Abstract
Health literacy is understudied in Asian Americans/Pacific Islanders (AA/PI). We used a population-based sample in Hawai'i to consider if low health literacy is associated with poor health outcomes in Japanese, Filipino, Native Hawaiians, and other AA/PI groups compared with Whites. In data weighted and adjusted for population undercounts and complex survey design, low health literacy varied significantly by group, from 23.9% among Filipinos, 20.6% in Other AA/PI, 16.0% in Japanese, 15.9% in Native Hawaiians, and 13.2% in Whites (χ2 (4) = 52.22; p < .001). In multivariate models, low health literacy was significantly associated with (a) poor self-reported health in Japanese, Filipinos, Other AA/PI, and Whites; (b) diabetes in Hawaiians and Japanese; and (c) depression for Hawaiians. Low health literacy did not significantly predict overweight/obesity in any ethnic grouping in multivariate models. The design and relevance of health literacy interventions, as well as the pathways that link health literacy to health status, may vary by race/ethnicity, culture, and health outcomes.
Adult health literacy is defined in Healthy People 2010 as the “the degree to which individuals have the capacity to obtain, process, and understand basic health information and services for appropriate health decisions” (U.S. Department of Health and Human Services, 2010). Health literacy is an important predictor of health status and behavior in the general population that has not been well studied in Asian Americans or Pacific Islander (AA/PI) populations.
Asian Americans and Pacific Islanders
The AA/PI group is one of the fastest growing ethnic groups in the United States (US), yet there is limited information on many health disparities and health correlates in these populations (Ghosh, 2003; Ghosh, 2010). A key barrier to population-level research on the AA/PI population generally is its small size relative to the U.S. population, presenting challenges for surveillance, stratification, and analysis (Ghosh, 2010). Another issue is the extreme diversity of the 50+ populations aggregated under the AA/PI label, which have distinct cultures, languages, histories, immigration patterns, and health outcomes. Thus, combining AA/PI groups masks significant health disparities and obscures important variations in health correlates (Ghosh, 2010; Panapasa, Mau, Williams, & McNally, 2010; Pourat, Kakawa-Singer, Breen, & Sripipatana, 2010, Tanjasiri, Wallace, & Shibata, 1995). For example, a detailed analysis of AA/PI subgroups in Hawai'i found a 12-year gap in expected longevity among AA/PI groups, with life expectancy at birth longest for Chinese residents and shortest for Native Hawaiian and other Pacific Islander residents (Park, Braun, Horiuchi, Tottori, & Onaka, 2009). Important differences also are seen in the strength and importance of health determinants (e.g., insurance, acculturation, usual source of care) and health outcomes by AA/PI subgroups, indicating that the most meaningful and appropriate interventions may vary across diverse AA/PI groups (Pourat et al., 2010).
Health Literacy
Although adult health literacy is becoming increasingly recognized as a key predictor of health behaviors and outcomes, research in this area has not focused on AA/PI populations. The few published studies directly concerning Asian groups and health literacy have examined specific Asian groups living in Asian countries (Lee, Kang, Lee, & Hyun, 2009; Tang, Pang, Chan, Yeung, & Yeung, 2008; Tokuda, Doba, Butler, & Paasche-Orlow, 2009). To our knowledge, no published studies report health literacy prevalence or outcomes among Pacific Islander (PI) groups. The few U.S.-based studies examining literacy that report data for AA/PI respondents (rather than including them in the “Other” racial/ethnic group) do not report data for distinct AA/PI subgroups. Rather, AA/PI data are aggregated. For example, the 2003 National Assessment of Adult Literacy (NAAL), which currently provides the strongest population-level evidence regarding low health literacy, reports health literacy prevalence data only from the combined AA/PI group (Kutner, Greenberg, Jin, Paulsen, 2006). Reports based on the NAAL data have not included detailed health outcomes for even the combined AA/PI group because of sample size limitations. In sum, the literature on prevalence and health outcomes of low health literacy in AA/PI is extremely limited.
Yet studies reporting data on predominantly White, African American, and Latino populations have typically found low health literacy to be linked to poorer self-reported health status, as well as many other health outcomes, such as diabetes and poorer mental health (Berkman et al., 2011; DeWalt, Berkman, Sheridan, Lohr, Pingnone, 2004; Kutner, Greenberg, Jin, & Paulsen, 2006). In a number of studies, health literacy also has been shown to be a more important determinant of health than race/ethnicity, income, and/or education, indicating that health literacy is a key factor for explaining health disparities (Bennett, Chen, Soroui, and White, 2009; Howard, Sentell, Gazmararian, 2006; Lindau, Tomori, Lyons, Langseth, Bennett, Garcia, 2002; Paasche-Orlow & Wolf, 2007; Sentell & Halpin, 2006).
It is important to clarify whether the relationship between health literacy that has been found in other samples holds for AA/PI for at least three reasons: First, AA/PI groups may have distinct health patterns and disparities when compared with Whites and with other racial/ethnic groups, as well as when compared across AA/PI groups (Ghosh, 2010; Panapasa et al., 2010; Park et al., 2009; Pourat et al., 2010, Tanjasiri et al., 1995). It would be useful to know if health literacy operates similarly as a health predictor for both Whites and AA/PI and if it works differently for distinct AA/PI groups.
Second, health literacy can be increased through intervention, which could mediate health disparities (Schillinger, Barton, Karter, Wang, & Adler, 2006). Yet distinct cultural communities may interact with health information, health concerns, and health access in different ways, which could lead to variation in the relationship of health literacy to health across racial/ethnic subgroups. This would have implications for health literacy interventions targeting AA/PI groups.
Third, racial/ethnic variation may help us to understand the theoretical pathways by which health literacy actually impacts health. Health literacy is theorized to impact health through a number of pathways, including instrumental ones, such as the ability to comprehend and access health care directly due to reading skills, and more diffuse pathways, such as self-efficacy or trust in providers (Osborn, Cavanaugh, Wallston, & Rothman, 2010; Osborn, Paasche-Orlow, Bailey, & Wolf, 2011; Saha, 2006). However, empirical findings about these pathways have been mixed (Saha, 2006). Understanding and specifying these theoretical pathways are critical to designing effective interventions for all groups. Noting variation by cultural history and ethnic identity may provide insight into the operation of the pathways by which lower health literacy would lead to poor health.
Existing evidence is extremely limited about whether the relationship of low health literacy to health would be similar in AA/PI compared with Whites. A recent study of Japanese adults in Japan found that poor self-reported health literacy was associated with poor physical and mental health status (Tokuda et al., 2009). This suggests that health literacy operates similarly to population-based U.S. samples in at least one Asian group. On the other hand, a study that examined African Americans and Latinos found no association between health literacy and physical and mental health status (Guerra & Shea, 2007). This led the authors to question the generalizability of the perceived link between health literacy and poor health status in minority communities. This may be particularly true for at least some AA/PI groups because of distinct cultural histories, contexts, and identities.
Purpose
As noted above, health determinants and outcomes vary dramatically across AA/PI groups (Ghosh, 2010; Panapasa et al., 2010; Park et al., 2009; Pourat et al., 2010; Tanjasiri et al., 1995). Thus, we might expect to see considerable variation in low health literacy prevalence and the links between low health literacy and health outcomes across AA/PI groups. We used a weighted and adjusted population-based sample to examine associations between low health literacy and poor health outcomes among AA/PI groups in Hawai'i—Japanese, Filipino, Hawaiian, and “other AA/PI”—compared with Caucasians. This furthers research on health literacy and health associations across AA/PI groups. Hawai'i, where two-thirds of residents are AA/PI, is an excellent location to examine variation across AA/PI groups because population-based studies in Hawai'i have larger AA/PI sample sizes compared with most other U.S. population-based studies.
Methods
Sample
The Hawai'i Health Survey (HHS), conducted annually by the Hawai'i State Department of Health, Office of Health Status Monitoring (OHSM), is a population-based phone survey modeled after the National Health Interview Survey. The HHS is stratified by island, and a random cluster sample is taken of island households and members (OHSM, 2011). In the HHS, respondents (18 years and older) report health and demographic information both on themselves and on other household members. For this study, only the results from the actual adult respondents were used, as the health literacy question was not asked by proxy. Because the HHS is administered in English, only adults who could speak English well enough to answer questions on the phone are included in the HHS sample. About 4% of households sampled were excluded for not having a respondent meeting these English-proficiency requirements. More information about this survey can be found at http://hawaii.gov/health/statistics/hhs/index.html.
The 2008 survey had a Council of American Survey Research Organizations (CASRO) completion rate of 40.1%, yielding data from 5,928 respondents. We reported weighted and adjusted data from the 5,399 adults who self-reported either an AA/PI subgroup or White (for comparison): 2,277 respondents were White and 3,122 were AA/PI of any group; specifically 1,201 were Japanese, 810 were Native Hawaiian, 731 were Filipino, and 380 were Other AA/PI. Because the HHIS is a population-based study, data from the HHS respondents were weighted based on population estimates to correct for under-sampling of certain populations and to be representative of the approximately 970,567-person adult population of Hawai'i.
Study Variables
Health Literacy
Although health literacy has traditionally been measured with an in-person test, this is not practical for most population-based surveys. Fortunately, a single self-reported health literacy item, “How confident are you filling out medical forms by yourself?” has been validated against the most commonly given in-person tests and shown to perform well in identifying low health literacy (area under the receiver operating characteristic curve [AUROC] ≥ 80) against both the Rapid Estimate of Adult Literacy in Medicine (REALM) and the Test of Functional Health Literacy in Adults (TOFHLA; Chew, Bradley, & Boyko, 2004; Chew et al., 2008; Wallace, Rogers, Roskos, Holiday, & Weiss, 2006). This item also has been used in other studies (Tokuda et al., 2009; Sarkar et al., 2010). Following the literature, we coded individuals as having low health literacy if they responded “not at all,” “a little bit,” or “somewhat” to this question (Chew et al., 2008). These were compared with those who reported being “quite a bit” or “extremely” confident filling out medical forms by themselves.
Race/Ethnicity
Ethnicity was self-reported. We included adults reporting White or any AA/PI group in the analysis. We also looked specifically within four AA/PI groups with sufficient sample size for detailed analysis: Native Hawaiian, Filipino, Japanese, and Other AA/PI. The Other AA/PI group was 80% Chinese, but the Chinese lacked sufficient sample sizes to be analyzed independently. The HHS asks a number of questions concerning race/ethnicity, as this can be a complex issue generally and is particularly in the population of Hawai'i, which is racially, ethnically, and culturally mixed. In this analysis, the first racial/ethnic group that subjects reported in response to the question: “What race do you consider yourself to be?” is used to describe race/ethnicity. The advantages and limitations of this approach have been considered in other studies (e.g., Park et al., 2009), with an overall finding that this is a useful metric for self-reported race in a multiethnic society.
Health Outcomes
All health outcomes were self-reported. Specifically, we considered poor health, depression, diabetes, and overweight/obesity. A person was coded as in poor health if they answered “fair” or “poor” to the question “Would you say your health in general is excellent, very good, good, fair, poor, or don't know?” Individuals who responded don't know (<1% across all groups) were excluded from the self-reported health analyses. Depression was coded “yes” if the person answered “all the time” or “most of the time” to the question: “In the past four weeks, have you felt down-hearted and depressed all the time, most of the time, some of the time, a little of the time, or none of the time?” A person was coded as having diabetes if he/she reported having been told by a physician or medical professional that he/she had this condition. A person was coded as overweight or obese based on standard National Heart, Lung, and Blood Institute (NHLBI) BMI thresholds (NHLBI, 2008) using self-reported weight and height without shoes.
Control Variables
Because health literacy and health status are associated with education, age, gender, living in a rural area, insurance status, and income (Nielsen-Bohlman, Panzer, & Kindig, 2004; Schillinger et al., 2006), these were used as control variables in multivariate models. Education was measured at three levels (less than HS, high school, and greater than high school) based on self-reported educational attainment. Age was self-reported as a continuous variable (18–105). Health insurance status was coded by OHSM from responses to seventeen questions on insurance status, including respondent policy information and self-reported coverage. This detail was used to create a dichotomous variable indicating insured (1) or not (0). Gender was ascertained as male or female. Being at or near poverty was estimated according to U.S. Department of Health and Human Services (USDHHS) poverty guidelines for Hawai'i for 2008 (USDHHS, 2008) from self-reported pre-tax income. Individuals who were at or under the threshold of 199% of poverty were coded as 1, and others were coded as 0. Marital status was recorded as married 1 = yes or 0 = no. Location was coded as 1 = living in Oahu compared with living on the other Hawaiian Islands = 0 as the other islands are more rural and have greater health care access barriers.
Statistical Analyses
The relationship of low health literacy to health outcomes was compared across the four AA/PI groups and Whites using chi-square analyses. Multivariate logistic models were run predicting each health outcome for each subgroup individually, adjusting for the control variables. All data were analyzed in Stata 11 (Stata Corp., 2009) using svy commands, which account for the complex survey design and provide weighted variance estimates. Data were also weighted to represent the adult population of Hawai'i (excluding households without telephones, group quarters, homeless, and the island of Ni'ihau). All percentages reported in this study represent adjusted and weighted estimates.
Results
Low health literacy, health outcomes, and demographics for everyone studied and by ethnic categories are shown in Table 1. Considering low health literacy, Filipinos had the highest rates of low health literacy (with 23.9% reporting low confidence filling out medical forms by themselves), followed by 20.6% of Other AA/PI, 16.0% of Japanese, 15.9% of Native Hawaiians, and 13.2% of Whites (χ2 (4) = 52.22, p < .001). All health outcomes except overweight varied across racial/ethnic groups. Native Hawaiians had the poorest self-reported health, followed by Filipinos and other AA/PIs. Whites had the lowest percentage of diabetes (5.4%), but the highest percentage of self-reported depression (4.6%). Native Hawaiians had the highest percentage of obesity (43.3%) compared to the other racial/ethnic groups. All demographic variables also varied across racial/ethnic groups except marital status.
Table 1. Demographics by ethnicity for AA/PI and White adults from the 2008 Hawai'i Health Survey
Table 2 shows the comparison of each health status variable with regard to low and adequate health literacy for each participant and for each racial/ethnic group specifically. Listings shown in bold designate statistical significance in the comparisons between low and adequate health literacy for the particular health status variable within each group. Of note are the extremely high percentages of those with low health literacy who reported poor health: 28.3% among Native Hawaiians, 24.5% among Filipinos, 28.5% among Japanese, 25.0% among other AA/PI, and 23.8% among Whites. Also of note, in unadjusted models, low health literacy is associated with: (a) poor self-reported health in all groups except Native Hawaiians; (b) diabetes in Hawaiians, Japanese, and Whites; and (c) depression in all groups except Japanese and other AA/PI. Low health literacy was significantly associated with overweight in Hawaiians and Japanese, but was not associated with obesity in any racial/ethnic group.
Table 2. Health outcomes by health literacy and self-reported race/ethnicity for AA/PI and White adults in the 2008 Hawai'i Health Survey
Multivariate Models
Poor Health
Multivariate models for poor self-reported health are shown in Table 3. Low health literacy was significantly associated with poor health status for all groups except Native Hawaiians. The effect size (odds ratios) were similar across the significant models for all groups, ranging from 2.31 (95% CI: 1.08–4.94) among Filipinos to 2.72 (95% CI: 1.23–6.02) among other AA/PIs. Also significant in at least one model were education, age, marital status, poverty, and insurance; however, different variables were significant for different ethnic groupings.
Table 3. Odds ratios and 95% confidence intervals for multivariate models predicting self-reported poor health for AA/PI and White adults in the 2008 Hawai'i Health Survey
Diabetes
Multivariate models for diabetes are shown in Table 4. Low health literacy was significantly associated with diabetes for Native Hawaiians and Japanese. Point estimates of odds ratios were 3.03 (95% CI: 1.34–6.83) for Native Hawaiians and 1.78 (95% CI: 1.00–3.16) in Japanese, but confidence intervals overlapped considerably. Higher age was significantly associated with diabetes in all groups as well.
Table 4. Odds ratios and 95% confidence intervals for multivariate models predicting diabetes for AA/PI and White adults in the 2008 Hawai'i Health Survey
Depression
Multivariate models for depression are shown in Table 5. Low health literacy was significantly associated with depression for Native Hawaiians. The odds ratio was extremely high for the low literacy variable in Native Hawaiians: 4.51 (95% CI: 1.72–11.81), but this may be because of the small sample sizes from those reporting depression and thus the statistical instability (indicated by the large CI) of the point estimate. Those who were married were less likely to report depression for Japanese and Whites, while poverty was associated with greater depression among Other AA/PI and Whites.
Table 5. Odds ratios and 95% confidence intervals for multivariate models predicting depression for AA/PI and White adults in the 2008 Hawai'i Health Survey
Obesity/Overweight
Low health literacy was not significantly associated with obesity/overweight in any group in adjusted models (not shown in tables).
Discussion
This article has five important findings. First, it confirms that low health literacy is a significant predictor of key health status variables (self-reported poor health, diabetes, and depression) in Hawai'i's adult AA/PI population, particularly within several AA/PI subgroups using a population-based sample. This is important baseline information for further studies on this topic and should be relevant to policy and clinical practice.
Our second key finding notes significant variation across AA/PI groups in the association of low health literacy across health status measures. Of particular note, lower health literacy did not predict poor self-reported health for Native Hawaiians, diabetes for Filipinos, or depression among Japanese or Filipinos. Thus, the relevance of low health literacy may vary across distinct AA/PI groups and health outcomes. This has implications for the effectiveness and meaningfulness of health literacy interventions in these particular groups, and these racial/ethnic differences may be an important area to consider in comparisons of the effectiveness of health literacy intervention on health outcomes (Berkman et al., 2011).
Third, we found that overweight and obesity, important determinants of health risk, were not associated with low literacy in any group in multivariate models. This may be because of inaccuracies in self-reported weight and height, which were used to determine BMI (McAdams, Van Dam, & Hu, 2007). However, further examination is needed regarding why self-reported health, depression, and diabetes were associated with low health literacy, whereas obesity (a risk factor for diabetes) was not.
Other curious patterns emerged. For example, although some groups had similar factors associated with health outcomes, such as marital status being protective of depression, this was not true across all racial/ethnic groups. In self-reported health, most established factors (poverty, insurance status, marital status) were significant only in Whites in the multivariate models. Although lack of significance in AA/PI groups might be because of small sample sizes, a growing literature identifies racial/ethnic differences in health status predictors in AA/PI (Ghosh, 2010; Panapasa et al., 2010; Pourat et al., 2010, Tanjasiri et al., 1995) and other racial/ethnic groups. Exploring reasons behind these differential patterns may help us understand the pathways by which low literacy might compromise health status, as well as the possible variation in these pathways by race/ethnicity. Understanding these variations in these predictors can advance our understanding the sources and predictors of health disparities, and provide critical insights needed for developing effective treatments and interventions.
Fourth, as part of this study, we report some baseline rates of poor self-reported health literacy for adult AA/PI populations. Previously, the only study to report population-level information on AA/PI groups was the 2003 NAAL, which reported aggregated AA/PI data. NAAL findings suggested that 13% of AA/PI had below-basic health literacy (Kutner et al., 2006). We found a higher prevalence—18%— for our aggregated AA/PI population (results not shown), as well as significant variation across AA/PI groups, from 24% of Filipinos to 16% of Japanese and 15.9% of Native Hawaiians. Thus, this study provides yet another example in of the importance of disaggregating AA/PI subgroups to understand health determinants and outcomes.
Finally, we highlight the large numbers of those with low health literacy who struggle with various types of poor health status. Regardless of whether the relationship between low health literacy and health outcome variables in this study held in multivariate models, evidence that 8.7% of Native Hawaiians with low health literacy self-report feeling depressed or 28.5% of Japanese with low health literacy have poor self-reported health provides important information about vulnerable population groups and the communication barriers they may face in receiving needed care.
Health literacy is a key social determinant of health. Like other social determinants of health, the ability to leverage this capacity and/or the disadvantages that accrue from poor health literacy may vary by racial/ethnic groups and may be mediated by the social and structural context in which these groups live (Saha, 2006; Schillinger et al., 2006). Considering the patterns of these predictors should be useful in designing interventions to target low health literacy. For instance, health literacy is not correlated with self-reported health status among Hawaiians, but is strongly associated with depression and diabetes in this group. This may be related to cultural differences in the way in which Hawaiians assess and describe self-reported health status and/or self-reported health literacy, or may be a unique finding in this particular sample. Future research including larger sample sizes and/or qualitative methods could help confirm this finding and/or identify reasons for this apparent paradox. Our findings also suggest that health literacy interventions among Hawaiians may be most useful for depression and diabetes specifically, and qualitative research could further solicit information to directly inform the development of these interventions.
Limitations
This study has a number of strengths, including a population-level sample, large sample sizes of AA/PI groups, and a validated measure of self-reported health literacy. However, it does have some limitations. For one, the 2008 Hawai'i Health Survey did not include a measure of limited English proficiency (LEP), which could explain some of the low health literacy variation within and across racial/ethnic groups. However, the HHS restricts respondents to those who are able to speak English, leaving individuals who are non-English speakers out of this sample for all racial/ethnic groups, making this less of a concern. Similarly, the 2008 HHS did not include measures of U.S. nativity, immigration, and/or acculturation, which are likely to be associated with low health literacy, and could interact with the variables included here. We hope this study provides evidence to demonstrate the importance of low health literacy in AA/PI populations and inspires more research about these overlapping correlates.
The health literacy measure used here, although validated in previous studies with non-AA/PI samples (Chew et al., 2004; Chew et al., 2008; Wallace et al., 2006), is a self-reported measure. It would be ideal to assess health literacy face-to-face using other well-validated measures, although this is expensive and time-consuming on a population level. The variation seen here by AA/PI subgroup and health outcomes may help provide evidence to support the importance of such time- and cost-intensive efforts.
Also, it is important to note that this is a cross-sectional study. Causation cannot be determined, nor can the presence of another factor for which low health literacy is a proxy (such as self-efficacy), nor can we determine the direction of the relationship between literacy and health. The relationship of low health literacy to health could signify reverse causation (poor health = poor test performance), and/or “inadequate literacy may itself reflect the trajectory of chronic disease” (Schillinger et al., 2006, p. 251). This is a common concern across cross-sectional studies of health literacy. Also, it would be ideal to have multiple years of data from the HHS specifically to have sufficient sample size to dampen the variability by year for point estimates and to confirm that these relationships hold across time. With a larger sample size, additional ethnic groups in Hawai'i, including those of mixed race, could also be considered in detail. Other health status measures, such as cancer, known to vary with health literacy, would be interesting to follow across longer time periods, or with larger cohorts.
A final issue is that this self-reported measure was validated in a non-AA/PI sample and may thus vary in its validity and reliability for AA/PI populations generally and among certain AA/PIs specifically. AA/PI are known to have distinct patterns of responding to heath measures and survey questions (Kandula, Lauderdale, & Baker, 2007), and may be less likely to self-report health communication challenges. If that is the case, our findings may be an underestimation of the burden of low health literacy in the AA/PI adult population.
Even considering these possible measurement concerns, we believe that the self-reported health literacy measure used in the HHS also has strong face validity as a measure of health communication challenges because it is a direct report of comprehension difficulties with written communication. Even if one considers this a measure of “health communication challenges” rather than “health literacy,” our key findings concerning variation in prevalence and associations by AA/PI groups remain relevant. Clearly, these data signify respondents' own reports of their struggles with health information, and these struggles are linked to key health outcomes.
Conclusions
Overall, this study should alert policy makers, health planners, clinicians, and researchers that low health literacy is associated with poorer health outcomes in adult AA/PI populations, but that these patterns vary by AA/PI group. Of particular note, lower health literacy did not predict poor self-reported health for Native Hawaiians (who had the highest level of self-reported poor health of all groups), diabetes for Filipinos, or depression among Japanese or Filipinos. These findings are useful in the design of low health literacy interventions for distinct AA/PI population groups. These findings also suggest that the pathways that link low health literacy to health status may vary by race/ethnicity, culture, and health outcome. Further research should explore these important topics in more detail.
| Native Hawaiian | Filipino | Japanese | Other AA/PI | White | Total | |
|---|---|---|---|---|---|---|
| Unweighted n | 810 | 731 | 1,201 | 380 | 2,777 | 5,399 |
| Weighted % of Total | 15.7 | 14.2 | 24.5 | 11.6 | 34.0 | 100 |
| Weighted % of AA/PI | 23.8 | 21.5 | 37.2 | 17.5 | 0 | |
| % | % | % | % | % | % | |
| Low Health Literacy * | 15.9 | 23.9 | 16.0 | 20.6 | 13.2 | 16.7 |
| Health Outcomes | ||||||
| Poor Health* | 19.4 | 14.9 | 13.9 | 14.3 | 10.6 | 13.9 |
| Diabetes* | 10.5 | 10.7 | 11.1 | 10.7 | 5.4 | 8.9 |
| Depression* | 2.8 | 3.0 | 1.3 | 2.7 | 4.6 | 3.1 |
| Overweight | 28.7 | 34.9 | 35.0 | 31.2 | 33.5 | 33.0 |
| Obese* | 43.3 | 15.7 | 11.9 | 16.6 | 17.0 | 19.7 |
| Demographics | ||||||
| Education* | ||||||
| Less than high school | 3.1 | 7.1 | 2.4 | 2.8 | 2.0 | 3.1 |
| High school | 45.3 | 35.8 | 25.9 | 30.8 | 19.8 | 28.8 |
| More than high school | 51.7 | 57.1 | 71.7 | 66.4 | 78.2 | 68.1 |
| Age Group* | ||||||
| Youngest (18–24) | 20.1 | 21.2 | 4.6 | 13.0 | 3.8 | 10.1 |
| Middle (25–64) | 70.1 | 67.4 | 66.6 | 73.4 | 74.0 | 70.6 |
| Older (65–84) | 8.8 | 10.8 | 22.5 | 9.6 | 18.7 | 15.9 |
| Elderly (85 +) | 1.0 | <1.0 | 6.2 | 4.0 | 3.5 | 3.4 |
| Female* | 54.7 | 59.2 | 52.6 | 50.4 | 46.1 | 51.4 |
| Below or Near Poverty* | 30.5 | 35.6 | 16.7 | 23.0 | 23.0 | 24.4 |
| Live in Oahu* | 66.6 | 72.6 | 80.6 | 88.9 | 58.5 | 70.7 |
| Insured* | 92.8 | 91.9 | 95.9 | 96.5 | 91.9 | 93.5 |
| Married | 49.4 | 54.3 | 54.3 | 55.8 | 57.5 | 54.7 |
| Note. All percentages are based on the weighted HHS estimates. *Variable significant at p < .05 in χ2 comparing across all racial/ethnic groups. | ||||||
| Race/Ethnicity | Health literacy | % Poor health | % Diabetes | % Depressed | % Overweight | % Obese |
|---|---|---|---|---|---|---|
| Native Hawaiian | Low HL | 28.3* | 18.3 | 8.7 | 18.2 | 46.1 |
| Adequate HL | 17.6* | 8.8 | 1.7 | 31.1 | 42.2 | |
| Filipino | Low HL | 24.5 | 15.5* | 5.9 | 36.8 | 12.2 |
| Adequate HL | 11.9 | 9.3* | 2.1 | 34.8 | 17.0 | |
| Japanese | Low HL | 28.5 | 16.9 | 3.2* | 24.6 | 12.1 |
| Adequate HL | 10.9 | 10.0 | 1.0* | 37.2 | 11.8 | |
| Other AA/PI | Low HL | 25.0 | 11.0 | 2.8 | 32.8 | 25.6 |
| Adequate HL | 11.8 | 9.4 | 2.8 | 30.5 | 14.5 | |
| White | Low HL | 23.8 | 9.0 | 10.3 | 36.8 | 22.1 |
| Adequate HL | 8.7 | 4.8 | 3.8 | 32.8 | 16.3 | |
| Total | Low HL | 25.9 | 13.9 | 6.4 | 30.6 | 21.9 |
| Adequate HL | 11.4 | 7.8 | 2.4 | 33.6 | 19.2 | |
| Note. All percentages are based on the weighted HHS estimates. Bold indicates χ2 comparison between low and adequate health literacy is significant at p < .05. HL = health literacy. *χ2 comparison between low and adequate health literacy was marginally significant, .05 ≤ p ≤ .10. | ||||||
| Native Hawaiian only model | Filipino only model | Japanese only model | Other AA/PI only model | Whites only model | |
|---|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Low Health Literacy | 1.54 (0.80–2.97) | 2.31 (1.08–4.94) | 2.62 (1.55–4.43) | 2.72 (1.23–6.02) | 2.49 (1.51–4.11) |
| Female | 1.47 (0.84–2.56) | 0.93 (0.44–1.93) | 0.95 (0.61–1.50) | 1.76 (0.84–3.67) | 1.03 (0.70–1.51) |
| Education | |||||
| Less than high school | 2.93 (1.07–8.04) | 1.55 (0.42–5.75) | 0.97 (0.32–2.98) | 0.84 (0.22–3.19) | 1.66 (0.56–4.97) |
| High school | 1.73 (0.92–3.22) | 0.92 (0.45–1.89) | 1.24 (0.73–2.10) | 0.90 (0.40–2.06) | 1.67 (1.06–2.61) |
| More than high school | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Age | 1.02 (1.00–1.03) | 1.02 (0.99–1.04) | 1.03 (1.01–1.04) | 1.02 (1.00–1.03) | 1.03 (1.01–1.05) |
| Married | 0.70 (0.40–1.22) | 0.85 (0.46–1.57) | 0.56 (0.35–0.89) | 0.91 (0.42–1.97) | 0.56 (0.38–0.83) |
| Live in Oahu | 1.13 (0.70–1.81) | 0.99 (0.56–1.74) | 1.02 (0.68–1.52) | 0.77 (0.40–1.49) | 0.94 (0.65–1.35) |
| Poor/Near Poor | 1.73 (0.97–3.08) | 1.28 (0.67–2.47) | 1.59 (0.93–2.70) | 1.07 (0.43–6.61) | 2.08 (1.34–3.22) |
| Insured | 1.30 (0.32–5.25) | 3.67 (0.89–15.09) | 1.32 (0.29–6.00) | 1.16 (0.31–4.32) | 2.54 (1.16–5.52) |
| Note. All percentages are based on the weighted HHS estimates. Bolded indicates comparison is significant at p < .05. | |||||
| Native Hawaiian only model | Filipino only model | Japanese only model | Other AA/PI only model | Whites only model | |
|---|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Low Health Literacy | 3.03 (1.34–6.83) | 1.80 (0.88–3.67) | 1.78 (1.00–3.16) | 1.29 (0.52–3.22) | 1.40 (0.71–2.76) |
| Female | 1.26 (0.69–2.30) | 0.97 (0.48–1.92) | 0.70 (0.43–1.11) | 0.85 (0.39–1.88) | 0.90 (0.57–1.43) |
| Education | |||||
| Less than high school | 1.95 (0.76–4.97) | 0.66 (0.23–1.86) | 0.30 (0.94–0.97) | 0.22 (0.02–2.18) | 1.05 (0.32–3.48) |
| High school | 1.04 (0.55–1.97) | 1.85 (0.97–3.51) | 1.36 (0.78–2.38) | 0.58 (0.27–1.26) | 2.10 (1.22–3.62) |
| More than high school | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Age | 1.04 (1.02–1.05) | 1.08 (1.06–1.10) | 1.04 (1.03–1.06) | 1.05 (1.03–1.07) | 1.05 (1.04–1.06) |
| Married | 1.65 (0.89–3.04) | 1.01 (0.52–1.97) | 0.75 (0.47–1.22) | 2.61 (1.08–6.30) | 0.98 (0.61–1.58) |
| Live in Oahu | 0.48 (0.28–0.81) | 0.79 (0.47–1.35) | 0.78 (0.51–1.17) | 1.15 (0.54–2.45) | 0.93 (0.61–1.42) |
| Poor/Near Poor | 0.84 (0.46–1.52) | 0.69 (0.36–1.32) | 0.53 (0.28–1.00) | 1.03 (0.30–3.57) | 1.61 (0.92–2.80) |
| Insured | 2.20 (0.70–6.88) | 0.94 (0.19–4.70) | 1.17 (0.19–7.24) | 1.10 (0.19–6.24) | 2.15 (0.62–7.41) |
| Note. All percentages are based on the weighted HHS estimates. Bold indicates comparisons are significant at p < .05. | |||||
| Native Hawaiian only model | Filipino only model | Japanese only model | Other AA/PI only model | Whites only model | |
|---|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Low Health Literacy | 4.51 (1.72–11.81) | 2.58 (0.72–9.22) | 2.00 (0.49–8.24) | 0.62 (0.11–3.47) | 2.24 (0.92–5.46) |
| Female | 0.54 (0.18–1.66) | 0.78 (0.24–2.54) | 0.61 (0.17–2.20) | 5.53 (0.95–32.18) | 0.69 (0.36–1.33) |
| Education | |||||
| Less than high school | 3.00 (0.73–12.38) | 1.18 (0.14–10.15) | 4.97 (0.38–64.22) | 2.50 (0.28–22.17) | 2.09 (0.26–17.03) |
| High school | 0.55 (0.14–2.14) | 2.84 (0.83–9.68) | 3.15 (0.67–14.76) | 1.38 (0.31–6.12) | 0.60 (0.26–1.36) |
| More than high school | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Age | 1.01 (0.98–1.04) | 1.04 (1.02–1.07) | 1.00 (0.97–1.04) | 1.00 (0.98–1.03) | 0.99 (0.97–1.01) |
| Married | 0.89 (0.30–2.62) | 0.90 (0.25–3.29) | 0.49 (0.10–0.23) | 0.27 (0.06–1.23) | 0.41 (0.22–0.77) |
| Live in Oahu | 0.59 (0.20–1.73) | 0.92 (0.33–2.51) | 0.35 (0.08–1.42) | 0.73 (0.16–3.40) | 0.85 (0.48–1.50) |
| Poor/Near Poor | 2.11 (0.71–6.23) | 0.99 (0.26–3.81) | 1.92 (0.55–6.63) | 4.18 (1.06–16.47) | 2.82 (1.42–5.58) |
| Insured | —* | 0.32 (0.54–1.93) | —* | 2.35 (0.12–45.80) | 3.81 (1.11–13.04) |
| Note. All percentages are based on the weighted HHS estimates. Bold indicates significant at p < .05. *Dropped because predicts failure perfectly. | |||||
Notes
Note. All percentages are based on the weighted HHS estimates.
*Variable significant at p < .05 in χ2 comparing across all racial/ethnic groups.
Note. All percentages are based on the weighted HHS estimates. Bold indicates χ2 comparison between low and adequate health literacy is significant at p < .05. HL = health literacy.
*χ2 comparison between low and adequate health literacy was marginally significant, .05 ≤ p ≤ .10.
Note. All percentages are based on the weighted HHS estimates. Bolded indicates comparison is significant at p < .05.
Note. All percentages are based on the weighted HHS estimates. Bold indicates comparisons are significant at p < .05.
Note. All percentages are based on the weighted HHS estimates. Bold indicates significant at p < .05.
*Dropped because predicts failure perfectly.
References
- Bennett , I. M. , Chen , J. , Soroui , J. S. , & White , S. ( 2009 ). The contribution of health literacy to disparities in self-rated health status and preventive health behaviors in older adults . Annals of Family Medicine , 7 ( 3 ), 204 – 211 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Berkman , N. D. , Sheridan , S. L. , Donahue , K. E. , Halpern , D. J. , Viera , A. , et al. . ( 2011 ). Health literacy interventions and outcomes: An updated systematic review. (Evidence Report/Technology Assessment No. 199. Prepared by RTI International–University of North Carolina Evidence–based Practice Center under contract No. 290–2007–10056–I. AHRQ Publication Number 11–E006.) Rockville, MD, Agency for Healthcare Research and Quality. Retrieved from http: //www.ahrq.gov/downloads/pub/evidence/pdf/literacy/literacyup.pdf [Google Scholar]
- Chew , L. D. , Bradley , K. A. , & Boyko , E. J. ( 2004 ). Brief questions to identify patients with inadequate health literacy . Family Medicine , 36 ( 8 ), 588 – 594 . [PubMed], [Web of Science ®], [Google Scholar]
- Chew , L. D. , Griffin , J. M. , Partin , M. R. , Noorbaloochi , S. , Grill , J. P. , et al.. (2008). Validation of screening questions for limited health literacy in a large VA outpatient population. Journal of General Internal Medicine , 23(5), 561–566. [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- DeWalt , D. , Berkman , N. D. , Sheridan , S. , Lohr , K. N. , & Pignone , M. P. ( 2004 ). Literacy and health outcomes: A systematic review of the literature . Journal of General Internal Medicine , 19 ( 12 ), 1228 – 1239 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Ghosh , C. ( 2003 ). Healthy People 2010 and Asian Americans/Pacific Islanders: Defining a baseline of information . American Journal of Public Health , 93 ( 12 ), 2093 – 2098 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Ghosh , C. ( 2010 ). A national health agenda for Asian Americans and Pacific Islanders . Journal of the American Medical Association , 304 ( 2 ), 1381 – 1382 . [Google Scholar]
- Guerra , C. E. , & Shea , J. A. ( 2007 ). Health literacy and perceived health status in Latinos and African Americans . Ethnicity & Disease , 17 ( 2 ), 305 – 312 . [PubMed], [Web of Science ®], [Google Scholar]
- Howard , D. H. , Sentell , T. , & Gazmararian , J. A. ( 2006 ). Impact of health literacy on socioeconomic and racial difference in health in an elderly population . Journal General Internal Medicine , 21 ( 8 ), 857 – 861 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Kandula , N. R. , Lauderdale , D. S. , & Baker , D. W. ( 2007 ). Differences in self-reported health among asians, latinos, and non-hispanic whites: The role of language and nativity . Annals of Epidemiology , 17 ( 3 ), 191 – 198 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Kutner , M. , Greenberg , E. , Jin , Y. , Paulsen , C. ( 2006 ). The health literacy of America's adults: Results from the 2003 National Assessment of Adult Literacy (National Center for Education, NCES 2006-483). Washington DC: U.S. Department of Education. [Google Scholar]
- Lee , T. W. , Kang , S. J. , Lee , H. J. , & Hyun , S. I. ( 2009 ). Testing health literacy skills in older Korean adults . Patient Education and Counseling , 75 ( 3 ), 302 – 307 . [Google Scholar]
- Lindau , S. T. , Tomori , C. , Lyons , T. , Langseth , L. , Bennett , C. L. , & Garcia , P. ( 2002 ). The association of health literacy with cervical cancer prevention knowledge and health behaviors in a multiethnic cohort of women . American Journal Obstetrics and Gynecology , 186 ( 5 ), 938 – 943 . [Crossref], [Google Scholar]
- McAdams , M. A. , Van Dam , R. M. , & Hu , F. B. ( 2007 ). Comparison of self-reported and measured BMI as correlates of disease markers in US adults . Obesity , 15 ( 1 ), 188 – 196 . [Crossref], [Web of Science ®], [Google Scholar]
- National Heart, Lung, and Blood Institute in cooperation with National Institute of Diabetes, & Digestive, and Kidney Diseases . ( 1998 ). On the identification, evaluation, and treatment of overweight and obesity in adults (The Evidence Report, NIH Publications No. 98-4083). Washington, DC: U.S. Department of Health and Human Services. [Google Scholar]
- Nielsen-Bohlman , L. , Panzer , A. M. , & Kindig , D. A. (Eds.). ( 2004 ). Health literacy, A prescription to end confusion (Institute of Medicine Report) . Washington , DC : National Academies Press . [Google Scholar]
- Office of Health Status Monitoring (OHSM). Hawaii Health Survey. Standard Output. Hawai'i Department of Health. Office of Health Status Monitoring. Retrieved July 3, 2011 from http: //hawaii.gov/health/statistics/hhs/ [Google Scholar]
- Osborn , C. Y. , Cavanaugh , K. , Wallston , K. A. , & Rothman , R. L. ( 2010 ). Self-efficacy links health literacy and numeracy to glycemic control . Journal of Health Communication , 15 ( Suppl 2 ), 146 – 158 . [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
- Osborn , C. Y. , Paasche-Orlow , M. K. , Bailey , S. C. , & Wolf , M. S. ( 2011 ). The mechanisms linking health literacy to behavior and health status . American Journal of Health Behavior , 35 ( 1 ), 118 – 128 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Paasche-Orlow , M. K. , & Wolf , M. S. ( 2007 ). The causal pathways linking health literacy to health outcomes . American Journal of Health Behavior , 31 ( Suppl 1 ), S19 – S26 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Panapasa , S. , Mau , M. K. , Williams , D. R. , & McNally , J. W. ( 2010 ). Mortality patterns of native Hawaiians across their lifespan. 1990–2000 . American Journal of Public Health , 111 ( 10 ), 2304 – 2310 . [Google Scholar]
- Park , C. B. , Braun , K. L. , Horiuchi , B. Y. , Tottori , C. , & Onaka , A. T. ( 2009 ). Longevity disparities in multiethnic Hawai'i: An analysis of 2000 life tables . Public Health Reports , 124 ( 4 ), 579 – 584 . [Web of Science ®], [Google Scholar]
- Pourat , N. , Kakawa-Singer , M. , Breen , N. , & Sripipatana , A. (2010). Access versus acculturation: Indentifying modifiable factors to promote cancer screening among Asian American women. Medical Care , 48(12), 1088–1096. [Google Scholar]
- Saha , S. ( 2006 ). Improving literacy as a means to reducing health disparities . Journal General Internal Medicine , 21 ( 8 ), 893 – 895 . [Crossref], [Web of Science ®], [Google Scholar]
- Sarkar , U. , Karter , A. , Liu , J. Y. , Adler , N. E. , Nguyen , R. , Lopez , A. , & Schillinger , D. ( 2010 ). The literacy divide: Health literacy and the use of an Internet-based patient portal in an integrated health system—results from the diabetes study of Northern California (DISTANCE) . Journal of Health Communication , 12 ( Supplement 2 ), 183 – 196 . [Google Scholar]
- Schillinger , D. , Barton , L. R. , Karter , A. J. , Wang , F. , & Adler , N. ( 2006 ). Does literacy mediate the relationship between education and health outcomes? A study of a low-income population with diabetes . Public Health Reports , 121 ( 3 ), 245 – 254 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Sentell , T. , & Halpin , H. ( 2006 ). The importance of adult literacy in understanding health disparities . Journal of General Internal Medicine , 21 ( 8 ), 852 – 856 . [Google Scholar]
- StataCorp . ( 2009 ). Stata Statistical Software: Release 11. College Station, TX: StataCorp. [Google Scholar]
- Tang , Y. H. , Pang , S. M. , Chan , M. F. , Yeung , G. S. , & Yeung , V. T. ( 2008 ). Health literacy, complication awareness, and diabetic control in patients with type 2 diabetes mellitus . Journal of Advanced Nursing , 62 ( 1 ), 74 – 83 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Tanjasiri , S. P. , Wallace , S. P. , & Shibata , K. ( 1995 ). Picture imperfect: Hidden problems among Asian Pacific Islander elderly . Gerontologist , 25 ( 6 ), 753 – 760 . [Google Scholar]
- Tokuda , Y. , Doba , N. , Butler , J. P. , & Paasche-Orlow , M. K. ( 2009 ). Health literacy and physical and psychological wellbeing in Japanese adults . Patient Education and Counseling , 75 ( 3 ), 411 – 417 . [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- U.S. Department of Health and Human Services . ( 2000 ). Healthy People 2010. (2nd ed). In Understanding and improving health and objectives for improving health (2 vols.). Washington, DC: U.S. Government Printing Office. [Google Scholar]
- U.S. Department of Health and Human Services . ( 2008 ) Poverty Guidelines. The 2008 HHS Poverty Guidelines. Retrieved from http://aspe.hhs.gov/poverty/08poverty.shtml [Google Scholar]
- Wallace , L. S. , Rogers , E. S. , Roskos , S. E. , Holiday , D. B. , & Weiss , B. D. ( 2006 ). Brief report: Screening items to identify patients with limited literacy skills . Journal of General Internal Medicine , 21 , 874 – 877 . [PubMed], [Web of Science ®], [Google Scholar]