|Home | About | Journals | Submit | Contact Us | Français|
The health disparities of Native Hawaiians and Pacific Islanders (NHPI) are well established for diabetes and cardiovascular disease, but less is known about disparities in arthritis. This study examined possible disparities in the prevalence of arthritis by age, sex, and severity comparing NHPI, Whites, and Asians. The study population included adult Hawai‘i participants in the 2013 Behavioral Risk Factor Surveillance Survey. NHPI males had a significantly higher prevalence of arthritis, which peaked twenty years earlier, than White and Asian males (P<.001). The prevalence of arthritis peaked at 65–79 years in males and females in all racial groups, except in NHPI males where it peaked at 45–54 years. The mean ages (years) for males with arthritis were 46.2 for NHPI, 59.1 for Whites, and 60.5 for Asians; the respective ages for females were 54.2, 60.5, and 58.8. NHPI males body mass index averaged 2.4 kg/m2 greater than White males (P<.001), and obese NHPI males had twice the age-adjusted odds of arthritis than obese White males. Although NHPI females had a greater body mass index than White females (P=.05), the prevalence of arthritis was only slightly and not significantly higher. NHPI males and females reported high pain scores more frequently than Whites did, but the differences did not reach statistical significance. Diabetes was a comorbidity more than twice as often in NHPI and Asians of both sexes than among Whites. This study demonstrated racial disparities in the prevalence of arthritis among NHPI, Whites, and Asians.
Arthritis affects 22.7% of US adults, or 52.5 million people,1 and is the most common cause of disability with 43.2% of those with arthritis reporting activity limitations.2 Inherent risk factors for arthritis include advancing age,2–4 female sex,1,2,4 and genetic conditions.5–10 The prevalence of arthritis has been reported to be 7.3% in those 18–44 years of age, 30.3% in the 45–64 age group, and 49.7% in those 65 years and older.2 Twenty six percent of females and 19.1% of males have ever been told by a physician that they have arthritis.2
Modifiable risk factors for arthritis include obesity,3,11–15 occupation,16,17 and joint trauma18,19 or infection.20 Obesity has been associated with a higher risk of osteoarthritis.3,11–15 Studies have demonstrated that being overweight preceded the onset of osteoarthritis in the knee11,12 and increased the rate of its radiographic progression.13,14 Also, weight loss has been shown to decrease knee osteoarthritis in women.15 Occupations with repetitive manual tasks have a higher rate of hand arthritis16 and those with heavy lifting or frequent knee bending have a higher rate of knee and hip arthritis.17 Joint trauma from a fracture18 or athletics19 has been associated with an increased risk of osteoarthritis. A variety of infectious pathogens can cause both acute and chronic arthritis.20 Diabetes and heart disease are common comorbidities associated with arthritis.1 Forty seven percent of adults with diabetes and 49% with heart disease are reported to have arthritis.1
The reasons for racial and ethnic health disparities are complex, but genetic, behavioral, environmental, cultural, and socioeconomic factors may contribute. Health disparities in Native Hawaiians have been reported in diabetes,21–23 obesity,23–27 ischemic28 and hemorrhagic stroke,29 and cardiovascular disease.27,30 Racial and ethnic differences in the prevalence, treatment, and outcome of different forms of arthritis have been reported.31–35 African Americans have a higher prevalence, more severe disease, and poorer outcomes due to systemic lupus erythematous than Whites.31,35 Non-Hispanic Blacks, Hispanics, and multiracial groups have been shown to have higher arthritis-attributable activity, and work limitations and more severe joint pain than non-Hispanic Whites with arthritis.32 However, previous studies have combined Asians and Pacific Islanders into one group.32 Similarly, prior studies and surveys have not routinely identified Native Hawaiians and Other Pacific Islanders (NHPI) as a distinct racial or ethnic grouping. This study analyzed the prevalence of arthritis by age, sex, and severity among NHPI, Whites, and Asians in Hawai‘i using 2013 Behavioral Risk Factor Surveillance System (BRFSS) data.
The study population was adult participants residing in Hawai‘i surveyed in the 2013 BRFSS who self-identified as NHPI, White, or Asian race. The study employed the dataset available nationally through the Centers for Disease Control and Prevention (CDC).36 The study population was restricted to those who answered that they either were or were not diagnosed with arthritis in response to the question asking if a doctor had ever told them that they had some form of arthritis (eg, arthritis, rheumatoid arthritis, gout, lupus, or fibromyalgia). Eligible ages ranged from 18 to 79. The study was designed to compare the prevalence of arthritis by race, age, sex, and categories of body mass index (BMI). A secondary objective was to assess possible racial differences in the prevalence of diabetes among participants with arthritis.
Race was the preferred race selected by participants. BMI was analyzed categorically: underweight (12 to <18.5 kg/m2), normal weight (18.5 to < 25 kg/m2), overweight (25 to < 30 kg/m2), and obese (>30 kg/m2). The presence of diabetes was based on the answer to a question asking participants if a doctor had ever told them they had diabetes. Age and sex were used as reported by participants. BRFSS includes three questions on the burden of arthritis: (1) Are you now limited in any way in any of your usual activities because of arthritis or joint symptoms?; (2) Do arthritis or joint symptoms now affect whether you work, the type of work you do, or the amount of work you do?; (3) During the past 30 days, to what extent has your arthritis or joint symptoms interfered with your normal social activities, such as going shopping, to the movies, or to religious or social gatherings? Answers for social limitations were categorized as none, a little, or a lot. The questionnaire also asked participants with arthritis to rank their joint pain in the past 30 days on a scale from 0 to 10 describing 0 as no pain or aching and 10 as pain or aching as bad as it can be. For analyses, the pain scores were categorized as five or higher versus 0 to 4.
Characteristics of the study participants are summarized by percentages and standard errors within race and gender categories. All analyses were performed accounting for the complex survey design by using the primary sampling units, strata, and weights provided by CDC. Differences in prevalence by race and gender were evaluated by chi-square tests. The analyses employed logistic regression to examine race differences by body mass category with adjustment for age. Models included indicator variables for combinations of race and sex using the combinations of White females or White males as the reference category. As an example, one analysis compared obese Asian and NHPI females to obese White females. A similar analysis was employed to model the prevalence of diabetes by sex and race groupings among people with arthritis. Limitations in activities due to arthritis comparing NHPI and Asians to Whites were analyzed employing separate models for females and males. Age-adjusted logistic regression models were used for the questions on activities with binary (yes/no) answers and multinomial logistic regression for the question on social activities with three outcome categories. Results of logistic regression models are reported as odds ratios with 95% confidence intervals.
All analyses were performed using SAS version 9.3 and accounted for the complex survey design.
The study results are based on the responses of 6,735 participants. Table 1 summarizes the characteristics of the study population. For both females and males, NHPI had the greatest proportions in the youngest age groups. NHPI males also had the highest prevalence of obesity at 49.4%, compared to Whites (22.7%) and Asians (17.9%). NHPI female obesity was 36.7%, compared to Whites (20.2%) and Asians (12.9%).
Prevalence of arthritis varied by age, race, and sex (Figure 1). NHPI males exhibited the highest prevalence across the adult age span, significantly greater than Whites and Asians (P<.001). At the peak age range (45–54 years), arthritis prevalence among NHPI males was 49.4% compared to White males (22.2%) and Asian males (17.9%). The prevalence for NHPI males remained high — from 40% to 50% — at the oldest ages. Arthritis prevalence among White and Asian males increased with age, coming closer to the prevalence of NHPI males at ages 55 and older. Arthritis prevalence did not differ significantly among females but increased with age for all three races; differences were less than 15% by age groups and smallest at the oldest ages.
Prevalence of arthritis varied by weight status (Table 2). Obese NHPI males had twice the age-adjusted odds of having arthritis compared to obese White males (P=.045). The BMI of obese NHPI males was on average 2.4 kg/m2 greater than obese White males (P <.001). Among females, obese White females had the highest prevalence of arthritis. The age-adjusted odds of arthritis for NHPI were about 50% lower than White females (P=.03). The BMI of NHPI females averaged 1.3 kg/m2 greater than White females (P=.05).
Limitations from arthritis were not significantly different when comparing NHPI males and females to Whites (Table 3). Asian females, however, had 60% lower odds of limitations in usual activities, in the type or amount of work, and in social activities (P-values <.05).
Differences in joint pain among those with arthritis did not vary significantly by sex and race. However, NHPI tended to report greater than average pain scores compared to Asians and Whites (Table 4). Although not reaching statistical significance, NHPI males and females had odds ratios for above average pain 60%–70% higher than Whites.
Among participants with arthritis, the age-adjusted odds of diabetes were two to three times greater among NHPI and Asians of both sexes than among Whites (Table 5, P-values <.05).
This study provides evidence that NHPI adults are more likely to have arthritis than Whites and Asians. This disparity may be attributed predominantly to the higher prevalence of arthritis among NHPI males. A contributing factor may be the significantly higher rate of obesity in NHPI males, a finding consistent with other studies.22,24–27 Obesity rates were significantly higher in both NHPI males and females compared to Whites and Asians in this study, however, the difference was much greater in NHPI males. Although there may be a reciprocal relationship between obesity and arthritis, a high BMI and bone mineral density have been shown to increase the risk of osteoarthritis, the most common form of arthritis.37,38 Obese NHPI males in this study were twice as likely to have arthritis than obese White males, but obese NHPI females were half as likely to have arthritis than obese White females. The reason for this sex disparity is unclear. Furthermore, Asians had the same odds of having arthritis in all weight categories, suggesting that BMI alone is not causative. BMI has also been shown to increase with increasing percentage of NHPI ancestry which suggests a genetic component.23
The average age at diagnosis of arthritis in NHPI was significantly lower than Whites and Asians. This may be attributed mostly to NHPI males whose average age at diagnosis was thirteen years younger than Whites and fourteen years younger than Asians. Compared to White and Asian females, NHPI females were four and six years younger, respectively. The prevalence of arthritis is known to increase with advancing age,2–4 peaking in this study at 65–79 years for White and Asian males and all females, but twenty years earlier in NHPI males. Despite peaking earlier, NHPI males continued to have a higher prevalence of arthritis than Whites and Asians at older ages. The decreasing prevalence of arthritis with advancing age in NHPI males over 55 years could be due to a shorter life span related to chronic illnesses. When combined with a shorter life expectancy,38 an earlier onset of arthritis may result in fewer quality adjusted life years. The reason for the earlier peak in the prevalence of arthritis in NHPI males is unclear. Further research into the possible causes of this age disparity is indicated.
A younger age of onset has also been reported in NHPI with ischemic28 and hemorrhagic stroke29 and cardiovascular mortality.40 In these studies, NHPI also had higher rates of diabetes and hypertension which are known risk factors for stroke41,42 and heart disease.41–43 In the current study, NHPI and Asians had a significantly higher prevalence of diabetes than Whites, but Whites had a higher prevalence of arthritis than Asians. Likewise, NHPI had a significantly higher prevalence of arthritis than Asians but not a significantly higher prevalence of diabetes. Additional studies are needed to more thoroughly explore the association between diabetes and arthritis.
NHPI have been previously reported to experience a more severe level of disability than other ethnic groups, most commonly attributing their disabling condition to stroke, whereas Japanese and Whites most commonly cited arthritis.44 In this study, both NHPI males and females, despite having the highest pain scores, demonstrated no significant difference in activity limitations attributed to arthritis compared to White males and females.
This study has several limitations. This was a cross-sectional study so causation cannot be established. Age adjusted analysis helped reduce prevalence bias. Self-identification with a specific racial/ethnic group does not allow the blood percentage of a particular ethnicity of the respondent to be determined. However, self-selection of a racial/ethnic group is the accepted method in determining race/ethnicity in such surveys. Our data was also based on very general race categories, particularly for Asians. Variations among ethnic groups might affect overall prevalence data. The Hawai‘i Department of Health collects state-level BRFSS data which separates Native Hawaiians from Pacific Islanders and separates Asians into specific ethnic groups, but this increases the number of groups and decreases the number in each group for analysis. Future studies with larger populations of distinct ethnic groups may reveal disparities by specific ethnicities.
BRFSS data is based on self-reporting which introduces the possibility of reporting bias. The validity of self-reported prevalence of arthritis using BRFSS data has been shown to be sensitive and highly reliable.45,46 Self-reported height has been found to be significantly overestimated and self-reported weight significantly underestimated, which would result in a lower calculated BMI.47 This would suggest that the prevalence of obesity may be higher than reported. Variations in reporting by race or ethnic group could also influence BMI.
This study included only Hawai‘i residents. According to the US Census 2010, 55% of NHPI, or 289,970 people, reside in Hawai‘i. About one-third of NHPI in the continental US reside in California.48 The 2010 Hawai‘i Health Survey cited 291,223 NHPI living in Hawai‘i.49 The investigators chose to analyze NHPI residing in Hawai‘i because Hawai‘i has the largest population of NHPI and self-reporting as NHPI should be the most accurate of the states surveyed.
The different types of arthritis and the joints involved could not be distinguished using 2013 BRFSS data. Future research with more detailed patient data will be important in determining whether there are specific arthritic conditions and joint locations that are more prevalent in NHPI.
This study demonstrated racial disparities in the prevalence of arthritis among NHPI, Whites, and Asians. NHPI adult males have a significantly higher prevalence of arthritis than White and Asian adult males in all age groups, and arthritis in this population peaks twenty years earlier than in other groups. Obesity may be a contributing factor.
In order to prevent or eliminate health disparities, they must first be identified. This is the first study to analyze health disparities in the prevalence of arthritis among NHPI. Future research into potential causal relationships and specific types of arthritis are warranted.
This study was supported by funds from the Bureau of Health Professions, Health Resources and Services Administration, Department of Health and Human Services grant D34HP16044, Native Hawaiian Center of Excellence, National Center on Minority Health and Health Disparities grant S21MD000228, and Myron P. Thompson Endowed Chair Program for research in Native Hawaiian Health. JD was partially supported by grants U54MD007584 and 5 U54 GM104944 from the National Institute on Minority Health and Disparities and grant P20GM103466 from the National Institute of General Medical Sciences from the National Institutes of Health. The authors also acknowledge guidance from the Native Hawaiian Health Summer Research Internship Program.
Kyle Obana is a junior Biology and Psychology double major at Amherst College in Amherst, Massachusetts. He was recently awarded the Kauffman Fellowship in Biomedical Research and will be conducting research in the Department of Orthopedic Surgery at the Columbia College of Physicians and Surgeons in the summer of 2016. He was coauthor of a publication addressing the willingness of young adults in Hawai‘i to favor aggressive care following severe traumatic brain injury. Having been born and raised in Hawai‘i, he aspires to attend medical school and practice medicine in Hawai‘i. He volunteers in the Amherst community and plays defensive back on the Amherst College football team.
Through the Department of Native Hawaiian Health Summer Research Internship program in 2015, he was fortunate to work with James Davis PhD, Associate Professor in Biostatistics & Quantitative Health Sciences. Consistent with his interest in arthritis and Native Hawaiian health, Kyle chose to study whether there is a racial disparity among Native Hawaiians, Whites, and Asians with arthritis.
Health disparities in Native Hawaiians and Pacific Islanders (NHPI) are well established for diabetes and cardiovascular disease, but less is known about disparities in arthritis. Arthritis is a chronic disease that affects millions of Americans and is the leading cause of disability.
This study examined possible disparities in the prevalence of arthritis by age, sex, and severity comparing NHPI, Whites, and Asians. The study population included 6,735 Hawai‘i adult participants in the 2013 Behavioral Risk Factor Surveillance Survey. This study found that NHPI adults are more likely to have arthritis than White and Asians. This disparity can be attributed mostly to the higher prevalence of arthritis among NHPI males. Obesity may be a contributing factor, since obesity rates were significant higher in NHPI males and females. The average age at diagnosis of arthritis in NHPI was significantly lower than Whites and Asians. Among NHPI males, this was 13 years younger than Whites and 14 years younger than Asians. NHPI females were diagnosed on average 4 years younger than Whites and 6 years younger than Asians.
This study is the first to demonstrate racial disparities in the prevalence of arthritis among NHPI, Whites, and Asians. NHPI adult males have a significantly higher prevalence of arthritis than White and Asian adult males in all age groups, and arthritis in this population peaks twenty years earlier than in other groups. In order to prevent or eliminate health disparities, they must first be identified. Future research into potential causal relationships and specific types of arthritis through longitudinal studies are warranted.
None of the authors identify a conflict of interest.