Abstract
Objective. To compare the health-related quality of life (HRQOL) and the functional ability by race in patients with gout.
Methods. In a 9-month prospective cohort multicentre study, patients with gout self-reported race, dichotomized as Caucasian or African American (others excluded). We calculated HRQOL/function scores adjusted for age, study site and college education for Short Form-36 (SF-36; generic HRQOL), Gout Impact Scale (GIS; disease-specific HRQOL) and HAQ-disability index (HAQ-DI; functional ability). Longitudinally adjusted scores were computed using multivariable mixed-effect regression models with a random patient effect and fixed sequential visit effect (3-monthly visits).
Results. Compared with Caucasians (n = 107), African Americans (n = 60) with gout were younger (61.1 vs 67.3 years) and had higher median baseline serum urate (9.0 vs 7.9 mg/dl) (P < 0.01). African Americans with gout had worse HRQOL scores on three SF-36 domains, the mental component summary (MCS) and two of the five GIS scales than Caucasians [mean (s.e.); P ⩽ 0.02 for all]: SF-36 mental health, 39.7 (1.1) vs 45.2 (0.9); SF-36 role emotional, 42.1 (4.2) vs 51.4 (4.2); SF-36 social functioning, 36.0 (1.1) vs 40.0 (0.9) (P = 0.04); SF-36 MCS, 43.2 (3.1) vs 50.0 (3.2); GIS unmet treatment need, 37.6 (1.6) vs 31.5 (1.4); and GIS concern during attacks, 53.3 (3.7) vs 47.4 (3.7). Differences between the respective HAQ-DI total scores were not statistically significant; 0.98 (0.1) vs 0.80 (1.0) (P = 0.11). Racial differences in SF-36 mental health, role emotional and MCS scales exceeded, and for HAQ-DI approached, the minimal clinically important difference thresholds.
Conclusions. African Americans with gout have significantly worse HRQOL compared with Caucasians. Further research is necessary in the form of studies targeted at African Americans on how best to improve these outcomes.
Keywords: health-related quality of life, HRQOL, function, gout, race, disparity, racial
Rheumatology key messages
African Americans with gout were demonstrated to have significantly worse generic mental and emotional health–related quality of life compared with Caucasians.
African Americans with gout had more functional limitations than Caucasians in unadjusted analyses, and this was statistically significant and clinically meaningful.
The adjusted functional limitation difference by race approached the minimal clinically important difference threshold of 0.22.
Introduction
Gout is the most common inflammatory arthritis in adults and it affects ∼8.3 million Americans [1]. Gout was associated with an estimated annual health care cost of $20 billion in the USA in 2006 [2]; costs were higher in patients with higher serum urate [3] or tophi [4]. Gout is frequently associated with comorbidities [5–7].
Racial disparities in gout epidemiology, medication adherence and health services utilization have been reported [8]. In the USA, gout prevalence is 1.3-fold and gout incidence 1.7-fold higher in African Americans compared with in Caucasians [1, 9]. This higher risk is partially attributable to higher rates of hypertension [9], obesity, diabetes and renal failure in African Americans [10]. Compared with Caucasians with gout, African Americans with gout had higher baseline serum urate (7.9 vs 7.1 mg/dl) [11], a 2.6-fold higher rate of emergency room visits/hospitalizations for gout [11], and 1.86-fold higher odds of being non-adherent with urate-lowering therapy (ULT) [12]; they were less likely to receive allopurinol (odds of 0.18) [13] or ULT in general (27 vs 39%) [11], or to achieve a target serum urate of <6 mg/dl [11]. Thus, not only do African Americans have a higher prevalence of gout, they are also less likely than Caucasians to achieve serum urate <6 mg/dl, a key target for gout treatment [14]. This indicates a higher disease burden in African Americans with gout.
Health-related quality of life (HRQOL) and functional disability have been recognized as core domains for studies of gout in international consensus [15, 16]. Gout is associated with lower HRQOL [17–19] and functional limitation [19, 20]. To our knowledge, study of racial differences in HRQOL and functional ability in gout patients has received little attention. Leading organizations have called for elimination of racial disparities in health care [21, 22]. Our objective was to assess whether HRQOL and functional ability in gout patients differs by patient race. In a prospective, multicentre, US cohort study, we examined the association between race and HRQOL and functional ability in patients with gout, adjusting for important factors. We hypothesized that only some differences will be attributable to age and education level differences by race.
Methods
We report study methods and results following the recommendations from the Strengthening of Reporting in Observational studies in Epidemiology statement [23]. The Institutional review boards of the Birmingham and Greater Los Angeles Veterans Affairs (VA) medical centres approved the study. All investigations were conducted in conformity with ethical principles of research.
Study design, setting and eligibility
This was a prospective, cohort multicentre US study conducted at two institutions. Patients >18 years of age with a documented diagnosis of gout were identified, using the electronic medical records at two medical centres, the VA Greater Los Angeles Healthcare System and the Birmingham VA Medical Center. Eligible gout patients were contacted via phone and invited to participate in the study. Patients were also recruited from outpatient primary care, rheumatology and other specialty clinics at the two sites; those identified by screening during clinic visits and those responding to study flyers. Patients were screened for eligibility during a clinic visit and those who met eligibility criteria, that is, adults 18–85 years with a physician diagnosis of gout who were able to provide consent were enrolled in the study. All patients also met the 1977 preliminary classification criteria for gout [24]. Study participants were evaluated every 3 months for four study visits (0, 3, 6 and 9 months).
Study assessments
During each study visit, patients completed a paper-based survey that included the following: generic HRQOL assessment with Short Form 36 (SF-36) [25]; disease-specific HRQOL with the Gout Impact Scale (GIS) of the Gout Assessment Questionnaire [26]; functional ability assessment with HAQ Disability Index (HAQ-DI) [27, 28]; patient global assessment [15, 29]; and the University of California at San Diego health care utilization survey. Patients also underwent physician assessment at each visit that included a physician global assessment of gout [30]. At each study visit, patients also self-reported comorbidities, which were used to calculate the Charlson index, a weighted index of 17 comorbidities, expressed as a summative score (where a higher score indicates more comorbidity) that has been validated [31]. Age and BMI (kg/m2) were extracted from the records on the day of the screening visit.
SF-36 is a widely used, self-administered general HRQOL instrument consisting of 36 items, summarized into 8 HRQOL domains (0–100; higher is better) and two summary scores: the physical component summary (PCS) and the mental component summary (MCS) scores, which are population- and norm-based, with a mean (s.d.) of 50 (10) [25, 32, 33]. Eight SF-36 domains are physical functioning, role physical, bodily pain, general health, vitality, role emotional (RE), mental health (MH) and social functioning. Minimal clinically important difference (MCID) is 5–10 points on domain scores and 2.5–5 on summary scale scores [34–37]; we used 5 and 2.5 points as thresholds, respectively.
GIS, a disease-specific HRQOL instrument, measures the impact of gout on HRQOL, both during and between acute gout attacks [26]. The GIS includes five scales measuring the potential impact of gout on patient’s lives: overall gout concern, gout medication side effects, unmet treatment needs, well-being during attacks and gout concern during attacks [26]. MCID is available for four scales and ranges 5–8 units: gout concern overall, 7.2; unmet gout treatment need, 6.9; gout well-being during attack, 5.2; and gout concern during attack, 7.6 [38].
Functional ability of the patients was assessed with the HAQ-DI [27, 28]. The HAQ-DI has questions within eight sections leading to eight scale scores: dressing and grooming, arising, eating, walking, hygiene, reach, grip and common daily activities. Scoring within each section is from 0 (without any difficulty) to 3 (unable to do). The score given to that section is the worst score within the section, that is, if one question is scored 1 and another 2, then the score for the section is 2. If an aid or device is used or if help is required from another individual, then the minimum score for that section is 2. The scores of the eight sections are summed and divided by 8, and the overall score ranges 0–3 (0 indicating no disability and 3 indicating extreme disability). The MCID for HAQ-DI is 0.22 [39].
Patients reported overall gout severity using a 10 cm visual analogue scale [15, 29]. Physicians also recorded global assessment of overall gout severity using a 10 cm visual analogue scale [30], and the presence or absence of tophi.
Independent variable and outcomes
The independent variable of interest was patient-reported race at the first visit. Race was dichotomized into African American and Caucasian (other races were excluded for analytic purposes), as determined a priori.
Outcomes of interest were HRQOL on SF-36 (eight domain scores, PCS and MCS; generic HRQOL) and GIS scales (five scales and overall score; disease-specific HRQOL), and functional limitation assessed by HAQ-DI (overall score and eight sections/domains). We compared these scores both cross-sectionally at baseline and longitudinally over the four study visits over 9 months.
Statistical analysis
Continuous variables were compared between African Americans and Caucasians with gout using the t test or Wilcoxon rank sum test, depending on the normality of the distribution of the data. Categorical variables were compared with the Chi-square test where expected frequencies were larger than five, and the Fisher exact test where expected frequencies were five or fewer. Mean differences in HRQOL and functional limitation measures were compared with MCID thresholds to assess clinical significance. Adjusted SF-36 t-score values, GIS scores and HAQ-DI scores were computed for African Americans and Caucasians with gout using the least squares means method (i.e. marginal means) with adjustment for key factors including age, study site and education. Multivariable linear regression models were used to compute the cross-sectional baseline-adjusted scores (see supplementary data, available at Rheumatology Online). Multivariable mixed-effect regression models with a random patient effect and fixed sequential visit effect were used to compute the longitudinally adjusted scores. We used variance components as the covariance structure applied to the longitudinal models, which model a different variance component for each patient. We considered a race and time interaction term for the longitudinal models, but this term was not statistically significant and was thus dropped from the analyses. Analyses were conducted using SAS software (SAS Institute, Cary, NC, USA). A P < 0.05 was considered statistically significant.
Results
Patient characteristics
A total of 186 patients were enrolled in the study, 112 at the Greater Los Angeles VA and 74 patients at the Birmingham VA. Of these, 107 were Caucasian and 60 were African American. The majority of the patients were male (97.9%), the mean (s.d.) age was 64.6 years (10.9) and 64% had at least some college level education (Table 1). Almost one-fifth had tophi, and the median serum urate was 7.9 mg/dl. Median physician and patient gout severity assessments were 3.0 and 6.0, respectively (0–10 scale). Comorbidities were common: hypertension (81.4%), diabetes (36.2%), congestive heart failure (15.4%), moderate or severe renal failure (24.4%) and hyperlipidaemia (58.9%; Table 2).
Table 1.
Baseline characteristics of study cohort
| Total cohort, n = 186 | Los Angeles (LA), n = 112 | Birmingham (BHAM), n = 74 | LA vs BHAM, P-values | White, n = 107 | African American (AA), n = 60 | White vs AA, P-values | |
|---|---|---|---|---|---|---|---|
| Age, mean (s.d.), years | 64.6 (10.9) | 65.7 (11.2) | 62.9 (10.2) | 0.09a | 67.3 (10.9) | 61.1 (9.3) | <0.01 |
| Charlson Comorbidity Index, mean (s.d.) | 5.0 (2.8) | 5.2 (3.1) | 4.7 (2.1) | 0.22a | 5.0 (2.8) | 5.3 (2.9) | 0.60a |
| Serum urate, median (IQR), mg/dl | 7.9 (3.5) | 7.5 (3.0) | 9.7 (8.2) | <0.01b | 7.9 (3.4) | 9.0 (5.7) | 0.03b |
| Patient gout severity assessment VAS, median (IQR), 0–10 cm | 6.0 (5.0) | 6.5 (4.5) | 4.0 (5.0) | 0.09b | 5.5 (5.0) | 6.5 (5.5) | 0.27b |
| Physician gout severity assessment VAS, median (IQR), 0–10 cm | 3.0 (4.0) | 3.0 (3.5) | 2.6 (5.0) | 0.73b | 3.0 (4.0) | 3.0 (4.0) | 0.53b |
| VA site, n (%) | |||||||
| Los Angeles, California | 112 (60.2) | – | – | – | 76 (71.0) | 17 (28.3) | <0.01c |
| Birmingham, Alabama | 74 (39.8) | – | – | – | 31 (28.9) | 43 (71.7) | |
| Education | |||||||
| No college degree | 134 (72.8) | 79 (71.8) | 55 (74.3) | 0.71c | 76 (72.4) | 44 (73.3) | 0.89c |
| College degree | 50 (27.2) | 31 (28.2) | 19 (25.7) | 29 (27.6) | 16 (26.7) | ||
| Education, n (%) | |||||||
| High school or less | 46 (25.9) | 25 (22.7) | 21 (28.4) | 0.39c | 32 (30.5) | 9 (15.0) | 0.04c |
| Some college | 138 (75.0) | 85 (77.3) | 53 (71.6) | 73 (69.5) | 51 (85.0) | ||
| Male, gender, n (%) | 182 (97.9) | 109 (97.3) | 73 (98.7) | >0.99d | 105 (98.1) | 59 (98.3) | 0.99d |
| Race, n (%) | |||||||
| American Indian | 1 (0.5) | 1 (0.9) | 0 (0.0) | <0.01d | – | – | – |
| Asian | 6 (3.2) | 6 (5.4) | 0 (0.0) | – | – | ||
| Pacific Islander | 1 (0.5) | 1 (0.9) | 0 (0.0) | – | – | ||
| African American | 60 (32.3) | 17 (15.2) | 43 (58.1) | – | – | ||
| White | 107 (57.5) | 76 (67.9) | 31 (41.9) | – | – | ||
| Other | 11 (5.9) | 11 (9.8) | 0 (0.0) | – | – | – | |
| Hispanic or Latino, n (%) | 21 (13.0) | 20 (18.7) | 1 (1.8) | 0.02d | 12 (12.2) | 1 (2.3) | 0.06d |
| Tophi diagnosed, n (%) | 37 (19.9) | 35 (31.3) | 2 (2.7) | <0.01d | 26 (26.3) | 4 (7.3) | <0.01d |
Bold represents characteristics in which the differences between groups, either by site or race, are statistically significant.
t-test;
Wilcoxon Rank Sum test;
Chi-square test;
Fisher exact test. IQR: interquartile range; VA: veterans association; VAS: visual analogue scale.
Table 2.
Baseline patient comorbidity characteristics
| Total cohort, n = 186 | Los Angeles (LA), n = 126 | Birmingham (BHAM),n = 74 | LA vs BHAM, P-values | White | African American (AA) | White vs AA, P-values | |
|---|---|---|---|---|---|---|---|
| Charlson comorbidities | n (%) | n (%) | n (%) | n (%) | n (%) | ||
| Myocardial infarction | 30 (17.1) | 19 (17.0) | 11 (17.2) | 0.97 | 19 (18.6) | 8 (14.6) | 0.51 |
| Congestive heart failure | 27 (15.4) | 18 (16.2) | 9 (14.1) | 0.70 | 16 (15.8) | 9 (16.4) | 0.93 |
| Peripheral vascular disease | 20 (11.4) | 16 (14.3) | 4 (6.3) | 0.14 | 16 (15.7) | 3 (5.5) | 0.07a |
| Cerebrovascular disease | 13 (7.4) | 13 (11.6) | 0 (0.0) | <0.01a | 9 (8.8) | 2 (3.6) | 0.33a |
| Dementia | 2 (1.1) | 2 (1.8) | 0 (0.0) | 0.53a | 2 (1.9) | 0 (0.0) | 0.54a |
| Chronic pulmonary disease | 28 (15.8) | 15 (13.4) | 13 (20.0) | 0.24 | 17 (16.5) | 9 (16.4) | 0.98 |
| CTD | 3 (1.7) | 3 (2.7) | 0 (0.0) | 0.55a | 1 (1.0) | 2 (3.6) | 0.28a |
| Ulcer disease | 22 (12.5) | 22 (19.6) | 0 (0.0) | <0.01a | 10 (9.8) | 10 (18.2) | 0.13 |
| Diabetes | 64 (36.2) | 46 (41.1) | 18 (27.7) | 0.07 | 34 (33.0) | 21 (38.2) | 0.52 |
| Diabetes with end organ damage | 21 (11.9) | 20 (17.9) | 1 (1.5) | <0.01a | 12 (11.7) | 6 (10.9) | 0.89 |
| Moderate or severe renal disease | 43 (24.4) | 24 (21.4) | 19 (29.7) | 0.22 | 26 (25.2) | 16 (29.6) | 0.56 |
| Hemiplegia | 8 (4.5) | 2 (1.8) | 6 (9.2) | 0.05a | 3 (2.9) | 5 (9.1) | 0.13a |
| Leukaemia | 0 (0.0) | 0 (0.0) | 0 (0.0) | – | 0 (0.0) | 0 (0.0) | – |
| Lymphoma | 0 (0.0) | 0 (0.0) | 0 (0.0) | – | 0 (0.0) | 0 (0.0) | – |
| Any tumour | 28 (15.8) | 14 (12.5) | 14 (21.5) | 0.11 | 15 (14.6) | 13 (23.6) | 0.15 |
| Metastatic cancer | 0 (0.0) | 0 (0.0) | 0 (0.0) | – | 0 (0.0) | 0 (0.0) | – |
| Mild liver disease | 10 (5.7) | 9 (8.2) | 1 (1.5) | 0.09a | 8 (7.8) | 2 (3.70) | 0.50a |
| Moderate or severe liver disease | 0 (0.0) | 0 (0.0) | 0 (0.0) | – | 0 (0.0) | 0 (0.0) | – |
| AIDS | 2 (1.1) | 1 (0.9) | 1 (1.5) | >0.99a | 1 (1.0) | 1 (1.8) | 0.99a |
| Other comorbidities | |||||||
| Depression | 54 (30.5) | 39 (34.8) | 15 (23.1) | 0.10 | 28 (27.2) | 18 (32.7) | 0.47 |
| Hypertension | 144 (81.4) | 98 (87.5) | 46 (70.8) | <0.01 | 87 (84.5) | 44 (80.0) | 0.48 |
| Skin ulcers/cellulitis | 5 (2.8) | 5 (4.5) | 0 (0.0) | 0.16a | 4 (3.9) | 0 (0.0) | 0.30a |
P-value using Fisher exact test; other P-values were obtained using the Chi-square test.
We noted statistically significant differences in patient characteristics by race (Table 1): compared with Caucasians, African Americans with gout were significantly younger, had lower education level, higher serum urate levels and lower prevalence of diagnosed tophi.
Generic HRQOL in gout by race
At baseline (cross-sectional analyses), in unadjusted analyses, African Americans with gout had lower mean scores (indicating poor health) on three SF-36 domains and the MCS relative to Caucasians, and these differences were statistically significant: MH, 40.8 vs 46.3 (P < 0.01); RE, 31.3 vs 40.6 (P < 0.01); social functioning, 37.3 vs 41.5 (P = 0.04); and MCS, 39.5 vs 46.2 (P < 0.01) (supplementary Table S1, available at Rheumatology Online). These differences met or exceeded the MCID, except for SF-36 social functioning. SF-36 role physical domain scores had a trend towards statistical significance, 33.8 vs 37.5, respectively (P = 0.06). PCS scores and other SF-36 domain scores were similar by race.
Cross-sectional analyses adjusted for age, site and education showed that the differences in SF-36 MH and social functioning by race were no longer statistically significant or clinically meaningful. The adjusted SF-36 RE score was lower in African Americans (P = 0.01), and this was both statistically significant and clinically meaningful (difference, 7.5; exceeded the MCID of 5); the MCS was lower (difference, 4.4; exceeded the MCID of 2.5) and this was clinically meaningful and had a non-significant statistical trend (P = 0.07; supplementary Table S2, available at Rheumatology Online).
Longitudinal analyses adjusted for age, site and education showed that five of the eight SF-36 domain scores (physical functioning, bodily pain, MH, RE and social functioning) and MCS were lower in African Americans compared with Caucasians (Table 3) and these differences were statistically significant; differences in the MH and RE domains and MCS scores also exceeded clinically important difference thresholds, that is, MCID.
Table 3.
Longitudinal short form-36 domain and component summary scores adjusted for age, site, education and visits (n = 164)
| White | African American | P-values | Difference in means for MCID comparisona | |
|---|---|---|---|---|
| Least square means | Least square means | |||
| (s.e.) | (s.e.) | |||
| SF-36 domains | ||||
| Physical functioning | 38.62b | 35.73b | 0.02b | −2.89b |
| (0.96)b | (1.11)b | |||
| Role physical | 38.83 | 36.50 | 0.07 | −2.33 |
| (0.97) | (1.13) | |||
| Bodily pain | 40.46b | 35.76b | 0.0001b | −4.70b |
| (0.93)b | (1.07)b | |||
| General health | 44.73 | 45.01 | 0.60 | 0.28 |
| (0.41) | (0.48) | |||
| Mental health | 45.21 | 39.68 | <0.0001 | −5.55 |
| (0.91) | (1.06) | |||
| Role emotional | 51.45 | 42.14 | <0.0001 | −9.31 |
| (4.24) | (4.16) | |||
| Social functioning | 40.02b | 36.04b | 0.001b | −3.98b |
| (0.92)b | (1.06)b | |||
| Energy/vitality | 48.18 | 46.47 | 0.08 | −1.71 |
| (1.88) | (1.87) | |||
| PCS | 39.08 | 38.03 | 0.34 | −1.05 |
| (0.86) | (0.99) | |||
| MCS | 50.02 | 43.24 | <0.0001 | −6.78 |
| (3.22) | (3.15) |
SF-36 is a generic HRQOL measure with higher scores indicating better health. Score differences that exceed both clinically meaningful and statistically significant thresholds are in bold.
Difference in means = (AA mean score − White mean score).
Score differences that are statistically significant, but do not exceed MCID. The MCID is 5-points on SF-36 domain scores and 2.5 points on summary scale scores, MCS and PCS [34–37]. HRQOL: health-related quality of life; MCID: minimal clinically important difference; MCS: mental component summary; PCS: physical component summary; SF-36: Short Form 36.
Gout-specific HRQOL by race
At baseline (cross-sectional analyses), compared with Caucasians, African Americans with gout scored higher (i.e. worse HRQOL) in four of the five GIS domains, including gout concern overall (P = 0.04), unmet treatment need (P = 0.01), well-being during attacks (P = 0.01) and concern during attacks (P = 0.01). Differences exceeded the MCID for all four scales. Concerns about medication side effects did not differ by race with statistical significance. The overall GIS score was higher in African Americans, and this was statistically significant (P < 0.01; supplementary Table S3, available at Rheumatology Online).
Cross-sectional analyses adjusted for age, site and education showed that the difference in only one of the five GIS scales, unmet treatment need, was worse/higher in African Americans compared with Caucasians, and this was both statistically significant (P = 0.03) and clinically meaningful (difference, 6.9) (supplementary Table S4, available at Rheumatology Online).
In longitudinal analyses adjusted for age, site and education, two GIS domains (unmet treatment need and concern during attacks) and overall GIS scale scores were worse/higher in African Americans compared with Caucasians (Table 4); these differences approached but did not exceed threshold for MCID, but they were statistically significant.
Table 4.
Longitudinal gout impact scale scores by race adjusted for age, site, some college education and visits (n = 164)
| White | African American (AA) | P-values | Difference in means for MCID comparisona | |
|---|---|---|---|---|
| Least square means (s.e.) | Least square means (s.e.) | |||
| GIS scales | ||||
| Gout concern overall | 62.99 (4.16) | 66.40 (4.19) | 0.17 | 3.41 |
| Medication side effects | 51.26 (2.15) | 50.77 (2.49) | 0.86 | −0.49 |
| Unmet treatment need | 31.48 (1.39)b | 37.59 (1.60) b | 0.0009b | 6.11b |
| Well-being during attack | 34.02 (6.42) | 38.64 (6.30) | 0.06 | 4.62 |
| Concern during attack | 47.39 (3.66)b | 53.33 (3.73)b | 0.02b | 5.94b |
| Overall GIS score | 42.58 (4.32) | 46.94 (4.25) | 0.014 | 4.36c |
GIS is a disease-specific HRQOL measure, with a higher score indicating worse health.
Difference in means = (AA mean score − White mean score); positive numbers indicate higher scores in the AA group, that is, worse GIS scores. MCID ranges 5–8 U on GIS scales: gout concern overall, 7.2; unmet gout treatment need, 6.9; gout well-being during attack, 5.2; and gout concern during attack, 7.6 [38].
Score differences that meet only statistical significance, but do not exceed MCID thresholds.
No MCID is available for overall GIS score; therefore, no comparisons could be made regarding MCID. GIS: Gout Impact Scale; HRQOL: health-related quality of life; MCID: minimal clinically important difference.
Functional limitation in gout by race
At baseline (cross-sectional analyses), the median total HAQ-DI score was higher, indicating worse disability in African Americans as compared with Caucasians, 0.92 vs 0.67 (P = 0.02; difference of means, 0.25) that also exceeded the MCID of 0.22 (supplementary Table S5, available at Rheumatology Online). African Americans had lower scores in five of the eight HAQ-DI activity domains, including dressing and grooming, arising, eating, walking and grip (no MCIDs were available for domain scores); these differences were statistically significant.
Cross-sectional analyses adjusted for age, site and education showed that there were statistically significant differences (P = 0.01 each) in only three HAQ-DI activity domains: dressing and grooming, walking, and grip, were lower in African Americans compared with Caucasians (supplementary Table S6, available at Rheumatology Online; no MCIDs were available for domain scores). The overall HAQ-DI scores were lower in African Americans than Caucasians (P = 0.04; supplementary Table S6, available at Rheumatology Online); these differences were both statistically significant and clinically meaningful.
In longitudinal analyses adjusted for age, site and education, there were statistically significant differences in five HAQ-DI activity domains: dressing and grooming, eating, hygiene, reach and grip (P < 0.0001, 0.0001, 0.03, 0.04 and 0.004) (Table 5). The difference in the overall HAQ-DI scores by race was not statistically significant, and it approached but did not exceed the threshold of clinical importance (difference of 0.18 vs MCID of 0.22).
Table 5.
Longitudinal HAQ disability index scores by race adjusted for age, site, some college education and visit (n = 163)
| White | African American | P-values | |
|---|---|---|---|
| Least square means (s.e.) | Least square means (s.e.) | ||
| HAQ sections | |||
| Dressing and grooming | 0.44 (0.05)a | 0.90 (0.06)a | <0.0001a |
| Arising | 0.75 (0.08) | 0.70 (0.08) | 0.53 |
| Eating | 0.34 (0.08)a | 0.77 (0.10)a | 0.0001a |
| Walking | 0.82 (0.09) | 0.85 (0.09) | 0.70 |
| Hygiene | 0.79 (0.09)a | 0.59 (0.10)a | 0.03a |
| Reach | 0.86 (0.08)a | 0.70 (0.08)a | 0.04a |
| Grip | 0.62 (0.08)a | 0.86 (0.09)a | 0.004a |
| Activities | 2.23 (0.54) | 2.21 (0.62) | 0.97 |
| HAQ-DI composite score | 0.80 (0.09) | 0.98 (0.11) | 0.12 |
Higher HAQ-DI score indicates worse health. The MCID for the HAQ-DI composite score was 0.22 [39]. The difference, i.e., the MCID in the overall HAQ-DI score was not statistically significant, although with a difference of 0.18, it approached clinically meaningful difference, as indicated by the MCID of 0.22 [39].
Score differences that meet statistical significance (no MCID thresholds have been defined for HAQ sections/domains). HAQ-DI: HAQ disability index; MCID: minimal clinically important difference.
Discussion
To our knowledge, this is the first multicentre, prospective, cohort study that has compared HRQOL and disability between African Americans and Caucasians with gout, and has adjusted for important covariates/confounders. Previous studies have examined predictors of overall worse HRQOL in predominantly Caucasian samples [17, 40]. No comparisons have been made with minorities and most were cross-sectional, except for one validation study [20]. Given that elimination of health care disparities is a national priority [21, 22], several novel study findings merit further discussion.
An important observation in our study was that African Americans with gout had worse unadjusted mental/emotional HRQOL scores compared with Caucasians with gout, and these differences were also clinically meaningful, that is, met or exceeded the MCID thresholds, in most cases. On the other hand, none of the physical HRQOL domain scores showed any significant difference, statistically or clinically. This is an important observation and indicates a higher mental and emotional HRQOL disease burden, but not physical HRQOL burden, in African Americans with gout compared with Caucasians. These results are concordant with the result of a qualitative study of HRQOL in gout with 10 nominal groups stratified by race and sex, which found that African Americans with gout ranked concern about the effect of gout on emotional health high more often than Caucasians [41]. So what are the underlying reasons for this?
In our analyses adjusted for age, education and study site, there was some attenuation of racial differences in these SF-36 domain scores, but some differences persisted. This supported our hypothesis that racial differences in age and education only partially explain racial differences in generic HRQOL in patients with gout. Whether the remaining differences are true differences by race or whether they can be explained by unmeasured differences in socio-economic status, social support and health care access across race remains to be determined. Nevertheless, our study shows that in multivariable-adjusted analyses, African Americans with gout had larger mental/emotional HRQOL deficits compared with Caucasians.
Unadjusted disease-specific HRQOL was also worse on four of the five GIS scales in African Americans, indicating more overall concern about gout, more concern during a gout attack, and higher unmet need and lower well-being scores in African Americans with gout. These differences attenuated after adjustment for age, education, etc., and were below the MCID, but they remained statistically significant. This indicated that the higher disease impact on gout-specific HRQOL in African Americans than Caucasians is explained partially due to differences in age, education, etc. The higher unmet need reported by African Americans with gout in our study might explain previously reported higher utilization of resources by African Americans with gout compared with that of Caucasians [11]. This study finding offers an important insight into gout-related health care burden differences by race/ethnicity. It remains to be tested whether the lower allopurinol use and ULT adherence in African Americans compared with Caucasians reported previously [11–13] contributes to a higher impact of gout-specific HRQOL in African Americans.
Another important observation in our study was that African Americans with gout were more disabled than their Caucasian counterparts. The difference in unadjusted disability scores was both statistically significant and clinically meaningful. Five of the eight HAQ-DI scale and the overall HAQ-DI scores were lower in African Americans in unadjusted analyses. There are no prior studies to our knowledge against which to compare these findings, and therefore our findings add to the current knowledge and provide a foundation for future studies. More suboptimal gout treatment and worse disease outcomes have been reported in African Americans compared with in Caucasians in previous studies [11–13]. Analyses adjusted for differences in age, education and site attenuated HAQ-DI scores differences only partially, and the overall HAQ-DI score was still worse in African Americans as compared with Caucasians (0.23 higher/worse for baseline-adjusted and 0.18 higher/worse for longitudinally adjusted scores; MCID = 0.22). Therefore, racial differences in HAQ-DI scores may either be attributed to race itself or to unmeasured factors such as socio-economic status, social support, health care access, etc.
Our study has several strengths. Most studies of HRQOL or functional limitation in gout have been cross-sectional, with the exception of one study examining the reproducibility of HAQ-DI in a Mexican cohort [20]. Prospective cohort design, multicentre study and the use of standardized, validated instruments allowed for systematic collection of valid data. We recruited patients from primary care and rheumatology clinics to increase the generalizability of our findings. The gout diagnosis was physician-reported and confirmed in accordance with the 1977 gout classification criteria, and not patient self-reported, making it more likely to be valid. HRQOL was assessed using both generic and disease-specific measures. Adjustment for several potential confounders increased the rigor of our study.
Our study has several limitations as well. The primarily male sample in our study implies that findings may not be generalizable to women with gout, and there is a growing female subset with elderly-onset gout. Veterans have higher comorbidity and poorer HRQOL than the general US population [42, 43]; therefore, these findings are likely generalizable to men in the general population with higher comorbidity. We likely missed patients with mild disease who do not see physicians frequently, skewing our results to reflect findings among more symptomatic gout patients. Our results were based on patients’ responses and are therefore subject to recall bias; however, the instruments used in this study are widely used in health services research, and we did not alter the recall period for the questions, making these data as valid as other studies using these assessments and comparable with them. A ceiling effect was noted with some GIS scale domains (gout concern overall, 17%; medication side effects, 9%; concern during attack, 6%) and a floor effect was noted with all HAQ domains (25–55%), which must be considered while interpreting these findings. We found that 22–25% of patients had missing data during follow-up visits for one or more variables, and therefore these patients were excluded from longitudinal analyses. We acknowledge that there was some indication of heteroscedasticity for SF-36 and GIS scores, even though the residuals were generally normally distributed. The findings should be interpreted considering this limitation.
Conclusions
In conclusion, this prospective, multicentre cohort study showed that in unadjusted analyses, compared with Caucasians, African Americans with gout had significantly worse generic HRQOL, disease-specific HRQOL and functional limitation; most differences were also clinically meaningful. Specifically, compared with Caucasians, African Americans with gout expressed higher disease concern and unmet need and lower mental/emotional HRQOL and functional ability to perform daily tasks. Age and education level explained these racial differences only partially, and many differences remained statistically significant in adjusted analyses, indicating either an independent association of race or unmeasured confounding due to differences in socio-economic status, social support and health care access (factors not measured in this study) by race. Further research is needed into the determinants of these racial differences to target interventions to modifiable mediators of these relationships. Efforts are also needed to reach out to disadvantaged minorities, such as African Americans with gout, to improve disease outcomes in gout.
Supplementary Material
Acknowledgements
D.K. was supported by the National Institutes of Health NIAMS K24 AR063120 (Outcomes Research in Rheumatic Diseases) and P.P.K. was supported by an ACR Clinical Investigator Fellowship Award.
Funding: This study was sponsored in part by an investigator-initiated study from Savient Pharmaceuticals Inc.
Disclosure statement: J.A.S. has received research grants from Takeda and Savient and consultant fees from Savient, Takeda, Regeneron, Iroko, Merz, Bioiberica, Crealta and Allergan pharmaceuticals and WebMD and UBM LLC, serves as the principal investigator (PI) for an investigator-initiated study funded by Horizon pharmaceuticals through a grant to DINORA, Inc. (a 501c3 entity), is a member of the executive of OMERACT, an organization that receives arms-length funding from 36 companies, is a member of the ACR’s Guidelines Subcommittee of the Quality of Care Committee and a member of the Veterans Affairs (VA)’s Rheumatology Field Advisory Committee. D.K. has received consultancy fees from Astra Zeneca and Takeda Pharmaceuticals and is an investigator on a planned investigator–initiated trial of Pegloticase in gout (PI: PP Khanna). P.P.K. serves as PI on industry-sponsored open-label extension trials from Astra-Zeneca and a phase 4 clinical trail by Crealta; he has received research funding on an investigator-initiated study from Astra-Zeneca and is PI on a planned investigator–trial of Pegloticase sponsored by Crealta; he has also received consultancy fees from Astra Zeneca and Takeda Pharmaceuticals and is funded by the ACR–Rheumatology Research Foundation for the development of Quality Measures in gout and serves as coordinating PI for the nationwide VA CRYSTAL Registry. The other authors have declared no conflicts of interest.
Supplementary data
Supplementary data are available at Rheumatology Online.
References
- 1.Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia in the US general population: the National Health and Nutrition Examination Survey 2007–2008. Arthritis Rheum 2011;63:3136–41. [DOI] [PubMed] [Google Scholar]
- 2.Wertheimer AI, Morlock R, Becker MA. A revised estimate of the burden of illness of gout. Curr Ther Res Clin Exp 2013;75:1–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Park H, Rascati KL, Prasla K, McBayne T. Evaluation of health care costs and utilization patterns for patients with gout. Clin Ther 2012;34:640–52. [DOI] [PubMed] [Google Scholar]
- 4.Khanna PP, Nuki G, Bardin T. et al. Tophi and frequent gout flares are associated with impairments to quality of life, productivity, and increased healthcare resource use: results from a cross-sectional survey. Health Qual Life Outcomes 2012;10:117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.MacFarlane LA, Kim SC. Gout: a review of nonmodifiable and modifiable risk factors. Rheum Dis Clin North Am 2014;40:581–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McAdams-DeMarco MA, Maynard JW, Baer AN, Coresh J. Hypertension and the risk of incident gout in a population-based study: the atherosclerosis risk in communities cohort. J Clin Hypertens 2012;14:675–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Burke BT, Köttgen A, Law A. et al. Physical function, hyperuricemia, and gout in older adults. Arthritis Care Res 2015;67:1730–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Singh JA. Racial and gender disparities among patients with gout. Curr Rheumatol Rep 2013;15:307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hochberg MC, Thomas J, Thomas DJ. et al. Racial differences in the incidence of gout. The role of hypertension. Arthritis Rheum 1995;38:628–32. [DOI] [PubMed] [Google Scholar]
- 10.Maynard W, Janet A, McAdams MN. et al. Racial differences in gout risk and uric acid levels in both men and women in the Atherosclerosis Risk in Communities (ARIC) study. [abstract]. Arthritis Rheum 2010; 62(Suppl 10):1556. [Google Scholar]
- 11.Coley K, Saul M, Pater K. Relationship between race, uric acid levels, urate-lowering therapy and resource use in patients with gout. Arthritis Rheum 2012;64(10 Suppl): S772. [Google Scholar]
- 12.Solomon DH, Avorn J, Levin R, Brookhart MA. Uric acid lowering therapy: prescribing patterns in a large cohort of older adults. Ann Rheum Dis 2008;67:609–13. [DOI] [PubMed] [Google Scholar]
- 13.Krishnan E, Lienesch D, Kwoh CK. Gout in ambulatory care settings in the United States. J Rheumatol 2008;35: 498–501. [PubMed] [Google Scholar]
- 14.Khanna D, Fitzgerald JD, Khanna PP. et al. 2012 American College of Rheumatology guidelines for management of gout. Part 1: systematic nonpharmacologic and pharmacologic therapeutic approaches to hyperuricemia. Arthritis Care Res 2012;64:1431–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Singh JA, Taylor WJ, Simon LS. et al. Patient-reported outcomes in chronic gout: a report from OMERACT 10. J Rheumatol 2011;38:1452–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Taylor WJ, Schumacher HR, Jr, Baraf HS. et al. A modified Delphi exercise to determine the extent of consensus with OMERACT outcome domains for studies of acute and chronic gout. Ann Rheum Dis 2008;67:888–91. [DOI] [PubMed] [Google Scholar]
- 17.Singh JA, Strand V. Gout is associated with more comorbidities, poorer health-related quality of life and higher healthcare utilisation in US veterans. Ann Rheum Dis 2008;67:1310–6. [DOI] [PubMed] [Google Scholar]
- 18.Roddy E, Zhang W, Doherty M. Is gout associated with reduced quality of life? A case–control study. Rheumatology 2007;46:1441–4. [DOI] [PubMed] [Google Scholar]
- 19.Khanna D, Ahmed M, Yontz D. et al. The disutility of chronic gout. Qual Life Res 2008;17:815–22. [DOI] [PubMed] [Google Scholar]
- 20.Alvarez-Hernandez E, Pelaez-Ballestas I, Vazquez-Mellado J. et al. Validation of the Health Assessment Questionnaire disability index in patients with gout. Arthritis Rheum 2008;59:665–9. [DOI] [PubMed] [Google Scholar]
- 21.Smedley BD, Stith AY, Nelson AR, eds. Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care. Unequal treatment: confronting racial and ethnic disparities in health care. Washington, DC: National Academies Press, 2003: 782 http://www.nap.edu/catalog/10260.html (24 September 2016, date last accessed). [PubMed] [Google Scholar]
- 22.U.S. Department of Health & Human Services. 2011 National Healthcare Quality and Disparities Reports. Rockville, MD: Agency for Healthcare Research and Quality, 2014. http://archive.ahrq.gov/research/findings/nhqrdr/nhqrdr11/qrdr11.html (24 September 2016, date last accessed). [Google Scholar]
- 23.STROBE Statement. Checklist of items that should be included in reports of cohort studies. Strengthening the Reporting of Observational Studies in Epidemiology. Bern, Germany: University of Bern, 2007. http://www.strobe-statement.org/fileadmin/Strobe/uploads/checklists/STROBE_checklist_v4_cohort.pdf (24 September 2016, date last accessed). [Google Scholar]
- 24.Wallace SL, Robinson H, Masi AT. et al. Preliminary criteria for the classification of the acute arthritis of primary gout. Arthritis Rheum 1977;20:895–900. [DOI] [PubMed] [Google Scholar]
- 25.Ware JE, Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30:473–83. [PubMed] [Google Scholar]
- 26.Hirsch JD, Terkeltaub R, Khanna D. et al. Gout disease-specific quality of life and the association with gout characteristics. Patient Relat Outcome Meas 2010;1:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fries JF, Spitz PW, Young DY. The dimensions of health outcomes: the health assessment questionnaire, disability and pain scales. J Rheumatol 1982;9:789–93. [PubMed] [Google Scholar]
- 28.Liang M, Schurman DJ, Fries J. A patient-administered questionnaire for arthritis assessment. Clin Orthop Relat Res 1978;131:123–9. [PubMed] [Google Scholar]
- 29.Singh JA, Yang S, Strand V. et al. Validation of pain and patient global scales in chronic gout: data from two randomised controlled trials. Ann Rheum Dis 2011;70: 1277–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Schlesinger N, De Meulemeester M, Pikhlak A. et al. Canakinumab relieves symptoms of acute flares and improves health-related quality of life in patients with difficult-to-treat Gouty Arthritis by suppressing inflammation: results of a randomized, dose-ranging study. Arthritis Res Ther 2011;13:R53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613–9. [DOI] [PubMed] [Google Scholar]
- 32.Ware JE, Jr, Kosinski M, Gandek B. SF-36 health survey: manual & interpretation guide. Lincoln, RI: QualityMetric; Boston, MA: Health Assessment Lab, 2000. [Google Scholar]
- 33.Ware JE, Kosinski M, Dewey J. How to score Version Two of the SF-36 Health Survey. 3rd edn Lincoln, RI: QualityMetric, 2000. [Google Scholar]
- 34.Kosinski M, Zhao SZ, Dedhiya S, Osterhaus JT, Ware JE., Jr. Determining minimally important changes in generic and disease-specific health-related quality of life questionnaires in clinical trials of rheumatoid arthritis. Arthritis Rheum 2000;43:1478–87. [DOI] [PubMed] [Google Scholar]
- 35.Strand V, Bombardier C,AM. Use of minimum clinically important differences [MCID] in evaluating patient responses to treatment of RA [abstract]. Arthritis Rheum 2001;44(Suppl):S187. [Google Scholar]
- 36.Zhao SZ, Fiechtner JI, Tindall EA. et al. Evaluation of health-related quality of life of rheumatoid arthritis patients treated with celecoxib. Arthritis Care Res 2000;13:112–21. [DOI] [PubMed] [Google Scholar]
- 37.Tugwell P, Wells G, Strand V. et al. Clinical improvement as reflected in measures of function and health-related quality of life following treatment with leflunomide compared with methotrexate in patients with rheumatoid arthritis: sensitivity and relative efficiency to detect a treatment effect in a twelve-month, placebo-controlled trial. Leflunomide Rheumatoid Arthritis Investigators Group. Arthritis Rheum 2000;43:506–14. [DOI] [PubMed] [Google Scholar]
- 38.Khanna D, Sarkin AJ, Khanna PP. et al. Minimally important differences of the gout impact scale in a randomized controlled trial. Rheumatology 2011;50:1331–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Wells GA, Tugwell P, Kraag GR. et al. Minimum important difference between patients with rheumatoid arthritis: the patient’s perspective. J Rheumatol 1993;20:557–60. [PubMed] [Google Scholar]
- 40.Chandratre P, Roddy E, Clarson L. et al. Health-related quality of life in gout: a systematic review. Rheumatology 2013;52:2031–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Singh JA. The impact of gout on patient’s lives: a study of African-American and Caucasian men and women with gout. Arthritis Res Ther 2014;16:R132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Agha Z, Lofgren RP, VanRuiswyk JV, Layde PM. Are patients at Veterans Affairs medical centers sicker? A comparative analysis of health status and medical resource use. Arch Intern Med 2000;160:3252–7. [DOI] [PubMed] [Google Scholar]
- 43.Kazis LE, Miller DR, Clark J. et al. Health-related quality of life in patients served by the Department of Veterans Affairs: results from the Veterans Health Study. Arch Intern Med 1998;158:626–32. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.