HIV Articles  
Back 
 
 
Quality-Of-Life Predicts Survival in HIV: HIV & aging
 
 
  (CID Jan 15 2010)
 
from Jules: It goes understood that if HIV accelerates aging with HIV with accelerated frailty, senescence, increased rates of comorbidities, bone loss, fractures, neurologic impairment, diabetes, CVD, survival rates in HIV WILL decline, as I have predicted I think death rates will soon start to increase, thus we must devote our energies to solving theses key HIV & aging questions which are- does HIV cause premature senescence & ongoing activation, are these causing or significantly contributing ccelerated aging & increased comorbidities & death, and can we intervene, this is the big question, can we intervene & how to prevent accelerared aging. We can no longer ignore this problem and minimizing this compelling problem by disproportionately diverting resources from patients needs like this for vaccine research and global HIV. Let me be clear I am not saying we should ignore global & vaccines, I am saying we need to increase resources to the HIV & Aging problem, and perhaps that means taking a little bit of the 100s of millions of dollars devoted to the vaccine programs, which everyone agrees has problems, and devoting it to aging & HIV research for answers. AND, aging patients WILL need greater resources, which we are not discussing for what & how much.
 
"In summary, physical HRQL (health-related quality of life) predicts survival in the long term in a population of HIV-infected patients receiving HAART independent of demographic and clinical variables.....Of patients with a PHS, 26 (20%) died in quartile 1 (indicating worst HRQL), 17 (13%) died in quartile 2, 10 (8%) died in quartile 3, and 5 (4%) died in quartile 4 (indicating best HRQL) ( p<.001)....investigators found that poorer health-related quality of life in the following areas were associated with an increased risk of death: physical functioning (p < 0.001), pain (p < 0.001), role functioning (p < 0.001), social functioning (p = 0.01), and general health (p = 0.01)....failure to complete all items of the questionnaire, resulting in missing summary scores, was associated with an increased risk of mortality.....investigators suggest that measuring self-reported health status is important because it "may capture the entire range of symptoms, including subtle signs of disease progression."....Multivariate analysis showed that higher PHS score (HR for each 5 points, 0.8; 95% CI, 0.7-0.9) and no treatment with antiretroviral therapy before the start of HAART (HR, 0.1; 95% CI, 0.05-0.2) were protective for the risk of death. In contrast, higher age (HR for each 5 years, 1.3; 95% CI, 1.1-1.5), having a CD4 cell count less then 200 x 106/L at the start of HAART (HR, 2.1; 95% CI, 1.2-3.8), and having a detectable viral load at study entry (HR, 5.2; 95% CI, 2.8-9.4) increased the risk of death. All variables met the proportional hazards assumption (p>.05)......Future research should determine the prognostic value of HRQL on survival in clinical settings. Also, the mechanism by which physical HRQL influences survival remains to be determined. Potentially, this information could be highly useful for physicians in assessing the prognosis of their patients in resource-rich and resource-poor countries......Although PHS was independently predictive of survival, this does not imply a causal relationship......the post hoc analyses regarding causes of death did not reveal information about the mechanism because causes of death were too diverse and patient numbers were too small......Twenty-two patients (33%) died of AIDS-related diseases, such as malignant tumors and infections, 12 (18%) of heart and lung conditions, 11 (17%) of non-AIDS-related malignant tumors, 6 (9%) of liver and pancreas conditions, 6 (9%) of suicide, accidents, or euthanasia, 4 (6%) of non-AIDS-related infections, and 5 (8%) of unknown conditions. There was no clear pattern of differences in PHS and MHS scores among the different causes of death......An interesting finding is that failure to complete all items of the questionnaire, resulting in missing summary scores, was associated with an increased risk of mortality. Failure of completion might be due to fatigue or concentration problems and thereby indicative of a poor health status. In particular, failure to complete items about cognitive functioning increased the risk of mortality (HR, >20; p<.001). These items focus on difficulties with reasoning and solving problems, forgetfulness, trouble paying attention, and difficulties with concentration and thinking.......whereas perceptions of poor health might lead to behaviors that positively influence health status, they predominantly have been found to induce behaviors that negatively influence health status, such as lesser involvement in preventive activities or self-care and nonadherence to screening recommendations, medication, and treatment. In addition, health perceptions may be influenced by resources in the external social environment, such as income and social networks, and within-person resources, such as control over health. Therefore, self-rated health is believed to reflect both perceptions and subsequent behaviors and resources....Results were scored on a scale from zero to 100. The higher the score, the better a patient's health-related quality of life. The results were stratified into quartiles, with the first quartile having the lowest score and the fourth the highest....Most of the patients (76%) were gay men and 83% were from the Netherlands. The investigators suggest that this was a potentially a limiting factor of their research....By the end of follow-up in March 2008, 66 patients (12%) had died"
 
Health-Related Quality of Life and Survival among HIV-Infected Patients Receiving Highly Active Antiretroviral Therapy: A Study of Patients in the AIDS Therapy Evaluation in the Netherlands (ATHENA) Cohort
 
Clinical Infectious Diseases Jan 15 2010;50:255-263
 
I. Marion de Boer-van der Kolk,1 Mirjam A. G. Sprangers,1 Jan M. Prins,2 Colette Smit,3 Frank de Wolf,3 and Pythia T. Nieuwkerk1 1Department of Medical Psychology and 2Division of Infectious Diseases, Tropical Medicine and AIDS, Department of Internal Medicine, Academic Medical Centre, and 3HIV Monitoring Foundation, Amsterdam, the Netherlands
 
ABSTRACT
 
Background. Previous studies have shown that health-related quality of life (HRQL) predicts survival in patients infected with human immunodeficiency virus (HIV). However, these studies predated the highly active antiretroviral therapy (HAART) era, included only a few patients receiving HAART, or had a limited duration of follow-up. This study investigates whether HRQL predicts survival among HIV-infected patients receiving HAART.
 
Methods. HIV-infected patients participating in the focus group of the AIDS Therapy Evaluation in the Netherlands (ATHENA) study and starting or already receiving HAART completed the Medical Outcomes Study HIV Health Survey at study entry (1 May 1998 through 31 December 2000). The physical health summary (PHS) and mental health summary (MHS) scores were calculated. All-cause mortality was established at 31 March 2008. Kaplan-Meier analysis and Cox regression models were performed to predict survival.
 
Results. The median follow-up was 8.4 years. Sixty-six patients (11.8%) died during follow-up. We found a significant relation between quartiles of PHS and survival (p<.001 , log-rank test). Of patients with a PHS, 26 (20%) died in quartile 1 (indicating worst HRQL), 17 (13%) died in quartile 2, 10 (8%) died in quartile 3, and 5 (4%) died in quartile 4 (indicating best HRQL) ( p<.001). The prediction of PHS on survival was independent of other (clinical) parameters (p<.001). No relation was found between MHS and survival (p=.13).
 
Conclusion. Patient-reported HRQL predicted survival among HIV-infected patients receiving HAART. This information could be highly useful for physicians in determining the prognosis of their patients.
 
Survival among patients infected with human immunodeficiency virus (HIV) has significantly improved since highly active antiretroviral therapy (HAART) became available in 1996 [1]. Yet, eradication of HIV infection cannot be achieved with current HAART regimens, and patients have to continue receiving HAART indefinitely. HAART regimens are now more effective, easier to take, and often better tolerated than the first HAART regimens. Still, patients are burdened with having to take lifetime treatment with potential adverse effects, which may diminish their health-related quality of life (HRQL). Measures of HRQL reflect the impact of both disease and treatment as perceived by the patient. Improving patients' HRQL is recognized by HIV treatment guidelines as 1 of the therapeutic objectives [2].
 
Evidence is accumulating to support the theory that HRQL predicts survival in several community samples and in patients with a variety of diseases [3-5]. To our knowledge, only 3 studies have investigated whether HRQL predicts survival among HIV-infected patients. These studies found that HRQL was a significant predictor of survival, even after adjustment for known clinical risk factors [6-8]. However, the relevance of these studies for today's clinical practice is uncertain because they predated the HAART era, had a limited duration of follow-up, or included only a few patients undergoing HAART.
 
If HRQL would independently predict mortality, physicians could consider HRQL in conjunction with clinical parameters when assessing the prognosis of HIV-infected patients. The aim of the present study was to investigate whether HRQL predicts survival in the national AIDS Therapy Evaluation in the Netherlands (ATHENA) cohort, which consists of HIV-infected patients receiving HAART with a long follow-up time.
 
Discussion
 
Patient-reported HRQL was a significant and independent predictor of survival among HIV-infected patients receiving HAART, with a median follow-up time of 8.4 years in the ATHENA cohort. The physical component of HRQL remained a significant predictor even after adjustment for demographic and clinical factors. Our findings suggest that HRQL could be used to assess the prognosis among HIV-infected patients in conjunction with demographic and clinical variables. These results might be of interest to both resource-rich and resource-poor countries.
 
The finding that HRQL predicts survival adds to evidence from other studies. A number of explanations have been proposed for why self-reported HRQL predicts survival over and above established demographic and clinical prognostic factors [3]. First, self-reported health status may capture the entire range of symptoms, including subtle signs of disease progression a person perceives. Second, whereas perceptions of poor health might lead to behaviors that positively influence health status, they predominantly have been found to induce behaviors that negatively influence health status, such as lesser involvement in preventive activities or self-care and nonadherence to screening recommendations, medication, and treatment. In addition, health perceptions may be influenced by resources in the external social environment, such as income and social networks, and within-person resources, such as control over health. Therefore, self-rated health is believed to reflect both perceptions and subsequent behaviors and resources [3].
 
It seems plausible that self-reported HRQL may influence survival in HIV infection, at least in part, via adherence to HAART. The association between suboptimal adherence to HAART and an increased risk of mortality is well documented [3]. There is only limited evidence on the relation between HRQL and subsequent adherence, with inconsistent results [16-18]. Several studies have established a relationship between a poorer self-reported mental health or a higher number of depressive symptoms and an increased risk of HIV disease progression and/or mortality [19-22]. Self-reported depressive symptoms have been found to affect survival both directly and via adherence [7, 19-22]. Our finding that the PHS but not the MHS predicted survival is therefore unexpected. Moreover, the subscales expected to measure depression to some extent (ie, mental health, health distress, quality of life, cognitive function, and vitality) were not predictive of survival. Our finding that the PHS but not the MHS predicted survival is consistent with 3 previous studies among HIV-infected patients that predate the HAART era or include only a few patients receiving HAART [6, 7, 23].
 
The subscales that were most predictive of survival were physical functioning, pain, role functioning, and social functioning. These subscales contribute to PHS and focus predominantly on limitations in usual daily activities, such as walking the stairs or carrying groceries. Self-reported PHS seems to be an inclusive and robust measure of health, reflecting health aspects relevant to survival that are not covered by common clinical indicators that we investigated in the present study. One could speculate that future studies might explore the possibility of stratifying patients on the basis of their baseline HRQL in clinical trials in which survival is an end point. Although PHS was independently predictive of survival, this does not imply a causal relationship. The mechanism by which self-reported HRQL influences survival remains to be determined [24]. Also, the post hoc analyses regarding causes of death did not reveal information about the mechanism because causes of death were too diverse and patient numbers were too small. Future research should determine the prognostic value of HRQL on survival in clinical settings. If this prognostic value would be established, it might be useful to provide patients with the prognostic HRQL information to encourage greater adherence.
 
An interesting finding is that failure to complete all items of the questionnaire, resulting in missing summary scores, was associated with an increased risk of mortality. Failure of completion might be due to fatigue or concentration problems and thereby indicative of a poor health status. In particular, failure to complete items about cognitive functioning increased the risk of mortality (HR, >20; ). These items focus on difficulties with reasoning and solving problems, forgetfulness, trouble paying attention, and difficulties with concentration and thinking.
 
Our study has several limitations. A percentage of our patients were already receiving HAART when they enrolled in ATHENA and completed the HRQL questionnaire. Ideally, we would have assessed HRQL in all patients at the start of HAART. Restricting our analysis to patients initiating HAART at enrollment in ATHENA would have reduced our sample size substantially. By investigating the effect on survival of the pre-HAART HIV RNA concentration, the pre-HAART CD4 cell count, and the use of antiretroviral treatment before starting HAART, we attempted to adjust for this limitation. In addition, we analyzed the prediction of HRQL on survival among those who were treatment naive at the start of this study (not displayed in results). The outcomes showed the same trend (PHS: HR for each 5 points, <1) as in the entire group but were not significant because of small numbers.
 
Inherent with the long follow-up period of our study, the HAART regimens do not reflect HAART regimens that are currently used. Current HAART regimens are more effective and often better tolerated and easier to take than those prescribed at the start of our study. Possibly, this may influence the association between self-reported HRQL and survival, because these newer regimens might have an effect on both HRQL and survival. However, knowing the relationship between HRQL and survival in patients following those HAART regimens remains of relevance because these patients still form a large percentage of today's population of HIV patients. Moreover, the HAART regimens used in the study period are still used in resource-poor settings.
 
Also inherent in our prolonged follow-up period is the homogeneous composition of our study population. Our study population mainly consisted of men having sex with men (76.1%), which reflects the composition of the HIV-infected population in the Netherlands at that time. Currently, heterosexuals, especially patients from ethnic minority groups, constitute a larger part of the HIV-infected population in the Netherlands [10]. Nevertheless, men having sex with men still form a substantial part of the population of HIV-positive patients. We therefore believe that our finding that HRQL predicts survival in this group is important. In line with the composition of the study population is the lack of data on other possible confounders, such as socioeconomic status and access to health care. Both have been found to strongly predict mortality. However, we think these factors might play a minor role in our study because the study sample consists of a homogenous group with an above average socioeconomic status. Also, in the Netherlands nearly everyone has health insurance and therefore access to health care. Future studies should corroborate our findings among other groups of patients and include measurement of other confounders that might predict survival, such as socioeconomic status.
 
We choose the MOS-HIV Health Survey for our study because it is the most widely used measure of HRQL in HIV-infected patients and has good reliability, validity, and responsiveness [11-13]. We previously demonstrated good reliability and validity of a Dutch version of the MOS-HIV Health Survey [25]. A potential limitation of the MOS-HIV Health Survey is a low sensitivity because it is a rather generic measure of HRQL. Nevertheless, the MOS-HIV Health Survey was a significant predictor of survival in the present study.
 
The most important strength of our study is the long follow-up period of more than 8 years, giving insight into the prediction in the long term. Other strengths include a large sample size, embedment of our HRQL study within the ATHENA cohort, which assesses a larger range of clinical measures than existing studies, and use of the clinically relevant and unambiguous end point all-cause mortality.
 
In summary, physical HRQL predicts survival in the long term in a population of HIV-infected patients receiving HAART independent of demographic and clinical variables. Future research should determine the prognostic value of HRQL on survival in clinical settings. Also, the mechanism by which physical HRQL influences survival remains to be determined. Potentially, this information could be highly useful for physicians in assessing the prognosis of their patients in resource-rich and resource-poor countries.
 
Results
 
Participants. Characteristics of patients at study entry are presented in Table 1. A total of 404 patients (72%) were already receiving HAART at study entry, and 156 patients (28%) started HAART at or just after study entry (data not shown). Sixty-six patients (11.8%) had died by 31 March 2008. The median survival time was 8.4 years (interquartile range, 7.4-8.9 years).
 
PHS and MHS scores and survival. Quartiles of PHS and MHS scores comprised 131 patients, except for quartile 1, which comprised 132 patients. For 35 patients, we could not calculate PHS and MHS scores because of missing values. Of patients with a PHS score, 26 (20%) died in quartile 1, 17 (13%) died in quartile 2, 10 (8%) died in quartile 3, and in 5 (4%) died in quartile 4 (p<.001). Of patients with an MHS score, 15 (11%) died in quartile 1, 17 (13%) died in quartile 2, 16 (12%) died in quartile 3, and 10 (8%) died in quartile 4 (p=.52). Of patients who had no summary scores, 8 (23%) died. We observed no differences in characteristics at study entry between patients with missing summary scores compared with patients with complete data. Figure 1 and Figure 2 display the Kaplan-Meier curves for patients in the different quartiles of PHS and MHS scores, including those who had no summary scores. A significant difference in survival was found among the quartiles of patients with a PHS score (p<.001) but not among the quartiles of patients with an MHS score (p=.13).
 
Twenty-two patients (33%) died of AIDS-related diseases, such as malignant tumors and infections, 12 (18%) of heart and lung conditions, 11 (17%) of non-AIDS-related malignant tumors, 6 (9%) of liver and pancreas conditions, 6 (9%) of suicide, accidents, or euthanasia, 4 (6%) of non-AIDS-related infections, and 5 (8%) of unknown conditions. There was no clear pattern of differences in PHS and MHS scores among the different causes of death.
 
Predictors of survival. Table 2 gives the results of the bivariate and multivariate Cox regression analyses for prediction of continuous scores of PHS and MHS and sociodemographic and clinical parameters on survival. The HRs indicate risk of death. The PHS predicted survival bivariately (HR for each 5 points, 0.8; 95% confidence interval [CI], 0.7-0.9), whereas the MHS did not significantly predict survival (HR for each 5 points, 0.9; 95% CI, 0.8-1.1; p=.13).
 
Of the 10 subscales, physical functioning (p<.001), pain (p<.001), role functioning (p<.001), social functioning (p<.001), and general health (p<.01) predicted survival. Health distress (p=.47), mental health (p=.30), vitality (p=.13), cognitive functioning (p=.13), and overall quality of life (p=.15) did not predict survival. When the summary scores were missing, the risk of death increased (HR, 2.3; 95% CI, 1.1-4.7). Lack of the following subscales increased the risk of death: cognitive functioning (p=.01), health distress (p=.03), vitality (p=.03), pain (p=.04), and mental health (p=.05).
 
Multivariate analysis showed that higher PHS score (HR for each 5 points, 0.8; 95% CI, 0.7-0.9) and no treatment with antiretroviral therapy before the start of HAART (HR, 0.1; 95% CI, 0.05-0.2) were protective for the risk of death. In contrast, higher age (HR for each 5 years, 1.3; 95% CI, 1.1-1.5), having a CD4 cell count less then 200 x 106/L at the start of HAART (HR, 2.1; 95% CI, 1.2-3.8), and having a detectable viral load at study entry (HR, 5.2; 95% CI, 2.8-9.4) increased the risk of death. All variables met the proportional hazards assumption (p>.05).
 
 
 
 
  iconpaperstack view older Articles   Back to Top   www.natap.org