- Email: [email protected]
Stemming the tide: Rising diabetes prevalence and ethnic subgroup variation among Asians in Los Angeles County
Preventive Medicine 63 (2014) 90–95
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Stemming the tide: Rising diabetes prevalence and ethnic subgroup variation among Asians in Los Angeles County Margaret Shih a,⁎, Yajun Du a, Amy S. Lightstone a, Paul A. Simon b,c, May C. Wang c a b c
Office of Health Assessment and Epidemiology, Los Angeles County Department of Public Health, 313 N. Figueroa St., Rm 127, Los Angeles, CA 90012, USA Division of Chronic Disease and Injury Prevention, Los Angeles County Department of Public Health, 3530 Wilshire Blvd., 8th Floor, Los Angeles, CA 90010, USA UCLA Fielding School of Public Health, 650 Charles E. Young Drive South, Los Angeles, CA 90095, USA
a r t i c l e
i n f o
Available online 20 March 2014 Keywords: Diabetes Asians Body mass index Ethnicity
a b s t r a c t Objective. The primary objective of this analysis was to examine the burden of diabetes among Asians and Asian subgroups in Los Angeles County, which has the largest county population of Asians in the U.S. Method. Data were analyzed from 6 cycles of the Los Angeles County Health Survey, 1997–2011 (n = 47,282). Asian adults (n = 4672) were categorized into the following ethnic subgroups: Chinese, Filipino, Korean, Japanese, Vietnamese, South Asian, and Other Asian. Descriptive and multivariable logistic regression analyses were conducted to examine trends in prevalence, prevalence among Asian subgroups, and factors associated with diabetes. Results. In 2005, we observed a rapid increase in diabetes prevalence among Asians compared to whites despite consistently lower BMI relative to other racial/ethnic groups. Diabetes prevalence was significantly higher among Filipinos and South Asians (N10%) compared to East Asians and Vietnamese (b 7%). After adjusting for all covariates, Asians who were older, non-drinkers, insured, and overweight or obese were found to have increased odds of diabetes. Conclusion. Diabetes prevalence is increasing more rapidly among Asians compared to whites despite overall lower BMI. The significant heterogeneity among Asian subgroups highlights the need for disaggregated data and additional research to develop culturally appropriate interventions for diabetes prevention and control. © 2014 Elsevier Inc. All rights reserved.
Introduction Diabetes is a well-recognized global public health crisis. The number of people affected worldwide is projected to increase from 366 million in 2011 to 552 million in 2030, with the largest increases expected in China and India (Chan et al., 2009; Danaei et al., 2011; International Diabetes Federation, 2011). Asians are the fastest growing race in the United States (US Census Bureau, 2012), and like other racial/ethnic groups, have been experiencing an increase in diabetes rates (Lee et al., 2011). However, the increase in diabetes among Asian Americans may be under-recognized. The lack of attention to diabetes as an increasing problem among Asians in the United States may partly be because Asians with diabetes are less likely to present with overt obesity. Obesity, clinically defined among adults as having a Body Mass Index (BMI) of ≥30 kg/m2 (WHO, 1995), is a major risk factor for diabetes. However, the appropriateness of this definition for identifying individuals with a high percentage of ⁎ Corresponding author. E-mail addresses: [email protected] (M. Shih), [email protected] (Y. Du), [email protected] (A.S. Lightstone), [email protected] (P.A. Simon), [email protected] (M.C. Wang).
http://dx.doi.org/10.1016/j.ypmed.2014.03.016 0091-7435/© 2014 Elsevier Inc. All rights reserved.
body fat has been questioned. Asians have a higher percentage of body fat than whites of the same age, sex, and BMI (Deurenberg et al., 2002; He et al., 2001; Wang et al., 1994; WHO Expert Consultation, 2004), and experience greater obesity-related health problems at lower BMIs relative to other racial/ethnic groups (Deurenberg-Yap et al., 2002; Lee et al., 2011; McNeely and Boyko, 2004; Yoon et al., 2006). These observations led the World Health Organization (WHO) to convene the Expert Consultation on BMI in Asian populations in 2002, which recommended retaining existing international BMI classifications of weight status, but adding additional (lower) cutoffs for informing and triggering public health and clinical actions among Asians (WHO, 2000; WHO Expert Consultation, 2004). However, there is still considerable controversy regarding BMI cutoffs, and these lower cut-points have not been officially adopted. Additionally, the extreme heterogeneity of the Asian population has made their burden of diabetes difficult to assess. Although there is increased recognition of the need for disaggregated ethnicity data, few studies in the United States report disaggregated data for Asians, and there has been a lack of standardized reporting on the prevalence of diabetes among Asian subgroups (Lee et al., 2011; Staimez et al., 2013). The primary objective of this analysis was to examine the burden of self-reported diabetes among Asians and Asian subgroups in Los
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
Angeles County, which has the largest county population of Asians in the United States. We examined trends in diabetes prevalence among Asians and non-Asians and report on the prevalence of diabetes among Asian ethnic subgroups and factors associated with diabetes among Asians overall. Methods Data source We used data from the Los Angeles County Health Survey (LACHS), a periodic, random-digit-dial telephone survey of the non-institutionalized population of Los Angeles County (Simon et al., 2001). The survey collects information on demographics, health conditions, health-related behaviors, health insurance coverage, and access to care among county residents. Due to the diversity of the population, interviews were conducted in English, Spanish, Mandarin, Cantonese, Korean, and Vietnamese, with one adult randomly selected from each household. Details regarding the survey design and weighting methodology are reported elsewhere (Simon et al., 2001). For this study, we analyzed data from the adult component of six iterations of the LACHS: 1997, 1999, 2002, 2005, 2007, and 2011. Two methodologic changes were made to the 2011 LACHS to maintain survey representativeness and validity. A dual frame sampling approach was used to incorporate cell phones into the sample, and a more sophisticated raking (i.e. sample balancing) procedure was adopted for weighting the survey (Los Angeles County Health Survey Methodology Report, 2012). These changes were similar to those made to the Behavioral Risk Factor Surveillance System in 2011 (CDC, 2012).
Study population From 1997 to 2011, a total of 42,610 non-Asian (white, Latino, black) and 4672 Asian adults (aged 18 years or older) residing in Los Angeles County were surveyed and included in this analysis. The Asian adults were categorized into the following subgroups: Chinese (n = 1693), Filipino (n = 737), Korean (n = 786), Japanese (n = 558), Vietnamese (n = 313), South Asian (n = 269) and Other Asian (n = 316). ‘South Asian’ included Asian Indians, Sri Lankans, Pakistanis, and Bangladeshis. ‘Other Asian’ included, but was not limited to, southeast Asians (i.e. persons from Cambodia, Thailand, Laos, etc.) and Asians who did not specify any ethnicity. Native Hawaiians and Pacific Islanders were excluded from the analysis. Statistical weighting was utilized to generalize the sample survey data to the overall Los Angeles County population (Los Angeles County Health Survey Methodology Report, 2012).
Variables Diabetes Individuals were classified as having diabetes if they gave a positive response to the following question: “Have you ever been told by a doctor or other health professional that you have diabetes or sugar diabetes?”
Covariates Independent variables that have previously been demonstrated to be associated with diabetes were included as covariates in the analysis. Sociodemographic variables examined included age group (18–49, 50–64, 65+ years), gender (male, female), education (b high school, high school graduate, some college, college graduate+), and household income (0–99% of the Federal Poverty Level (FPL), 100–199% FPL, ≥ 200% FPL). Other covariates of interest included nativity and years living in the U.S. (US born, foreign born living in the US for 0–9 years, foreign-born living in the US for ≥ 10 years), smoking status (non- or former smoker, current smoker), alcohol use (nondrinker, low-moderate drinker, heavy or binge drinker), insurance status (uninsured, insured), physical activity [meets national physical activity guidelines (US DHHS, 2000), some activity but not meeting guidelines, inactive], and weight status (under-to-normal weight, overweight, obese). For non-Asians, weight status classifications were defined using standard BMI cutoffs recommended by the CDC and WHO (BMI b 25, 25 ≤ BMI b 30, BMI ≥ 30 kg/m2) (WHO, 1995). For Asians, we applied these standard cutoffs and also the recommended Asianspecific BMI cutoffs (BMI b 23, 23 ≤BMI b 27.5, BMI ≥ 27.5 kg/m2) (WHO Expert Consultation, 2004).
91
Statistical analyses We examined trends in diabetes prevalence by race/ethnicity using the Cochran–Armitage test for trend. We also compared diabetes prevalence and mean BMI between Asians and whites, controlling for age. For comparisons among Asian subgroups, we merged all 6 cycles of the LACHS and examined diabetes prevalence among Chinese, Filipinos, Korean, Japanese, Vietnamese, South Asians, and Other Asians compared to whites. We used logistic regression modeling to compare odds of diabetes by Asian subgroups using whites as the reference group and adjusting for age, gender, and BMI. To investigate factors associated with diabetes among the full sample of Asians, we constructed additional logistic regression models to obtain crude and adjusted odds ratios (ORs). For this stage of modeling, we merged 4 cycles of the LACHS (2002–2011), excluding the 1997 and 1999 surveys because comparable data on smoking status, insurance, and physical activity were not available for these earlier survey years. Education level was not found to be significant in preliminary models and was not included in the final models. A significance level of α = 0.05 was used throughout. The Hosmer–Lemeshow Goodness-of-Fit test was used to assess model fit. All analyses were performed in SAS version 9.3 (SAS Institute, Inc., Cary, NC).
Results Summary characteristics In 2007–2011, proportionately more Asians than non-Asians were aged ≥ 65 years (15.2% vs. 13.9%); this pattern was not observed in 1997–1999 (Table S1). For all survey years, a higher percentage of Asians than non-Asians had a college degree and a household income greater than 200% FPL, while a lower percentage of Asians were born in the U.S. than non-Asians; at the same time, among those who were foreign-born, proportionately more Asians had lived in the U.S. for ten years or more than non-Asians. Compared to non-Asians, Asians were also less likely to be current smokers and heavy or binge drinkers, more likely to have health insurance, and less likely to meet physical activity guidelines. When using Asian-specific BMI cutoffs, 46% of Asians were categorized as overweight or obese compared to only 26% when using CDC/WHO standard cutoffs. Trends in diabetes prevalence and BMI among Asians vs. non-Asians The age-adjusted prevalence of diabetes increased from 1997 to 2011 for all major racial/ethnic groups (p b 0.005). Among Asians overall, the prevalence of diabetes increased by 66%, from 5.8% in 1997, to 9.6% in 2011 (Table 1). The prevalence of diabetes among Asians was lower than that of Latinos and blacks, and was similar to that of whites from 1997 to 2005, after which it appeared to increase more rapidly and was significantly higher among Asians compared to whites (Fig. 1). Mean BMI also increased among Asians during this period, but it remained the lowest compared to other racial/ethnic groups for all years, and was significantly lower than that of whites, who had the second lowest mean BMI (Table 1). Factors associated with diabetes among Asians Diabetes prevalence among Asians was highest among adults who were aged 65 years or older, living below poverty, foreign-born but living in the U.S. for 10 years or more, non-drinkers, insured, physically inactive, and overweight or obese (Table 2). Asians with diabetes who were categorized as overweight and obese using either Asian-specific or CDC/WHO standard BMI cutoffs were two times and five times more likely to be diagnosed with diabetes, respectively, than those who were under-to-normal weight. After adjusting for all covariates, Asians who were older, non-drinkers, insured, and overweight or obese were found to have increased odds of diabetes. Asian adults who were 65 years or older were 13 times more likely to report being diagnosed with diabetes compared to those aged 18–
92
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
Table 1 Trend in age-adjusted diabetes prevalence and mean BMI by race/ethnicity, Los Angeles County, 1997–2011 (N = 47,282).a Race/ethnicity
1997 (n = 7840)
1999 (n = 8166)
2002 (n = 7948)
2005 (n = 8364)
2007 (n = 7020)
2011 (n = 7944)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Latino White Black Asianb
9.5 (7.9, 11.0) 4.6 (3.9, 5.4) 10.1 (8.0, 12.2) 5.8 (3.8, 7.8)
11.3 (9.6, 13.1) 5.5 (4.7, 6.4) 9.5 (7.2, 11.8) 5.4 (3.4, 7.5)
11.4 (9.8, 13.0) 5.4 (4.6, 6.2) 9.4 (7.3, 11.5) 5.0 (3.2, 6.8)
12.3 (10.8, 13.8) 5.6 (4.8, 6.5) 12.0 (9.7, 14.4) 7.1 (5.3, 9.0)
12.8 (11.4, 14.3) 5.7 (5.0, 6.5) 11.4 (8.7, 14.1) 9.1 (7.1, 11.2)⁎⁎
13.5 (11.8, 15.2) 6.7 (5.6, 7.9) 12.4 (9.8, 15.1) 9.6 (7.4, 11.8)⁎
Race/ethnicity
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Latino White Black Asianb
26.5 (26.3, 26.7) 24.9 (24.7, 25.0) 26.9 (26.6, 27.3) 22.9 (22.6, 23.2)⁎⁎⁎
27.0 (26.8, 27.2) 25.4 (25.2, 25.6) 26.9 (26.5, 27.4) 23.3 (22.9, 23.6)⁎⁎⁎
27.3 (27.1, 27.5) 25.4 (25.2, 25.6) 27.9 (27.5, 28.4) 23.5 (23.1, 23.8)⁎⁎⁎
27.9 (27.6, 28.1) 25.5 (25.2, 25.7) 27.7 (27.2, 28.3) 23.7 (23.4, 24.0)⁎⁎⁎
27.8 (27.6, 28.1) 25.9 (25.6, 26.2) 27.8 (27.2, 28.3) 23.9 (23.4, 24.4)⁎⁎⁎
28.3 (28.0, 28.5) 25.6 (25.3, 25.9) 28.4 (27.8, 28.9) 24.5 (24.1, 24.9)⁎⁎⁎
a
Test for trend: p = 0.002 for blacks and p b 0.001 for Latinos, whites, and Asians. Comparing Asians vs. whites only, controlling for age. ⁎ p b 0.05. ⁎⁎ p = 0.001. ⁎⁎⁎ p b 0.0001. b
49 years, and Asian adults who were categorized as overweight or obese using Asian-specific BMI cutoffs were, respectively, two times and six times more likely to report being diagnosed with diabetes compared to their under-to-normal weight peers. Asian adults who drank alcohol were less likely to be diagnosed with diabetes compared to non-drinkers. Prevalence of diabetes among Asian subgroups Disaggregating the Asians revealed significant heterogeneity in diabetes prevalence among Asian ethnicities. The age-adjusted diabetes prevalence was highest among Filipinos (10.7%), South Asians (11.8%) and Other Asians (14.3%), and lowest among East Asian ethnicities (Chinese, Vietnamese, Korean, and Japanese), whose rates were comparable to that of whites (Table 3). A similar pattern was seen in mean BMI among Asian ethnicities. After adjusting for age, gender, and BMI, Filipino, South Asian, and Other Asian adults were found to be twice as likely as whites to report a diagnosis of diabetes. Discussion
Age-Adjusted Diabetes Prevalence (%, 95% CI)
In this analysis of data gathered between 1997 and 2011 from a racially and ethnically diverse sample of 47,282 Los Angeles residents, of whom 10% were Asian, we observed a rapid increase in the prevalence of diabetes among Asians compared to whites beginning in 2005. 16
12
* *
8
Asian White
4
0 1995
1997
1999
2001
2003
Year
2005
2007
2009
2011 *P<0.05
Fig. 1. Age-adjusted diabetes prevalence (±95% CI) among Asians and whites, Los Angeles County, 1997–2011.
Specifically, while diabetes rates for both Asians and whites hovered around 5% in 2002, the prevalence of diabetes for Asians increased to 7.1% in 2005, accelerating to 9.6% in 2011. In comparison, the prevalence of diabetes for whites increased minimally from 5.4% in 2002 to 6.7% in 2011. Our finding – that the prevalence of diabetes among Asians residing in Los Angeles County has been increasing rapidly – parallels reports of increasing rates of diabetes among Asians living in rapidly urbanizing countries such as China and India (Chan et al., 2009; Cockram, 2000; Yoon et al., 2006). Indeed, diabetes has emerged as a growing public health problem in many Asian countries. Changing lifestyles that have adversely affected diet and physical activity and led to increased obesity risk have been blamed (Ramachandran et al., 2012). Further, the epidemiologic literature has consistently reported that Asians living in their native countries and Asian immigrants living in western countries show a higher risk of diabetes and other obesity-related chronic health conditions at lower levels of BMI (Odegaard et al., 2009; Razak et al., 2005, 2007; Wang et al., 1994). Consistent with this finding, we observed that BMI, while increasing over the years in all racial/ethnic groups, was consistently lower among Asians compared to all other racial/ethnic groups, even when the prevalence of diabetes for Asians exceeded that of whites. There has been ongoing debate about the need for country- or ethnicity-specific BMI cut-points, particularly among Asian populations (Deurenberg et al., 2002; Low et al., 2009). BMI is an imperfect proxy for body fatness, and may not account adequately for variation in percent body fat and lean muscle mass, bone density, or fat distribution. Other measures of body fatness have been investigated, including bioimpedance, waist-to-hip ratio, waist circumference, and sagittal abdominal diameter. The sagittal abdominal diameter is a promising anthropometric proxy measure for predicting visceral adiposity, which has high relevance for Asian populations (Yim et al., 2010). Further research is needed to determine which of these, separately or in combination, is most predictive of cardiometabolic risk in multi-ethnic populations. The findings from our study suggest that the increasing trend in diabetes prevalence among Asians residing in highly urbanized Los Angeles County varies substantially among the various Asian subgroups. In particular, Filipinos and South Asians have considerably higher rates of diabetes (N10%) than East Asians and Vietnamese (b7%). This heterogeneity across ethnic subgroups has been noted in previous studies (Lam and LeRoith, 2012; Ma and Chan, 2013) and calls for the need for more comparative studies to examine genetic factors, gene–environment interactions, and factors during early development. Factors identified by our study as being associated with diabetes among all Asians after adjusting for covariates are older age (≥65 years), being overweight or obese,
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
being a non-drinker, and being insured. The inverse association between diabetes and heavy or binge drinking in our study was unexpected. Previous studies have suggested a J-shaped relationship between alcohol intake and diabetes risk, with mildly protective effects at low-tomoderate levels of consumption, and increased risk with heavy drinking and binge drinking among both men and women (Baliunas et al., 2009; Pietraszek et al., 2010). A potential explanation for our finding is that we were unable to identify lifetime abstainers and thus were unable to account for a potential sick quitter effect (i.e. former drinkers who quit drinking due to being diagnosed with an illness such as diabetes). This would tend to bias results toward overestimating the benefits of moderate consumption and underestimating the risks of heavy consumption. Additionally, since the number of Asians in the sample who were heavy drinkers was small, we were unable to distinguish respondents with high levels of consumption from those with very high consumption. This may have masked potential negative effects among drinkers with very high consumption. That increased diabetes risk was associated with having insurance may reflect the increased likelihood of being tested for diabetes (as well as other diseases) among those with health insurance (Barcellos et al., 2012). There are several limitations to the present study. First, the data source did not contain information on what type of diabetes a respondent
93
had, so we were unable to distinguish between Type 1 and Type 2 diabetes. Second, the data were cross-sectional and cannot be used to infer causation. Third, since the data were self-reported, there is the possibility of recall and social desirability bias, which may have led to misclassification of diabetes or underestimation of BMI (Merrill and Richardson, 2009). However, self-reported diabetes has been shown to have reasonable validity for public health surveillance (Margolis et al., 2008; Martin et al., 2000), and we found no evidence that Asians differentially report diabetes or BMI compared to other racial/ethnic groups. Diabetes is likely underreported, but Asians appear to have a similar rate of undiagnosed diabetes as whites (Thorpe et al., 2009). Fourth, the cooperation rate of the most recent LACHS in 2011 was 66% and reflects the ongoing challenges of conducting telephone-based surveys. However, the LACHS response rates have been comparable to those reported for other large state-based telephone health surveys (CDC, 2013), and the weighted demographic profile of the respondents, including the Asian subgroups, was similar to the population in Los Angeles County (unpublished results). Further, rigorous adjustments were made to address residual bias. Finally, while a strength of our study was the use of a large, unique dataset which allowed us to examine diabetes prevalence by Asian subgroups, the sample size was insufficient to further analyze risk factors for diabetes by Asian subgroup.
Table 2 Unadjusted diabetes prevalence, crude and adjusted odds of diabetes in Asians (N = 3234), Los Angeles County, 2002–2011. Characteristic
Age (years) 18–49 50–64 65+ Gender Male Female Household income 0–99% FPL 100–199% FPL 200% or above FPL Nativity US Born Foreign born in US 0–9 years Foreign born in US 10+ years Smoking status Non-smoker Current smoker Alcohol drinking Non-drinkers Low-mod drinkers Heavy or binge drinkers Insurance status Uninsured Insured Physical activityd Inactive Some PA Meets the guidelines BMI (Asian standard)e Under/normal weight Overweight Obesity BMI (CDC/WHO standard) Under/normal weight Overweight Obesity a
% (95% CI)
Crude OR
Adjusted OR: model 2a
Adjusted OR: model 3b
Asian
OR (95% CI)
OR (95% CI)
OR (95% CI)
2.4 (1.6, 3.3) 12.3 (9.8, 14.9) 22.4 (18.3, 26.5)
1.00 5.72 (3.75, 8.71) 11.77 (7.71, 17.98)
1.00 6.37 (4.10, 9.89) 15.15 (9.66, 23.78)
1.00 5.81 (3.71, 9.11) 12.63 (7.78, 20.50)
8.0 (6.5, 9.5) 7.3 (5.9, 8.7)
1.00 0.92 (0.69, 1.23)
1.00 1.04 (0.75, 1.45)
1.00 0.91 (0.64, 1.29)
10.1 (7.2, 13.0) 7.6 (5.3, 9.8) 7.0 (5.8, 8.3)
1.00 0.74 (0.47, 1.17) 0.67 (0.46, 0.98)
1.00 1.00 (0.56, 1.78) 1.14 (0.68, 1.91)
6.1 (4.3, 7.8) 2.7 (1.1, 4.2)c 10.2 (8.6, 11.8)
1.00 0.41 (0.21, 0.80) 1.76 (1.24, 2.51)
1.00 0.75 (0.33, 1.70) 1.46 (0.96, 2.22)
7.6 (6.6, 8.7) 6.7 (3.7, 9.7)
1.00 0.84 (0.51, 1.39)
1.00 1.54 (0.87, 2.72)
9.7 (8.2, 11.2) 4.7 (3.3, 6.2) 2.8 (0.8, 4.8)c
1.00 0.45 (0.31, 0.65) 0.26 (0.12, 0.55)
1.00 0.44 (0.29, 0.68) 0.38 (0.17, 0.85)
3.1 (1.7, 4.6)c 8.6 (7.4, 9.8)
1.00 2.98 (1.80, 4.93)
1.00 1.97 (1.09, 3.58)
8.9 (7.2, 10.6) 7.3 (4.6, 10.1) 6.4 (5.0, 7.7)
1.00 0.81 (0.51, 1.29) 0.70 (0.51, 0.96)
1.00 0.98 (0.57, 1.66) 0.83 (0.58, 1.20)
3.9 (2.8, 5.0) 8.0 (6.2, 9.7) 18.2 (14.0, 22.4)
1.00 2.12 (1.45, 3.11) 5.49 (3.66, 8.25)
5.1 (4.0, 6.1) 9.8 (7.6, 12.1) 22.7 (15.6, 29.8)
1.00 2.00 (1.43, 2.81) 5.43 (3.45, 8.56)
1.00 2.09 (1.42, 3.09) 6.43 (4.14, 9.98)
1.00 1.93 (1.29, 2.89) 6.43 (4.07, 10.17)
Model 2: Adjusted for age (categorical), gender, and BMI (Asian standard). Model 3 (full model): Adjusted for age (categorical), gender, BMI (Asian standard), Federal Poverty Level, nativity and years living in U.S., smoking status, alcohol drinking status, insurance status, and physical activity. c Estimate has a relative standard error ≥ 23%. d To meet physical activity guidelines at least one of the following criteria must be fulfilled: 1) vigorous activity for 20+ min, N3 days/week, 2) moderate activity for 30+ min, N5 days/ week, and 3) a combination of vigorous and moderate activity meeting the time criteria for N5 days/week. e Asian BMI standard: Overweight: 23 ≤ BMI b 27.5 kg/m2; obesity: BMI ≥ 27.5 kg/m2. b
94
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
Table 3 Age-adjusted mean BMI, diabetes prevalence, and adjusted odds of diabetes among Asian subgroups, Los Angeles County, 1997–2011. Race/ethnicity
n
Mean BMI (95% CI)
Diabetes prevalence % (95% CI)
Adjusted ORb (95% CI)
White Asian Chinese Vietnamese Korean Japanese Filipino South Asianc Other Asiand
20,027 4672 1693 313 786 558 737 269 316
25.4 (25.3, 25.5) 23.7 (23.5, 23.8) 23.0 (22.8, 23.3) 22.9 (22.3, 23.5) 23.4 (23.1, 23.7) 23.9 (23.6, 24.3) 24.5 (24.2, 24.9) 25.0 (24.5, 25.5) 25.1 (24.5, 25.8)
5.7 (5.3, 6.0) 7.3 (6.4, 8.1) 5.2 (4.1, 6.4) 5.9 (3.0, 8.8)a 6.3 (4.6, 7.9) 6.7 (4.8, 8.7) 10.7 (7.9, 13.4) 11.8 (6.5, 17.1) 14.3 (8.8, 19.8)
1.00 – 1.03 (0.79, 1.34) 1.17 (0.65, 2.11) 1.39 (0.98, 1.97) 1.37 (0.96, 1.94) 1.98 (1.43, 2.72) 1.96 (1.11, 3.49) 2.10 (1.20, 3.70)
a
Estimate has a relative standard error ≥23%. Model included Asian subgroups, age (categorical), gender, and BMI (Asian standard: under-to-normal weight: BMI b 23.0 kg/m2; overweight: 23.0 ≤ BMI b 27.5 kg/m2; obese: BMI ≥ 27.5 kg/m2), using whites as the reference group. c Includes Asian Indian, Sri Lankan, Pakistani, and Bangladeshi. d Includes Asians not listed in the other subgroups, including southeast Asians (i.e. persons from Cambodia, Thailand, Laos, etc.) and Asians who did not specify any ethnicity. b
Despite these limitations, this is one of the first studies to report the rapid increase in diabetes among Asians using a large population-based sample, and to report differences in diabetes rates among Asian ethnic subgroups. Los Angeles County is home to the largest and one of the most diverse Asian populations in the U.S., allowing us to report disaggregated results for Asians. We were able to identify only one other study which reported differences in diabetes rates among Asian subgroups in the U.S. using a large population-based sample (Lee et al., 2011). Our study confirmed Lee's finding that Asian Americans appear to be at higher risk of diabetes compared to whites, despite having lower BMI. We were able to extend these findings by examining other Asian subgroups, and showing that diabetes is increasing more rapidly among Asians compared to whites. We conclude that diabetes among Asian Americans is a serious public health problem that has not been adequately recognized. There are substantial disparities among Asian subgroups that are masked by reporting data in aggregated form, and the current international BMIcutoffs for defining obesity may miss a significant percentage of diabetic Asians when used for targeted screening. More research is needed to identify which screening methods are most effective for ethnic populations and to ascertain the ethnicity specific needs of the various Asian subgroups. Such information is necessary for the development of culturally appropriate intervention programs to prevent and control diabetes in a population that is experiencing rapidly increased rates of diabetes. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ypmed.2014.03.016. Conflict of interest statement The authors declare that there are no conflicts of interest.
Acknowledgments The authors wish to thank Dr. Yan Cui for her assistance with the data analysis. References Baliunas, D.O., Taylor, B.J., Irving, H., Roerecke, M., Patra, J., Mohapatra, S., Rehm, J., 2009. Alcohol as a risk factor for type 2 diabetes: a systematic review and meta-analysis. Diabetes Care 32, 2123–2132. Barcellos, S.H., Goldman, D.P., Smith, J.P., 2012. Undiagnosed disease, especially diabetes, casts doubt on some of reported health ‘advantage’ of recent Mexican immigrants. Health Aff. (Millwood) 31, 2727–2737. Centers for Disease Control and Prevention, 2012. Methodologic changes in the Behavioral Risk Factor Surveillance System in 2011 and potential effects on prevalence estimates. MMWR 61, 410–413. Centers for Disease Control and Prevention, 2013. Behavioral Risk Factor Surveillance System 2012 Summary Data Quality Report. Accessed January 16, 2014. Available from http://www.cdc.gov/brfss/annual_data/2012/pdf/SummaryDataQualityReport2012_ 20130712.pdf.
Chan, J.C., Malik, V., Jia, W., Kadowaki, T., Yajnik, C.S., Yoon, K.H., Hu, F.B., 2009. Diabetes in Asia: epidemiology, risk factors, and pathophysiology. JAMA 301, 2129–2140. Cockram, C.S., 2000. The epidemiology of diabetes mellitus in the Asia-Pacific region. Hong Kong Med. J. 6, 43–52. Danaei, G., Finucane, M.M., Lu, Y., Singh, G.M., Cowan, M.J., Paciorek, C.J., Lin, J.K., Farzadfar, F., Khang, Y.H., et al., Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Blood Glucose), 2011. National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants. Lancet 378, 31–40. Deurenberg, P., Deurenberg-Yap, M., Guricci, S., 2002. Asians are different from Caucasians and from each other in their body mass index/body fat per cent relationship. Obes. Rev. 3, 141–146. Deurenberg-Yap, M., Chew, S.K., Deurenberg, P., 2002. Elevated body fat percentage and cardiovascular risks at low body mass index levels among Singaporean Chinese, Malays, and Indians in Singapore. Obes. Rev. 3, 209–215. He, M., Tan, K.C., Li, E.T., Kung, A.W., 2001. Body fat determination by dual energy X-ray absorptiometry and its relation to body mass index and waist circumference in Hong Kong Chinese. Int. J. Obes. Relat. Metab. Disord. 25, 748–752. International Diabetes Federation, 2011. IDF Diabetes Atlas, 5th ed. (Brussels, Belgium. Accessed July 3, 2013: Available from: http://www.idf.org/diabetesatlas). Lam, D.W., LeRoith, D., 2012. The worldwide diabetes epidemic. Curr. Opin. Endocrinol. Diabetes Obes. 19, 93–96. Lee, J.W., Brancati, F.L., Yeh, H.C., 2011. Trends in the prevalence of type 2 diabetes in Asians versus whites: results from the United States National Health Interview Survey, 1997–2008. Diabetes Care 34, 353–357. 2011 Los Angeles County Health Survey (LACHS): survey methodology report. Accessed July 3, 2013. Available from: http://publichealth.lacounty.gov/ha/docs/2011% 20LACHS/LACHS%20Methodological%20Report_REV121029.pdf. Low, S., Chin, M.C., Ma, S., Heng, D., Deurenberg-Yap, M., 2009. Rationale for redefining obesity in Asians. Ann. Acad. Med. Singapore 38, 66–69. Ma, R.C.W., Chan, J.C.N., 2013. Type 2 diabetes in East Asians: similarities and differences with populations in Europe and the United States. Ann. N. Y. Acad. Sci. 1281, 64–91. Margolis, K.L., Lihong, Q., Brzyski, R., Bonds, D.E., Howard, B.V., Kempainen, S., Simin, L., Robinson, J.G., Safford, M.M., et al., 2008. Validity of diabetes self-reports in the Women's Health Initiative: comparison with medication inventories and fasting glucose measurements. Clin. Trials 5, 240–247. Martin, L.M., Leff, M., Calonge, N., Garrett, C., Nelson, D.E., 2000. Validation of self-reported chronic conditions and health services in a managed care population. Am. J. Prev. Med. 18, 215–218. McNeely, M.J., Boyko, E.J., 2004. Type 2 diabetes prevalence in Asian Americans: results of a national health survey. Diabetes Care 27, 66–69. Merrill, R.M., Richardson, J.S., 2009. Validity of self-reported height, weight, and body mass index: findings from the National Health and Nutrition Examination Survey, 2001–2006. Prev. Chronic Dis. 6 (A121) (Epub 2009 Sep 15). Odegaard, A.O., Koh, W.P., Vazquez, G., Arakawa, K., Lee, H.P., Yu, M.C., Pereira, M.A., 2009. BMI and diabetes risk in Singaporean Chinese. Diabetes Care 32, 1104–1106. Pietraszek, A., Gregersen, S., Hermansen, K., 2010. Alcohol and type 2 diabetes. A review. Nutr. Metab. Cardiovasc. Dis. 20, 366–375. Ramachandran, A., Chamukuttan, S., Shetty, S.A., Arun, N., Susairaj, P., 2012. Obesity in Asia—is it different from rest of the world. Diabetes Metab. Res. Rev. 28 (Suppl. 2), 47–51. Razak, F., Anand, S.S., Vuksan, V., Davis, B., Jacobs, R., Teo, K.K., Yusuf, S., Investigators, S.H. A.R.E., 2005. Ethnic differences in the relationships between obesity and glucosemetabolic abnormalities: a cross-sectional population-based study. Int. J. Obes. (Lond.) 29, 656–667. Razak, F., Anand, S.S., Shannon, H., Vuksan, V., Davis, B., Jacobs, R., Teo, K.K., McQueen, M., Yusuf, S., 2007. Defining obesity cut points in a multiethnic population. Circulation 115, 2111–2118. Simon, P.A., Wold, C.M., Cousineau, M.R., Fielding, J.E., 2001. Meeting the data needs of a local health department: the Los Angeles County Health Survey. Am. J. Publ. Health 91, 1950–1952.
M. Shih et al. / Preventive Medicine 63 (2014) 90–95 Staimez, L.R., Weber, M.B., Narayan, K.M., Oza-Frank, R., 2013. A systematic review of overweight, obesity, and type 2 diabetes among Asian American subgroups. Curr. Diabetes Rev. 9, 312–331. Thorpe, L.E., Upadhyay, U.D., Chamany, S., Garg, R., Mandel-Ricci, J., Kellerman, S., Berger, D.K., Frieden, T.R., Gwynn, C., 2009. Prevalence and control of diabetes and impaired fasting glucose in New York City. Diabetes Care 32, 57–62. U.S. Department of Health and Human Services, 2000. Healthy People 2010, Understanding and improving health and objectives for improving health2nd ed. United States Government Printing Office, Washington DC. United States Census Bureau, U.S. Department of Commerce, Economics and Statistics Administration, 2012. The Asian Population: 2010. 2010 Census Briefs. March. Accessed July 1, 2013: Available from: http://www.census.gov/prod/cen2010/briefs/c2010br11.pdf. Wang, J., Thornton, J.C., Russell, M., Burastero, S., Heymsfield, S., Pierson Jr., R.N., 1994. Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am. J. Clin. Nutr. 60, 23–28.
95
WHO Expert Consultation, 2004. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 363, 157–163. Erratum in: Lancet 363, 902. World Health Organization, 1995. Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Consultation. WHO Technical Report Series Number 854. World Health Organization, Geneva. World Health Organization, 2000. WHO: the Asia-Pacific perspective: redefining obesity and its treatment. Western Pacific Regional Office. World Health Organization, Geneva. Yim, J.Y., Kim, D., Lim, S.H., Park, M.J., Choi, S.H., Lee, C.H., Kim, S.S., Cho, S.H., 2010. Sagittal abdominal diameter is a strong anthropometric measure of visceral adipose tissue in the Asian general population. Diabetes Care 33, 2665–2675. Yoon, K.H., Lee, J.H., Kim, J.W., Cho, J.H., Choi, Y.H., Ko, S.H., Zimmet, P., Son, H.Y., 2006. Epidemic obesity and type 2 diabetes in Asia. Lancet 368, 1681–1688.
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Stemming the tide: Rising diabetes prevalence and ethnic subgroup variation among Asians in Los Angeles County Margaret Shih a,⁎, Yajun Du a, Amy S. Lightstone a, Paul A. Simon b,c, May C. Wang c a b c
Office of Health Assessment and Epidemiology, Los Angeles County Department of Public Health, 313 N. Figueroa St., Rm 127, Los Angeles, CA 90012, USA Division of Chronic Disease and Injury Prevention, Los Angeles County Department of Public Health, 3530 Wilshire Blvd., 8th Floor, Los Angeles, CA 90010, USA UCLA Fielding School of Public Health, 650 Charles E. Young Drive South, Los Angeles, CA 90095, USA
a r t i c l e
i n f o
Available online 20 March 2014 Keywords: Diabetes Asians Body mass index Ethnicity
a b s t r a c t Objective. The primary objective of this analysis was to examine the burden of diabetes among Asians and Asian subgroups in Los Angeles County, which has the largest county population of Asians in the U.S. Method. Data were analyzed from 6 cycles of the Los Angeles County Health Survey, 1997–2011 (n = 47,282). Asian adults (n = 4672) were categorized into the following ethnic subgroups: Chinese, Filipino, Korean, Japanese, Vietnamese, South Asian, and Other Asian. Descriptive and multivariable logistic regression analyses were conducted to examine trends in prevalence, prevalence among Asian subgroups, and factors associated with diabetes. Results. In 2005, we observed a rapid increase in diabetes prevalence among Asians compared to whites despite consistently lower BMI relative to other racial/ethnic groups. Diabetes prevalence was significantly higher among Filipinos and South Asians (N10%) compared to East Asians and Vietnamese (b 7%). After adjusting for all covariates, Asians who were older, non-drinkers, insured, and overweight or obese were found to have increased odds of diabetes. Conclusion. Diabetes prevalence is increasing more rapidly among Asians compared to whites despite overall lower BMI. The significant heterogeneity among Asian subgroups highlights the need for disaggregated data and additional research to develop culturally appropriate interventions for diabetes prevention and control. © 2014 Elsevier Inc. All rights reserved.
Introduction Diabetes is a well-recognized global public health crisis. The number of people affected worldwide is projected to increase from 366 million in 2011 to 552 million in 2030, with the largest increases expected in China and India (Chan et al., 2009; Danaei et al., 2011; International Diabetes Federation, 2011). Asians are the fastest growing race in the United States (US Census Bureau, 2012), and like other racial/ethnic groups, have been experiencing an increase in diabetes rates (Lee et al., 2011). However, the increase in diabetes among Asian Americans may be under-recognized. The lack of attention to diabetes as an increasing problem among Asians in the United States may partly be because Asians with diabetes are less likely to present with overt obesity. Obesity, clinically defined among adults as having a Body Mass Index (BMI) of ≥30 kg/m2 (WHO, 1995), is a major risk factor for diabetes. However, the appropriateness of this definition for identifying individuals with a high percentage of ⁎ Corresponding author. E-mail addresses: [email protected] (M. Shih), [email protected] (Y. Du), [email protected] (A.S. Lightstone), [email protected] (P.A. Simon), [email protected] (M.C. Wang).
http://dx.doi.org/10.1016/j.ypmed.2014.03.016 0091-7435/© 2014 Elsevier Inc. All rights reserved.
body fat has been questioned. Asians have a higher percentage of body fat than whites of the same age, sex, and BMI (Deurenberg et al., 2002; He et al., 2001; Wang et al., 1994; WHO Expert Consultation, 2004), and experience greater obesity-related health problems at lower BMIs relative to other racial/ethnic groups (Deurenberg-Yap et al., 2002; Lee et al., 2011; McNeely and Boyko, 2004; Yoon et al., 2006). These observations led the World Health Organization (WHO) to convene the Expert Consultation on BMI in Asian populations in 2002, which recommended retaining existing international BMI classifications of weight status, but adding additional (lower) cutoffs for informing and triggering public health and clinical actions among Asians (WHO, 2000; WHO Expert Consultation, 2004). However, there is still considerable controversy regarding BMI cutoffs, and these lower cut-points have not been officially adopted. Additionally, the extreme heterogeneity of the Asian population has made their burden of diabetes difficult to assess. Although there is increased recognition of the need for disaggregated ethnicity data, few studies in the United States report disaggregated data for Asians, and there has been a lack of standardized reporting on the prevalence of diabetes among Asian subgroups (Lee et al., 2011; Staimez et al., 2013). The primary objective of this analysis was to examine the burden of self-reported diabetes among Asians and Asian subgroups in Los
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
Angeles County, which has the largest county population of Asians in the United States. We examined trends in diabetes prevalence among Asians and non-Asians and report on the prevalence of diabetes among Asian ethnic subgroups and factors associated with diabetes among Asians overall. Methods Data source We used data from the Los Angeles County Health Survey (LACHS), a periodic, random-digit-dial telephone survey of the non-institutionalized population of Los Angeles County (Simon et al., 2001). The survey collects information on demographics, health conditions, health-related behaviors, health insurance coverage, and access to care among county residents. Due to the diversity of the population, interviews were conducted in English, Spanish, Mandarin, Cantonese, Korean, and Vietnamese, with one adult randomly selected from each household. Details regarding the survey design and weighting methodology are reported elsewhere (Simon et al., 2001). For this study, we analyzed data from the adult component of six iterations of the LACHS: 1997, 1999, 2002, 2005, 2007, and 2011. Two methodologic changes were made to the 2011 LACHS to maintain survey representativeness and validity. A dual frame sampling approach was used to incorporate cell phones into the sample, and a more sophisticated raking (i.e. sample balancing) procedure was adopted for weighting the survey (Los Angeles County Health Survey Methodology Report, 2012). These changes were similar to those made to the Behavioral Risk Factor Surveillance System in 2011 (CDC, 2012).
Study population From 1997 to 2011, a total of 42,610 non-Asian (white, Latino, black) and 4672 Asian adults (aged 18 years or older) residing in Los Angeles County were surveyed and included in this analysis. The Asian adults were categorized into the following subgroups: Chinese (n = 1693), Filipino (n = 737), Korean (n = 786), Japanese (n = 558), Vietnamese (n = 313), South Asian (n = 269) and Other Asian (n = 316). ‘South Asian’ included Asian Indians, Sri Lankans, Pakistanis, and Bangladeshis. ‘Other Asian’ included, but was not limited to, southeast Asians (i.e. persons from Cambodia, Thailand, Laos, etc.) and Asians who did not specify any ethnicity. Native Hawaiians and Pacific Islanders were excluded from the analysis. Statistical weighting was utilized to generalize the sample survey data to the overall Los Angeles County population (Los Angeles County Health Survey Methodology Report, 2012).
Variables Diabetes Individuals were classified as having diabetes if they gave a positive response to the following question: “Have you ever been told by a doctor or other health professional that you have diabetes or sugar diabetes?”
Covariates Independent variables that have previously been demonstrated to be associated with diabetes were included as covariates in the analysis. Sociodemographic variables examined included age group (18–49, 50–64, 65+ years), gender (male, female), education (b high school, high school graduate, some college, college graduate+), and household income (0–99% of the Federal Poverty Level (FPL), 100–199% FPL, ≥ 200% FPL). Other covariates of interest included nativity and years living in the U.S. (US born, foreign born living in the US for 0–9 years, foreign-born living in the US for ≥ 10 years), smoking status (non- or former smoker, current smoker), alcohol use (nondrinker, low-moderate drinker, heavy or binge drinker), insurance status (uninsured, insured), physical activity [meets national physical activity guidelines (US DHHS, 2000), some activity but not meeting guidelines, inactive], and weight status (under-to-normal weight, overweight, obese). For non-Asians, weight status classifications were defined using standard BMI cutoffs recommended by the CDC and WHO (BMI b 25, 25 ≤ BMI b 30, BMI ≥ 30 kg/m2) (WHO, 1995). For Asians, we applied these standard cutoffs and also the recommended Asianspecific BMI cutoffs (BMI b 23, 23 ≤BMI b 27.5, BMI ≥ 27.5 kg/m2) (WHO Expert Consultation, 2004).
91
Statistical analyses We examined trends in diabetes prevalence by race/ethnicity using the Cochran–Armitage test for trend. We also compared diabetes prevalence and mean BMI between Asians and whites, controlling for age. For comparisons among Asian subgroups, we merged all 6 cycles of the LACHS and examined diabetes prevalence among Chinese, Filipinos, Korean, Japanese, Vietnamese, South Asians, and Other Asians compared to whites. We used logistic regression modeling to compare odds of diabetes by Asian subgroups using whites as the reference group and adjusting for age, gender, and BMI. To investigate factors associated with diabetes among the full sample of Asians, we constructed additional logistic regression models to obtain crude and adjusted odds ratios (ORs). For this stage of modeling, we merged 4 cycles of the LACHS (2002–2011), excluding the 1997 and 1999 surveys because comparable data on smoking status, insurance, and physical activity were not available for these earlier survey years. Education level was not found to be significant in preliminary models and was not included in the final models. A significance level of α = 0.05 was used throughout. The Hosmer–Lemeshow Goodness-of-Fit test was used to assess model fit. All analyses were performed in SAS version 9.3 (SAS Institute, Inc., Cary, NC).
Results Summary characteristics In 2007–2011, proportionately more Asians than non-Asians were aged ≥ 65 years (15.2% vs. 13.9%); this pattern was not observed in 1997–1999 (Table S1). For all survey years, a higher percentage of Asians than non-Asians had a college degree and a household income greater than 200% FPL, while a lower percentage of Asians were born in the U.S. than non-Asians; at the same time, among those who were foreign-born, proportionately more Asians had lived in the U.S. for ten years or more than non-Asians. Compared to non-Asians, Asians were also less likely to be current smokers and heavy or binge drinkers, more likely to have health insurance, and less likely to meet physical activity guidelines. When using Asian-specific BMI cutoffs, 46% of Asians were categorized as overweight or obese compared to only 26% when using CDC/WHO standard cutoffs. Trends in diabetes prevalence and BMI among Asians vs. non-Asians The age-adjusted prevalence of diabetes increased from 1997 to 2011 for all major racial/ethnic groups (p b 0.005). Among Asians overall, the prevalence of diabetes increased by 66%, from 5.8% in 1997, to 9.6% in 2011 (Table 1). The prevalence of diabetes among Asians was lower than that of Latinos and blacks, and was similar to that of whites from 1997 to 2005, after which it appeared to increase more rapidly and was significantly higher among Asians compared to whites (Fig. 1). Mean BMI also increased among Asians during this period, but it remained the lowest compared to other racial/ethnic groups for all years, and was significantly lower than that of whites, who had the second lowest mean BMI (Table 1). Factors associated with diabetes among Asians Diabetes prevalence among Asians was highest among adults who were aged 65 years or older, living below poverty, foreign-born but living in the U.S. for 10 years or more, non-drinkers, insured, physically inactive, and overweight or obese (Table 2). Asians with diabetes who were categorized as overweight and obese using either Asian-specific or CDC/WHO standard BMI cutoffs were two times and five times more likely to be diagnosed with diabetes, respectively, than those who were under-to-normal weight. After adjusting for all covariates, Asians who were older, non-drinkers, insured, and overweight or obese were found to have increased odds of diabetes. Asian adults who were 65 years or older were 13 times more likely to report being diagnosed with diabetes compared to those aged 18–
92
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
Table 1 Trend in age-adjusted diabetes prevalence and mean BMI by race/ethnicity, Los Angeles County, 1997–2011 (N = 47,282).a Race/ethnicity
1997 (n = 7840)
1999 (n = 8166)
2002 (n = 7948)
2005 (n = 8364)
2007 (n = 7020)
2011 (n = 7944)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Diabetes prevalence % (95% CI)
Latino White Black Asianb
9.5 (7.9, 11.0) 4.6 (3.9, 5.4) 10.1 (8.0, 12.2) 5.8 (3.8, 7.8)
11.3 (9.6, 13.1) 5.5 (4.7, 6.4) 9.5 (7.2, 11.8) 5.4 (3.4, 7.5)
11.4 (9.8, 13.0) 5.4 (4.6, 6.2) 9.4 (7.3, 11.5) 5.0 (3.2, 6.8)
12.3 (10.8, 13.8) 5.6 (4.8, 6.5) 12.0 (9.7, 14.4) 7.1 (5.3, 9.0)
12.8 (11.4, 14.3) 5.7 (5.0, 6.5) 11.4 (8.7, 14.1) 9.1 (7.1, 11.2)⁎⁎
13.5 (11.8, 15.2) 6.7 (5.6, 7.9) 12.4 (9.8, 15.1) 9.6 (7.4, 11.8)⁎
Race/ethnicity
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Mean BMI (95% CI)
Latino White Black Asianb
26.5 (26.3, 26.7) 24.9 (24.7, 25.0) 26.9 (26.6, 27.3) 22.9 (22.6, 23.2)⁎⁎⁎
27.0 (26.8, 27.2) 25.4 (25.2, 25.6) 26.9 (26.5, 27.4) 23.3 (22.9, 23.6)⁎⁎⁎
27.3 (27.1, 27.5) 25.4 (25.2, 25.6) 27.9 (27.5, 28.4) 23.5 (23.1, 23.8)⁎⁎⁎
27.9 (27.6, 28.1) 25.5 (25.2, 25.7) 27.7 (27.2, 28.3) 23.7 (23.4, 24.0)⁎⁎⁎
27.8 (27.6, 28.1) 25.9 (25.6, 26.2) 27.8 (27.2, 28.3) 23.9 (23.4, 24.4)⁎⁎⁎
28.3 (28.0, 28.5) 25.6 (25.3, 25.9) 28.4 (27.8, 28.9) 24.5 (24.1, 24.9)⁎⁎⁎
a
Test for trend: p = 0.002 for blacks and p b 0.001 for Latinos, whites, and Asians. Comparing Asians vs. whites only, controlling for age. ⁎ p b 0.05. ⁎⁎ p = 0.001. ⁎⁎⁎ p b 0.0001. b
49 years, and Asian adults who were categorized as overweight or obese using Asian-specific BMI cutoffs were, respectively, two times and six times more likely to report being diagnosed with diabetes compared to their under-to-normal weight peers. Asian adults who drank alcohol were less likely to be diagnosed with diabetes compared to non-drinkers. Prevalence of diabetes among Asian subgroups Disaggregating the Asians revealed significant heterogeneity in diabetes prevalence among Asian ethnicities. The age-adjusted diabetes prevalence was highest among Filipinos (10.7%), South Asians (11.8%) and Other Asians (14.3%), and lowest among East Asian ethnicities (Chinese, Vietnamese, Korean, and Japanese), whose rates were comparable to that of whites (Table 3). A similar pattern was seen in mean BMI among Asian ethnicities. After adjusting for age, gender, and BMI, Filipino, South Asian, and Other Asian adults were found to be twice as likely as whites to report a diagnosis of diabetes. Discussion
Age-Adjusted Diabetes Prevalence (%, 95% CI)
In this analysis of data gathered between 1997 and 2011 from a racially and ethnically diverse sample of 47,282 Los Angeles residents, of whom 10% were Asian, we observed a rapid increase in the prevalence of diabetes among Asians compared to whites beginning in 2005. 16
12
* *
8
Asian White
4
0 1995
1997
1999
2001
2003
Year
2005
2007
2009
2011 *P<0.05
Fig. 1. Age-adjusted diabetes prevalence (±95% CI) among Asians and whites, Los Angeles County, 1997–2011.
Specifically, while diabetes rates for both Asians and whites hovered around 5% in 2002, the prevalence of diabetes for Asians increased to 7.1% in 2005, accelerating to 9.6% in 2011. In comparison, the prevalence of diabetes for whites increased minimally from 5.4% in 2002 to 6.7% in 2011. Our finding – that the prevalence of diabetes among Asians residing in Los Angeles County has been increasing rapidly – parallels reports of increasing rates of diabetes among Asians living in rapidly urbanizing countries such as China and India (Chan et al., 2009; Cockram, 2000; Yoon et al., 2006). Indeed, diabetes has emerged as a growing public health problem in many Asian countries. Changing lifestyles that have adversely affected diet and physical activity and led to increased obesity risk have been blamed (Ramachandran et al., 2012). Further, the epidemiologic literature has consistently reported that Asians living in their native countries and Asian immigrants living in western countries show a higher risk of diabetes and other obesity-related chronic health conditions at lower levels of BMI (Odegaard et al., 2009; Razak et al., 2005, 2007; Wang et al., 1994). Consistent with this finding, we observed that BMI, while increasing over the years in all racial/ethnic groups, was consistently lower among Asians compared to all other racial/ethnic groups, even when the prevalence of diabetes for Asians exceeded that of whites. There has been ongoing debate about the need for country- or ethnicity-specific BMI cut-points, particularly among Asian populations (Deurenberg et al., 2002; Low et al., 2009). BMI is an imperfect proxy for body fatness, and may not account adequately for variation in percent body fat and lean muscle mass, bone density, or fat distribution. Other measures of body fatness have been investigated, including bioimpedance, waist-to-hip ratio, waist circumference, and sagittal abdominal diameter. The sagittal abdominal diameter is a promising anthropometric proxy measure for predicting visceral adiposity, which has high relevance for Asian populations (Yim et al., 2010). Further research is needed to determine which of these, separately or in combination, is most predictive of cardiometabolic risk in multi-ethnic populations. The findings from our study suggest that the increasing trend in diabetes prevalence among Asians residing in highly urbanized Los Angeles County varies substantially among the various Asian subgroups. In particular, Filipinos and South Asians have considerably higher rates of diabetes (N10%) than East Asians and Vietnamese (b7%). This heterogeneity across ethnic subgroups has been noted in previous studies (Lam and LeRoith, 2012; Ma and Chan, 2013) and calls for the need for more comparative studies to examine genetic factors, gene–environment interactions, and factors during early development. Factors identified by our study as being associated with diabetes among all Asians after adjusting for covariates are older age (≥65 years), being overweight or obese,
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
being a non-drinker, and being insured. The inverse association between diabetes and heavy or binge drinking in our study was unexpected. Previous studies have suggested a J-shaped relationship between alcohol intake and diabetes risk, with mildly protective effects at low-tomoderate levels of consumption, and increased risk with heavy drinking and binge drinking among both men and women (Baliunas et al., 2009; Pietraszek et al., 2010). A potential explanation for our finding is that we were unable to identify lifetime abstainers and thus were unable to account for a potential sick quitter effect (i.e. former drinkers who quit drinking due to being diagnosed with an illness such as diabetes). This would tend to bias results toward overestimating the benefits of moderate consumption and underestimating the risks of heavy consumption. Additionally, since the number of Asians in the sample who were heavy drinkers was small, we were unable to distinguish respondents with high levels of consumption from those with very high consumption. This may have masked potential negative effects among drinkers with very high consumption. That increased diabetes risk was associated with having insurance may reflect the increased likelihood of being tested for diabetes (as well as other diseases) among those with health insurance (Barcellos et al., 2012). There are several limitations to the present study. First, the data source did not contain information on what type of diabetes a respondent
93
had, so we were unable to distinguish between Type 1 and Type 2 diabetes. Second, the data were cross-sectional and cannot be used to infer causation. Third, since the data were self-reported, there is the possibility of recall and social desirability bias, which may have led to misclassification of diabetes or underestimation of BMI (Merrill and Richardson, 2009). However, self-reported diabetes has been shown to have reasonable validity for public health surveillance (Margolis et al., 2008; Martin et al., 2000), and we found no evidence that Asians differentially report diabetes or BMI compared to other racial/ethnic groups. Diabetes is likely underreported, but Asians appear to have a similar rate of undiagnosed diabetes as whites (Thorpe et al., 2009). Fourth, the cooperation rate of the most recent LACHS in 2011 was 66% and reflects the ongoing challenges of conducting telephone-based surveys. However, the LACHS response rates have been comparable to those reported for other large state-based telephone health surveys (CDC, 2013), and the weighted demographic profile of the respondents, including the Asian subgroups, was similar to the population in Los Angeles County (unpublished results). Further, rigorous adjustments were made to address residual bias. Finally, while a strength of our study was the use of a large, unique dataset which allowed us to examine diabetes prevalence by Asian subgroups, the sample size was insufficient to further analyze risk factors for diabetes by Asian subgroup.
Table 2 Unadjusted diabetes prevalence, crude and adjusted odds of diabetes in Asians (N = 3234), Los Angeles County, 2002–2011. Characteristic
Age (years) 18–49 50–64 65+ Gender Male Female Household income 0–99% FPL 100–199% FPL 200% or above FPL Nativity US Born Foreign born in US 0–9 years Foreign born in US 10+ years Smoking status Non-smoker Current smoker Alcohol drinking Non-drinkers Low-mod drinkers Heavy or binge drinkers Insurance status Uninsured Insured Physical activityd Inactive Some PA Meets the guidelines BMI (Asian standard)e Under/normal weight Overweight Obesity BMI (CDC/WHO standard) Under/normal weight Overweight Obesity a
% (95% CI)
Crude OR
Adjusted OR: model 2a
Adjusted OR: model 3b
Asian
OR (95% CI)
OR (95% CI)
OR (95% CI)
2.4 (1.6, 3.3) 12.3 (9.8, 14.9) 22.4 (18.3, 26.5)
1.00 5.72 (3.75, 8.71) 11.77 (7.71, 17.98)
1.00 6.37 (4.10, 9.89) 15.15 (9.66, 23.78)
1.00 5.81 (3.71, 9.11) 12.63 (7.78, 20.50)
8.0 (6.5, 9.5) 7.3 (5.9, 8.7)
1.00 0.92 (0.69, 1.23)
1.00 1.04 (0.75, 1.45)
1.00 0.91 (0.64, 1.29)
10.1 (7.2, 13.0) 7.6 (5.3, 9.8) 7.0 (5.8, 8.3)
1.00 0.74 (0.47, 1.17) 0.67 (0.46, 0.98)
1.00 1.00 (0.56, 1.78) 1.14 (0.68, 1.91)
6.1 (4.3, 7.8) 2.7 (1.1, 4.2)c 10.2 (8.6, 11.8)
1.00 0.41 (0.21, 0.80) 1.76 (1.24, 2.51)
1.00 0.75 (0.33, 1.70) 1.46 (0.96, 2.22)
7.6 (6.6, 8.7) 6.7 (3.7, 9.7)
1.00 0.84 (0.51, 1.39)
1.00 1.54 (0.87, 2.72)
9.7 (8.2, 11.2) 4.7 (3.3, 6.2) 2.8 (0.8, 4.8)c
1.00 0.45 (0.31, 0.65) 0.26 (0.12, 0.55)
1.00 0.44 (0.29, 0.68) 0.38 (0.17, 0.85)
3.1 (1.7, 4.6)c 8.6 (7.4, 9.8)
1.00 2.98 (1.80, 4.93)
1.00 1.97 (1.09, 3.58)
8.9 (7.2, 10.6) 7.3 (4.6, 10.1) 6.4 (5.0, 7.7)
1.00 0.81 (0.51, 1.29) 0.70 (0.51, 0.96)
1.00 0.98 (0.57, 1.66) 0.83 (0.58, 1.20)
3.9 (2.8, 5.0) 8.0 (6.2, 9.7) 18.2 (14.0, 22.4)
1.00 2.12 (1.45, 3.11) 5.49 (3.66, 8.25)
5.1 (4.0, 6.1) 9.8 (7.6, 12.1) 22.7 (15.6, 29.8)
1.00 2.00 (1.43, 2.81) 5.43 (3.45, 8.56)
1.00 2.09 (1.42, 3.09) 6.43 (4.14, 9.98)
1.00 1.93 (1.29, 2.89) 6.43 (4.07, 10.17)
Model 2: Adjusted for age (categorical), gender, and BMI (Asian standard). Model 3 (full model): Adjusted for age (categorical), gender, BMI (Asian standard), Federal Poverty Level, nativity and years living in U.S., smoking status, alcohol drinking status, insurance status, and physical activity. c Estimate has a relative standard error ≥ 23%. d To meet physical activity guidelines at least one of the following criteria must be fulfilled: 1) vigorous activity for 20+ min, N3 days/week, 2) moderate activity for 30+ min, N5 days/ week, and 3) a combination of vigorous and moderate activity meeting the time criteria for N5 days/week. e Asian BMI standard: Overweight: 23 ≤ BMI b 27.5 kg/m2; obesity: BMI ≥ 27.5 kg/m2. b
94
M. Shih et al. / Preventive Medicine 63 (2014) 90–95
Table 3 Age-adjusted mean BMI, diabetes prevalence, and adjusted odds of diabetes among Asian subgroups, Los Angeles County, 1997–2011. Race/ethnicity
n
Mean BMI (95% CI)
Diabetes prevalence % (95% CI)
Adjusted ORb (95% CI)
White Asian Chinese Vietnamese Korean Japanese Filipino South Asianc Other Asiand
20,027 4672 1693 313 786 558 737 269 316
25.4 (25.3, 25.5) 23.7 (23.5, 23.8) 23.0 (22.8, 23.3) 22.9 (22.3, 23.5) 23.4 (23.1, 23.7) 23.9 (23.6, 24.3) 24.5 (24.2, 24.9) 25.0 (24.5, 25.5) 25.1 (24.5, 25.8)
5.7 (5.3, 6.0) 7.3 (6.4, 8.1) 5.2 (4.1, 6.4) 5.9 (3.0, 8.8)a 6.3 (4.6, 7.9) 6.7 (4.8, 8.7) 10.7 (7.9, 13.4) 11.8 (6.5, 17.1) 14.3 (8.8, 19.8)
1.00 – 1.03 (0.79, 1.34) 1.17 (0.65, 2.11) 1.39 (0.98, 1.97) 1.37 (0.96, 1.94) 1.98 (1.43, 2.72) 1.96 (1.11, 3.49) 2.10 (1.20, 3.70)
a
Estimate has a relative standard error ≥23%. Model included Asian subgroups, age (categorical), gender, and BMI (Asian standard: under-to-normal weight: BMI b 23.0 kg/m2; overweight: 23.0 ≤ BMI b 27.5 kg/m2; obese: BMI ≥ 27.5 kg/m2), using whites as the reference group. c Includes Asian Indian, Sri Lankan, Pakistani, and Bangladeshi. d Includes Asians not listed in the other subgroups, including southeast Asians (i.e. persons from Cambodia, Thailand, Laos, etc.) and Asians who did not specify any ethnicity. b
Despite these limitations, this is one of the first studies to report the rapid increase in diabetes among Asians using a large population-based sample, and to report differences in diabetes rates among Asian ethnic subgroups. Los Angeles County is home to the largest and one of the most diverse Asian populations in the U.S., allowing us to report disaggregated results for Asians. We were able to identify only one other study which reported differences in diabetes rates among Asian subgroups in the U.S. using a large population-based sample (Lee et al., 2011). Our study confirmed Lee's finding that Asian Americans appear to be at higher risk of diabetes compared to whites, despite having lower BMI. We were able to extend these findings by examining other Asian subgroups, and showing that diabetes is increasing more rapidly among Asians compared to whites. We conclude that diabetes among Asian Americans is a serious public health problem that has not been adequately recognized. There are substantial disparities among Asian subgroups that are masked by reporting data in aggregated form, and the current international BMIcutoffs for defining obesity may miss a significant percentage of diabetic Asians when used for targeted screening. More research is needed to identify which screening methods are most effective for ethnic populations and to ascertain the ethnicity specific needs of the various Asian subgroups. Such information is necessary for the development of culturally appropriate intervention programs to prevent and control diabetes in a population that is experiencing rapidly increased rates of diabetes. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ypmed.2014.03.016. Conflict of interest statement The authors declare that there are no conflicts of interest.
Acknowledgments The authors wish to thank Dr. Yan Cui for her assistance with the data analysis. References Baliunas, D.O., Taylor, B.J., Irving, H., Roerecke, M., Patra, J., Mohapatra, S., Rehm, J., 2009. Alcohol as a risk factor for type 2 diabetes: a systematic review and meta-analysis. Diabetes Care 32, 2123–2132. Barcellos, S.H., Goldman, D.P., Smith, J.P., 2012. Undiagnosed disease, especially diabetes, casts doubt on some of reported health ‘advantage’ of recent Mexican immigrants. Health Aff. (Millwood) 31, 2727–2737. Centers for Disease Control and Prevention, 2012. Methodologic changes in the Behavioral Risk Factor Surveillance System in 2011 and potential effects on prevalence estimates. MMWR 61, 410–413. Centers for Disease Control and Prevention, 2013. Behavioral Risk Factor Surveillance System 2012 Summary Data Quality Report. Accessed January 16, 2014. Available from http://www.cdc.gov/brfss/annual_data/2012/pdf/SummaryDataQualityReport2012_ 20130712.pdf.
Chan, J.C., Malik, V., Jia, W., Kadowaki, T., Yajnik, C.S., Yoon, K.H., Hu, F.B., 2009. Diabetes in Asia: epidemiology, risk factors, and pathophysiology. JAMA 301, 2129–2140. Cockram, C.S., 2000. The epidemiology of diabetes mellitus in the Asia-Pacific region. Hong Kong Med. J. 6, 43–52. Danaei, G., Finucane, M.M., Lu, Y., Singh, G.M., Cowan, M.J., Paciorek, C.J., Lin, J.K., Farzadfar, F., Khang, Y.H., et al., Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Blood Glucose), 2011. National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants. Lancet 378, 31–40. Deurenberg, P., Deurenberg-Yap, M., Guricci, S., 2002. Asians are different from Caucasians and from each other in their body mass index/body fat per cent relationship. Obes. Rev. 3, 141–146. Deurenberg-Yap, M., Chew, S.K., Deurenberg, P., 2002. Elevated body fat percentage and cardiovascular risks at low body mass index levels among Singaporean Chinese, Malays, and Indians in Singapore. Obes. Rev. 3, 209–215. He, M., Tan, K.C., Li, E.T., Kung, A.W., 2001. Body fat determination by dual energy X-ray absorptiometry and its relation to body mass index and waist circumference in Hong Kong Chinese. Int. J. Obes. Relat. Metab. Disord. 25, 748–752. International Diabetes Federation, 2011. IDF Diabetes Atlas, 5th ed. (Brussels, Belgium. Accessed July 3, 2013: Available from: http://www.idf.org/diabetesatlas). Lam, D.W., LeRoith, D., 2012. The worldwide diabetes epidemic. Curr. Opin. Endocrinol. Diabetes Obes. 19, 93–96. Lee, J.W., Brancati, F.L., Yeh, H.C., 2011. Trends in the prevalence of type 2 diabetes in Asians versus whites: results from the United States National Health Interview Survey, 1997–2008. Diabetes Care 34, 353–357. 2011 Los Angeles County Health Survey (LACHS): survey methodology report. Accessed July 3, 2013. Available from: http://publichealth.lacounty.gov/ha/docs/2011% 20LACHS/LACHS%20Methodological%20Report_REV121029.pdf. Low, S., Chin, M.C., Ma, S., Heng, D., Deurenberg-Yap, M., 2009. Rationale for redefining obesity in Asians. Ann. Acad. Med. Singapore 38, 66–69. Ma, R.C.W., Chan, J.C.N., 2013. Type 2 diabetes in East Asians: similarities and differences with populations in Europe and the United States. Ann. N. Y. Acad. Sci. 1281, 64–91. Margolis, K.L., Lihong, Q., Brzyski, R., Bonds, D.E., Howard, B.V., Kempainen, S., Simin, L., Robinson, J.G., Safford, M.M., et al., 2008. Validity of diabetes self-reports in the Women's Health Initiative: comparison with medication inventories and fasting glucose measurements. Clin. Trials 5, 240–247. Martin, L.M., Leff, M., Calonge, N., Garrett, C., Nelson, D.E., 2000. Validation of self-reported chronic conditions and health services in a managed care population. Am. J. Prev. Med. 18, 215–218. McNeely, M.J., Boyko, E.J., 2004. Type 2 diabetes prevalence in Asian Americans: results of a national health survey. Diabetes Care 27, 66–69. Merrill, R.M., Richardson, J.S., 2009. Validity of self-reported height, weight, and body mass index: findings from the National Health and Nutrition Examination Survey, 2001–2006. Prev. Chronic Dis. 6 (A121) (Epub 2009 Sep 15). Odegaard, A.O., Koh, W.P., Vazquez, G., Arakawa, K., Lee, H.P., Yu, M.C., Pereira, M.A., 2009. BMI and diabetes risk in Singaporean Chinese. Diabetes Care 32, 1104–1106. Pietraszek, A., Gregersen, S., Hermansen, K., 2010. Alcohol and type 2 diabetes. A review. Nutr. Metab. Cardiovasc. Dis. 20, 366–375. Ramachandran, A., Chamukuttan, S., Shetty, S.A., Arun, N., Susairaj, P., 2012. Obesity in Asia—is it different from rest of the world. Diabetes Metab. Res. Rev. 28 (Suppl. 2), 47–51. Razak, F., Anand, S.S., Vuksan, V., Davis, B., Jacobs, R., Teo, K.K., Yusuf, S., Investigators, S.H. A.R.E., 2005. Ethnic differences in the relationships between obesity and glucosemetabolic abnormalities: a cross-sectional population-based study. Int. J. Obes. (Lond.) 29, 656–667. Razak, F., Anand, S.S., Shannon, H., Vuksan, V., Davis, B., Jacobs, R., Teo, K.K., McQueen, M., Yusuf, S., 2007. Defining obesity cut points in a multiethnic population. Circulation 115, 2111–2118. Simon, P.A., Wold, C.M., Cousineau, M.R., Fielding, J.E., 2001. Meeting the data needs of a local health department: the Los Angeles County Health Survey. Am. J. Publ. Health 91, 1950–1952.
M. Shih et al. / Preventive Medicine 63 (2014) 90–95 Staimez, L.R., Weber, M.B., Narayan, K.M., Oza-Frank, R., 2013. A systematic review of overweight, obesity, and type 2 diabetes among Asian American subgroups. Curr. Diabetes Rev. 9, 312–331. Thorpe, L.E., Upadhyay, U.D., Chamany, S., Garg, R., Mandel-Ricci, J., Kellerman, S., Berger, D.K., Frieden, T.R., Gwynn, C., 2009. Prevalence and control of diabetes and impaired fasting glucose in New York City. Diabetes Care 32, 57–62. U.S. Department of Health and Human Services, 2000. Healthy People 2010, Understanding and improving health and objectives for improving health2nd ed. United States Government Printing Office, Washington DC. United States Census Bureau, U.S. Department of Commerce, Economics and Statistics Administration, 2012. The Asian Population: 2010. 2010 Census Briefs. March. Accessed July 1, 2013: Available from: http://www.census.gov/prod/cen2010/briefs/c2010br11.pdf. Wang, J., Thornton, J.C., Russell, M., Burastero, S., Heymsfield, S., Pierson Jr., R.N., 1994. Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am. J. Clin. Nutr. 60, 23–28.
95
WHO Expert Consultation, 2004. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 363, 157–163. Erratum in: Lancet 363, 902. World Health Organization, 1995. Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Consultation. WHO Technical Report Series Number 854. World Health Organization, Geneva. World Health Organization, 2000. WHO: the Asia-Pacific perspective: redefining obesity and its treatment. Western Pacific Regional Office. World Health Organization, Geneva. Yim, J.Y., Kim, D., Lim, S.H., Park, M.J., Choi, S.H., Lee, C.H., Kim, S.S., Cho, S.H., 2010. Sagittal abdominal diameter is a strong anthropometric measure of visceral adipose tissue in the Asian general population. Diabetes Care 33, 2665–2675. Yoon, K.H., Lee, J.H., Kim, J.W., Cho, J.H., Choi, Y.H., Ko, S.H., Zimmet, P., Son, H.Y., 2006. Epidemic obesity and type 2 diabetes in Asia. Lancet 368, 1681–1688.