI quality-of-life scores was noted. It is hard to understand why the relation is nonlinear rather than linear. Kahn and colleagues19 found CEAP class was predictive of a disease-specific quality-of-life score but not the generic SF-36 score. The authors’ previous work33 with persons with venous leg ulcers and a disease-specific quality-of-life instrument reported a larger leg ulcer size negatively impacted quality of life, but only three of 32 subjects were HIV-positive. Only the clinical component of the CEAP in its basic form of the highest number was used; the more comprehensive clinical description where each number of a clinical manifestation was placed together was not utilized.43 A more comprehensive scoring may result in a more understandable/linear relation between disease severity and leg function. Funding was not available for Doppler studies to document the other components of the CEAP. The CEAP classification is not static and a person’s disease can be reclassified.43 Research needs to continue to explore the reliability of the CEAP in terms of individual components of the clinical classification. Further research with a larger sample is needed regarding the impact of CVI in the HIV-positive persons.
Leg Function scores were moderate in terms of pain, interference, and difficulty. The Leg Function score was associated with lower quality of life. This may reflect other leg problems, such as peripheral neuropathy, that may occur in HIV-positive persons. Peripheral neuropathy may be caused from the HIV-positivity, its treatment, or other conditions. Pain is a common problem for HIV-positive persons; the legs are frequently affected. In a previous study42 of lower extremity changes, pain, and function in injection drug users, leg pain and function were related to CVI severity. Leg pain was a mediator of the relationship between CVI and functioning, controlling for the effects of other chronic diseases. Only eight out of 100 persons were HIV-positive; thus, its impact could not be examined.
This study has implications for clinical practice. HIV-positive persons who injected drugs are at risk for CVI. Both HIV and CVI can negatively affect function of the legs. Leg function can impact ability to work, daily activities, and leisure activities — all critical in terms of quality of life. Practitioners need to assess for CVI, as well as for HIV and hepatitis C, with regard to changes to the legs; this assessment will allow for comprehensive planning in terms of leg health.
The sample size was small. Persons in this study had to return to the clinic a second time to complete the study — returning to the clinic meant additional transportation costs and time away from home or work. If the study could have been completed while participants were in the clinic for routine HIV care, the sample size may have increased. Future studies need to examine the best way to obtain participants. Also, it was not noted whether participants were obtaining treatment for their CVI, such as wearing support stockings or having their leg ulcer treated. Chronic venous insufficiency treatment may affect quality of life, pain levels, and function.
Future research should include CVI treatment methods. Information about peripheral neuropathy was taken from the medical record and self-report. Physical assessment of peripheral neuropathy should be examined since it is prevalent in HIV-positive persons.
Chronic venous insufficiency has a low prevalence in the general population, primarily affecting the elderly, but it is a major health problem for young/middle aged persons who are or have chronically abused injected drugs. Drug abuse is a causative factor for HIV as well as CVI. One of the most frustrating aspects of this “early-onset” form of CVI is the fact that those who have stopped abusing drugs remain at risk; damage that occurred during the active period of injection persists and advances long after drug use ceases. Routine examination of HIV-positive persons does not include CVI, yet it is another comorbid health problem that can affect functioning. Additional research is needed regarding CVI in the HIV-positive person in terms of quality of life and leg function.
This project was funded in part by Blue Cross/Blue Shield of Michigan Foundation, Excellence in Research Award (McDevitt Award).