Abstract
Low health-related quality of life (HRQOL) has been associated with increased risk for hospitalization and death in ESRD. However, the relationship of HRQOL with outcomes in predialysis CKD is not well understood. We evaluated the association between HRQOL and renal and cardiovascular (CV) outcomes in 1091 African Americans with hypertensive CKD enrolled in the African American Study of Kidney Disease and Hypertension (AASK) trial and cohort studies. Outcomes included CKD progression (doubling of serum creatinine/ESRD), CV events/CV death, and a composite of CKD progression or death from any cause (CKD progression/death). We assessed HRQOL, including mental health composite (MHC) and physical health composite (PHC), using the Short Form-36 survey. Cox regression analyses were used to assess the relationship between outcomes and five-point decrements in MHC and PHC scores using measurements at baseline, at the most recent annual visit (time-varying), or averaged from baseline to the most recent visit (cumulative). During approximately 10 years of follow-up, lower mean PHC score was associated with increased risk of CV events/CV death and CKD progression/death across all analytic approaches, but only time-varying and cumulative decrements were associated with CKD progression. Similarly, lower mean MHC score was associated with increased risk of CV events/CV death regardless of analytic approach, while only time-varying and cumulative decrements in mean MHC score was associated with CKD progression and CKD progression or death. In conclusion, lower HRQOL is associated with a range of adverse outcomes in African Americans with hypertensive CKD.
- AASK (African American Study of Kidney Disease and Hypertension)
- chronic kidney disease
- hypertension
- quality of life
Health-related quality of life (HRQOL) is diminished in patients with CKD and those with ESRD receiving hemodialysis compared with healthy individuals.1–3 In patients undergoing maintenance hemodialysis, low HRQOL is associated with a greater inflammatory state, poor nutritional status, increased hospitalization, and higher mortality.4–8 Although the relationship between HRQOL and adverse outcomes has been well studied in patients treated with hemodialysis,4,6,7 much less is known about the effect of HRQOL on patients with predialysis CKD.9–11 One exception is a small Taiwanese study of patients with CKD that found low baseline HRQOL scores to be associated with an increased risk of ESRD and death. However, similar studies are lacking in diverse CKD populations in the United States.10
Compared with whites, African Americans have a higher prevalence of CKD12 and are at increased risk for ESRD.13 Despite the magnitude of this health problem, HRQOL in this population has not been well studied. Prior studies have examined HRQOL among a cohort of African Americans with CKD enrolled in a randomized clinical trial of two levels of BP control targets and three antihypertensive drug regimens, the African American Study of Kidney Disease and Hypertension (AASK) trial. A cross-sectional analysis of African Americans with hypertensive CKD enrolled in the AASK trial reported that HRQOL scores, particularly the physical domain, were lower than the general population mean.14 However, the influence of HRQOL on clinical outcomes among this high-risk population is unknown.
To improve our understanding of the influence of HRQOL on health outcomes in adults with CKD, we examined the association between HRQOL and CKD progression, cardiovascular (CV) events, and all-cause death in a cohort of African Americans with hypertensive CKD with over 10 years of follow-up in the AASK trial and cohort studies.
Results
Participants and Characteristics
Patients completed the Medical Outcomes Study 36-Item Short Form (SF-36) quality of life (QOL) instrument at baseline and annual visits. At baseline, 1088 of 1094 participants (99.5%) had a mental health composite (MHC) score (mean baseline MHC score, 47.7±11.4). A total of 1085 of 1094 participants (99.2%) had a physical health composite (PHC) score (mean baseline PHC score, 43.5±10.9). Detailed baseline demographic and clinical characteristics of the cohort have been reported elsewhere.14 In brief, at baseline, the mean age±SD was 54.6±10.7 years, 39% of patients were female, the mean eGFR was 47.5±13.9 ml/min per m2, and the median protein-to-creatinine ratio was 0.08 g/g.
Cumulative Incidence and Rates of Outcomes
During 8.8–12.2 years of follow-up, the numbers of CV events/CV deaths, CKD progression events, and CKD progression events/deaths were 223, 419, and 563, respectively. Patients were censored at death. The rates of both CV events/CV death and CKD progression/death were significantly higher in participants with PHC scores in the lowest quartile than in those with scores in the highest quartile (4.04 versus 2.21/100 person-years for CV events and 8.92 versus 6.14/100 person-years for CKD progression/death) (Table 1). In contrast, the rates for CKD progression did not differ significantly by PHC quartiles (P>0.05). Moreover, none of the outcome rates varied significantly by MHC quartiles (P>0.05) (Table 1). However, an overall trend toward higher event rates in the lower quartiles of PHC and MHC was observed.
Cumulative incidence and rate of renal and cardiovascular events in MHC and PHC quartiles
Association between MHC Scores and Outcomes
In analyses of baseline, time-varying, and cumulative MHC scores (Table 2), we observed that for each five-point decrement in MHC, the risk for CV event/CV death was significantly increased (hazard ratios [HRs], 1.08, 1.13, and 1.13, respectively; P<0.05) after adjustment for other important factors. Similarly, in fully adjusted analyses, both lower time-varying and cumulative MHC score was associated with CKD progression (for both: HR, 1.07; P<0.05) and the composite outcome of CKD progression/death (HR, 1.09 and 1.08, respectively; P<0.05 for both) but not baseline score.
Association between MHC score and outcomes
Association between PHC Scores and Outcomes
Lower PHC score was associated with increased risk for CV events/CV death (Tables 3 and 4). In fully adjusted analyses, the lowest baseline PHC quartile was associated with almost a 2-fold higher risk of CV events/CV death compared with the highest quartile (HR, 1.94; P<0.05) (Table 4). For each five-point decrement in time-varying and cumulative PHC scores, the risk for the composite CV event increased, with HRs of 1.19 and 1.21, respectively (P<0.001) (Table 3). In fully adjusted analysis, both lower time-varying and cumulative PHC scores were associated with increased risk of CKD progression (HR, 1.12 and 1.10, respectively; P<0.05 for both), but baseline PHC score was not. For each five-point decrease in baseline, time-varying, and cumulative PHC scores, risk for the composite outcome of CKD progression/death was increased, with HRs of 1.06, 1.15, and 1.13 in fully adjusted analyses, respectively (P<0.05).
Association between PHC score and outcomes
Association between baseline PHC and CV events/CV death
Discussion
HRQOL is an important patient-centered outcome, and poor HRQOL is associated with adverse outcomes in a wide range of chronic diseases, including ESRD.4,5 Ours is the largest prospective study of QOL in patients with predialysis CKD. Among an exclusively African American cohort, we found that decrements in summary scores of the mental and physical components of the SF-36 (MHC and PHC) were consistently associated with increased risk of CV events/CV death across multiple methods of analysis. In addition, we found that lower PHC and MHC scores were associated with CKD progression in several analytic approaches (time-varying and cumulative analyses).
Because CVD is a major cause of death and morbidity in patients with CKD, a better understanding of the association between HRQOL and CV outcomes is critically important.15 The observed association between lower MHC and PHC scores and an increased risk of CV events may be related to several factors. As noted in studies of depression, poor mental health may be associated with a range of physiologic abnormalities that contribute to CVD, including changes in platelet function, dysregulation of the autonomic nervous system and hypothalamic-pituitary-adrenal axis, endothelial dysfunction, and inflammation.16 The factors underlying the association between poor physical health and CV outcomes are less clear. Similar to our findings, poor physical health has been shown to be an independent predictor of adverse outcomes both in the general population and in patients with ESRD.17 Poor self-reported physical health may be a proxy for significantly greater disease burden, which may increase risk for adverse CV outcomes. Alternatively, low PHC as well as MHC scores may each be associated with poor self-care, resulting in nonadherence to treatment regimens for CVD or its risk factors, thereby increasing the risk of adverse CVD outcomes.18
We also observed a relationship between worse MHC and PHC scores and the risk of CKD progression, although findings varied by analytic approach. Because risk factors for CV events and CKD progression risk are closely related,19 it is reasonable to speculate that the potential mechanisms discussed above regarding the association between HRQOL and CV outcomes could also potentially explain the association between HRQOL and CKD progression. Alternatively, reverse causality may underlie the stronger observed relationship between CKD progression and time-varying and cumulative HRQOL scores in contrast to baseline scores. Mental and physical disease burden may worsen in patients with declining kidney function, especially as they progress to ESRD.2 Similar to our findings, in a cohort of 423 Taiwanese patients with CKD, Tsai et al. observed an association between low HRQOL and increased rates of ESRD and death.10 In contrast to our findings, they found that baseline measures of HRQOL were predictive of progression to ESRD. This discrepancy may relate to differences in study populations, HRQOL instruments, and follow-up time.
Because we did not follow participants after reaching ESRD to ascertain death, our study included a composite outcome of CKD progression or death. Similar to CKD progression, most of our analytic approaches revealed a significant relationship between lower MHC and PHC scores and an increased risk of CKD progression or death. This is also consistent with the findings of the study by Tsai et al., in which low baseline HRQOL was associated with an increased risk of death.10 Furthermore, poor HRQOL is associated with increased mortality rates in patients with ESRD, further underscoring the importance of this association.4,6,7
Our study has several limitations. First, our cohort included only African Americans with hypertensive CKD, which may limit generalizability to patients with other causes of CKD and other racial/ethnic groups. However, because African Americans with hypertensive CKD are a large population at particularly high risk of progression to ESRD, the findings of this study have relevance to an important segment of the CKD population in the United States.13,20 Second, although this study was a prospective observational analysis, we cannot assess causality between HRQOL and the observed outcomes. However, observational studies are powerful tools to assess epidemiologic relationships, and we capitalized on complementary analytic techniques to robustly examine the relationship of HRQOL to clinically relevant outcomes.21 Third, despite robust risk adjustment, our study is subject to residual bias and confounding, as are other observational studies. Finally, these observations were made within the context of a randomized clinical trial and during a subsequent related observational study with a defined level of BP control, which may limit generalizability of these findings to the community setting.
In conclusion, low HRQOL measures were associated with increased risk of CKD progression and CV events in a large cohort of African American patients with CKD. These findings suggest that measurement of HRQOL has an important prognostic value for CKD patients in the clinical setting. Our findings underscore the need for future studies to build on these results with carefully designed translational studies to determine the mechanisms linking HRQOL and adverse outcomes and for interventions to potentially improve HRQOL and clinical outcomes.
Concise Methods
Study Design
We conducted a longitudinal analysis of African Americans with hypertensive CKD to examine the association of HRQOL with progression of CKD, occurrence of CV events, and death over a period of approximately 10 years. As previously reported, AASK had two phases: a randomized clinical trial conducted from 1995 to 2001 followed by a quasi-observational cohort study.22,23 The AASK trial included 1094 African Americans ages 18–70 years with hypertensive CKD (GFR, 20–65 ml/min per 1.73 m2) who were randomly assigned to one of two levels of BP control and to one of three different drug regimens to examine the effect of BP control on CKD progression.23 Individuals who did not develop ESRD during the trial were invited to join the cohort study. The latter was initiated in April 2002, at which point the 691 enrolled patients were switched from randomized therapy to ramipril and received standard protocol-driven BP management (<130/80 mmHg).22 Baseline and follow-up HRQOL measurements were available for 1091 participants, who make up the final analytic cohort for these analyses. All study participants provided written informed consent, and the institutional review boards of the participating centers approved the study.
Variables and Data Sources
Patients provided demographic and clinical information at enrollment. Variables included age, sex, marital status, income, insurance status, level of education, and comorbid medical conditions. eGFR was calculated using a formula derived from data from study participants.24 Urine protein-to-creatinine ratio was also assessed at baseline for each participant.22
HRQOL was measured annually with the Medical Outcomes SF-36, which is a generic questionnaire for QOL.25,26 The survey includes individual scale scores of the following eight domains: physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health. These eight scales can be combined as summary measures: PHC and MHC. The mean score for PHC and MHC is 50 in the general population, and lower scores are consistent with worse PHC and MHC.
Outcomes
The primary outcomes were (1) progression of CKD (defined as doubling of serum creatinine from baseline or development of ESRD), (2) a CV composite consisting of CV events or CV death, and (3) a composite outcome of progression of CKD or all-cause death. ESRD was defined by the initiation of dialysis or receipt of a kidney transplant. CV events were myocardial infarction, new-onset or worsened coronary heart disease, new-onset or worsened congestive heart failure, new-onset or worsened peripheral artery disease, and stroke. Each CV event was adjudicated by a subgroup of study investigators unaware of treatment assignment (trial phase).27
Statistical Analyses
Event cumulative incidences by percentages and event rates were characterized as per 100 person-years, and both were reported overall and by quartiles of MHC and PHC scores. Cox proportional hazards regression analyses were used to assess the association between baseline MHC and PHC scores and each of the outcomes in iterative models. Model 1 adjusted solely for randomized group; model 2 adjusted additionally for age, sex, and baseline eGFR; and model 3 adjusted additionally for proteinuria and history of CV disease. Additionally, time-dependent Cox regression was used to relate the hazard ratio for the same outcomes to the most recent assessment preceding each follow-up time point (i.e., time-varying) and the cumulative average of assessments preceding each follow-up time point (i.e., cumulative) among participants with at least one MHC or PHC during follow-up. The sample sizes for the time-dependent analyses (n=1094 for MHC and 1091 for PHC) exceeded the sample size for analyses of baseline MHC and PHC (n=1088 for MHC and 1085 for PHC) because the former analyses included patients with missing baseline MHC and PHC scores as long as they had at least one follow-up SF-36 measurement.
The assumption of proportional hazards in the Cox regression models was checked using Schoenfeld residuals for all included covariates. Significant violations of the proportional hazards assumption were found for baseline eGFR with the CKD progression/ESRD. Therefore, a linear interaction term between baseline eGFR and follow-up time was added to the corresponding models.
The assumption of linearity in the Cox regression models was checked using restricted cubic smoothing splines for all included covariates. Baseline PHC was found to have a significant nonlinear relationship with the CV composite outcome. Therefore, baseline PHC was categorized into quartiles. The effect of baseline PHC on the CV composite outcome was presented as the hazard ratio between the first (lowest), second, and third quartile group versus the fourth quartile group (highest), respectively. The effect of baseline PHC on the CV composite outcome was also presented as a nonparametric smoothing curve wherein the HR and its 95% confidence interval between the hazard of CV composite outcome when PHC was at a certain level and the hazard of CV composite outcome when PHC was 46 (median level) was plotted.
Disclosures
K.N. has consulted with Amgen, Pfizer, Merck, King Pharmaceuticals, and Abbott; has received grants from the National Institutes of Health and King Pharmaceuticals; and has received honoraria from Amgen.
Acknowledgments
AASK was supported by grants to each clinical center and the coordinating center from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). In addition, AASK was supported by the Office of Research in Minority Health (now the National Center on Minority Health and Health Disparities) and the following institutional grants from the National Institutes of Health: M01-RR00080, M01-RR00071, M01-00032, P20-RR11145, M01-RR00827, M01-RR00052, 2P20-RR11104, RR029887, and DK2818-02. King Pharmaceuticals provided monetary support and antihypertensive medications to each clinical center. Pfizer, Inc., AstraZeneca Pharmaceuticals, GlaxoSmithKline, Forest Laboratories, Pharmacia, and Upjohn also donated antihypertensive medications. J.P.L. was supported by NIDDK K24-DK092290. The project described was supported by Award Number KM1CA156717 (A.P.) from the National Cancer Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
The authors would like to thank the AASK Collaborative Research Group which includes the following institutions: Case Western Reserve University (Principal Investigators, Jackson T. Wright, Jr., Mahboob Rahman; Study Coordinator, Renee Dancie, Louise Strauss); Emory University (Principal Investigator, Janice Lea; Study Coordinators, Beth Wilkening, Arlene Chapman, Diane Watkins); Harbor-UCLA Medical Center (Principal Investigator, Joel D. Kopple; Study Coordinators, Linda Miladinovich, Jooree Choi, Patricia Oleskie, Connie Secules); Harlem Hospital Center (Principal Investigator, Velvie Pogue; Study Coordinator, Donna Dowie, Jen-Tse Cheng); Howard University (Principal Investigator, Otelio Randall, Tamrat Retta; Study Coordinators, Shichen Xu, Muluemebet Ketete, Debra Ordor, Carl Tilghman); Johns Hopkins University (Steering Committee Chair, Lawrence Appel; Principal Investigators, Edgar Miller, Brad Astor; Study Coordinators, Charalett Diggs, Jeanne Charleston, Charles Harris, Thomas Shields); Charles R. Drew University (Principal Investigators, Keith Norris, David Martins; Study Coordinators, Melba Miller, Holly Howell Laurice Pitts); Medical University of South Carolina (Principal Investigator, DeAnna Cheek; Study Coordinator, Deborah Brooks); Meharry Medical College (Principal Investigators, Marquetta Faulkner, Olufemi Adeyele; Study Coordinators, Karen Phillips, Ginger Sanford, Cynthia Weaver); Morehouse School of Medicine (Principal Investigators, William Cleveland, Kimberly Chapman; Study Coordinators, Winifred Smith, Sherald Glover); Mount Sinai School of Medicine and University of Massachusetts (Principal Investigators, Robert Phillips, Michael Lipkowitz, Mohammed Rafey; Study Coordinators, Avril Gabriel, Eileen Condren, Natasha Coke); Ohio State University (Principal Investigators, Lee Hebert, Ganesh Shidham; Study Coordinators, Leena Hiremath, Stephanie Justice); University of Chicago (Principal Investigators, George Bakris, James Lash; Study Coordinators, Linda Fondren, Louise Bagnuolo, Janet Cohan, Anne Frydrych); University of Alabama, Birmingham (Principal Investigators, Stephen Rostand, Denyse Thornley-Brown; Study Coordinator, Beverly Key); University of California, San Diego (Principal Investigators, Francis B. Gabbai, Daniel T. O'Connor; Study Coordinator, Brenda Thomas); University of Florida (Principal Investigators, C. Craig Tisher, Geraldine Bichier; Study Coordinators, Cipriano Sarmiento, Amado Diaz, Carol Gordon); University of Miami (Principal Investigators, Gabriel Contreras, Jacques Bourgoignie, Dollie Florence-Green; Study Coordinators, Jorge Junco, Jacqueline Vassallo); University of Michigan (Principal Investigators, Kenneth Jamerson, Akinlou Ojo, Tonya Corbin; Study Coordinators, Denise Cornish-Zirker, Tanya Graham, Wendy Bloembergen); University of Southern California (Principal Investigators, Shaul Massry, Miroslav Smogorzewski; Study Coordinators, Annie Richardson, Laurice Pitts).
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
- Copyright © 2014 by the American Society of Nephrology