Obesity Research & Clinical Practice (2007) 1, 267—272
ORIGINAL ARTICLE
Australian residence and heart disease risk in Vietnamese-born women Kaye E. Brock ∗, Margaret A. Bermingham, Dieu Tran, Aline Nguyen, Hoang Tran Dinh Faculty of Health Sciences, Lidcombe Campus, The University of Sydney, Australia Received 19 March 2007 ; received in revised form 13 May 2007; accepted 6 October 2007
KEYWORDS Vietnamese; Obesity; Body mass index; Body composition; Australia; Heart disease
Summary The aim of this study was to compare the disease risk profile of Vietnamese women who have lived in Australia for 2—15 years with a newly arrived group of Vietnamese women. The design was a comparison of two cross-sectional surveys (n = 256); one newly arrived (n = 159) and one (n = 97) who had lived in Australia for 2—15 years. The main outcome measures were body mass index (BMI), waist to hip ratio (WHR), total cholesterol (TC), high density lipoprotein cholesterol (HDL-C) and TC/HDL ratio (atherogenic index). The longer-stay residents had similar BMI (21.5 ± 3.5 kg/m2 vs. 21.1 ± 3.1 kg/m2 , p = 0.2); lower waist (69.3 ± 7.5 cm vs. 71.4 ± 7.6 cm, p = 0.8), WHR (0.76 ± 0.06 vs. 0.80 ± 0.06, p = 0.0001), TC (4.7 ± 1.0 mmol/L vs. 4.9 ± 0.9 mmol/L, p = 0.001), TC/HDL (3.0 ± 2.0 vs. 4.7 ± 2.0, p = 0.006) and higher hip measurement (91.1 ± 7.4 cm vs. 89.1 ± 5.6 cm, p = 0.009) than newly arrived Vietnamese women. After adjustment for BMI and age the odds of having a higher waist and WHR was significantly less for longer-stay residents, while the odds of having larger hips was doubled. The odds of having a high atherogenic index as estimated by the TC/HDL ratio was halved for the longer-stay residents (p = 0.15). We conclude that Vietnamese women we surveyed who have lived in Australia for 2—15 years have the same BMI, but lower levels of abdominal obesity and lower atherogenic index than newly arrived Vietnamese women surveyed. © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
Introduction Australia is a multiethnic country of rapidly changing demography. It is estimated that people of Asian ∗ Corresponding author. Tel.: +61 2 9351 9124; fax: +61 2 9351 9540. E-mail address: [email protected] (K.E. Brock).
origin will comprise up to 16% of Australia’s population by 2025 [1]; about 40% of this group will be Vietnamese-born. Descriptive migration studies have shown that migrants often change their diet and lifestyle to that of the host country [2]. While there are positive effects of migration from third world countries, such as improvement in life expectancy, migrants often subsequently develop
1871-403X/$ — see front matter © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2007.10.003
268 the chronic disease patterns of the new country [3]. Unfavourable changes in diet and lifestyle and lack of screening have been documented among Vietnamese migrants in the US [4]. We have already reported that Vietnamese who had lived for a time in Australia had started to eat more take away foods compared with new arrivals [5]. Increase in take away foods has been linked to obesity. This factor, together with potential exposure to an abundance of energy dense food, has the potential to lead to weight gain in any immigrant population, especially if exercise is low [2]. Obesity and overweight are defined by body mass index (BMI, weight in kg/height in m2 ) of greater than 30 and 25 kg/m2 , respectively. Generalised obesity is linked to risk of coronary heart disease (CHD), and, if the body fat is predominantly centrally distributed (abdominal obesity). This is a particularly potent risk factor for diabetes and CHD [6]. In a previous study [7] we found that Vietnamese females who were newly arrived in Australia (average time from arrival, 2 weeks), were of low BMI compared with the Australian national average, but had a high level of abdominal obesity as measured by waist to hip ratio (WHR), and a high prevalence of WHR above a recommended risk threshold [8]. These Vietnamese women also had low mean levels of high density lipoprotein cholesterol (HDL-C), the fraction of blood cholesterol which is anti-atherogenic, in comparison with an Australian national sample. A study of gestational diabetes in Australia found that women of Vietnamese ethnicity were among the highest risk group [9]. If Vietnamese women are at risk when they arrive first in Australia and are still of low BMI, the question arises, does this risk change with time of residence in Australia? As data on detailed blood lipid levels along with anthropometric measures in Vietnamese in both Australia and USA is very sparse we took the opportunity to compare these measures in Vietnamese women who have lived in Australia for up to 15 years and compare their profiles to a newly arrived group of Vietnamese women on whom we have already reported [7].
K.E. Brock et al.
Methods
through sequential sampling of patients at the surgery of a general practitioner with a relatively high proportion of Vietnamese patients (n = 57). Since separate analyses comparing these groups with new arrivals yielded similar results, the two groups were combined giving a total of 97 Vietnamese women aged 24—66 years who had been resident for at least 2 years in Australia (called longer-stay residents). These were compared with 159 Vietnamese women; this population was recruited via the auspices of the Refugee Health Screening Program, which all newly arriving immigrants are invited to attend soon after their arrival in Australia (usually within 2 weeks). A Vietnamese-speaking researcher attended screening services in Sydney from 1992 to 1994 and invited all Vietnamese subjects over age 18 to participate in the study [7] (newly arrived women). A Vietnamese-speaking researcher invited subjects to participate in the study, which had the approval of the Human Ethics Committee of the University of Sydney. Subjects had the study explained to them individually in Vietnamese and their voluntary participation was made clear. They were also given a subject information sheet in Vietnamese, the translation of which had been approved by an official Department of Health translator. All documents which were translated into Vietnamese and given to subjects were back translated by our Vietnamese assistant. All subjects signed informed consent before being accepted into the study. Subjects’ date of birth was recorded and their Vietnamese ethnicity confirmed by one of the authors (HTD) who is familiar with Vietnamese language and culture (Vietnamese people of Chinese origin were not included in these studies in order to exclude confounding by genetics and occupation). A questionnaire was administered which sought basic information on smoking and alcohol intake, time spent in vigorous (sports and heavy housework) and non vigorous (walking and light housework) physical activity, some dietary practices such as dairy intake and addition of salt to food, years of living in Australia and duration of residence in refugee camps before coming to Australia. Any subjects had clinical events associated with cardiovascular disease were excluded from the study.
Subjects
Measurements
Two groups of Vietnamese women who had lived in Australia for between 2 and 15 years were recruited. One group was recruited from friends and relatives of a researcher of Vietnamese origin and background (DT) by snowball community sampling (n = 40). The second group was accessed
Anthropometric measurements were taken with the subjects dressed in light indoor clothing with shoes removed. Weight was measured to the nearest 0.5 kg using a calibrated bathroom scales. Height was measured to the nearest 0.5 cm using a wall-mounted tape; BMI, weight in kg/height
Vietnamese women in Australia
269
in m2 , was calculated. The prevalence of chronic energy deficiency (CED) was determined on the BMI classification proposed by Ferro-Luzzi et al. [10]. BMI > 18.5 kg/m2 is classified as normal, 17—18.4 kg/m2 as grade 1, 16—16.9 kg/m2 as grade 2 and less than 16 kg/m2 as grade 3, CED. Waist and hip circumferences were measured using a flexible tape measure at the smallest circumference between the ribs and the iliac crest and the largest circumference in the buttock/gluteal area, respectively, and WHR calculated. All measurements were taken by a Vietnamese-speaking trained investigator.
Lipid Standardisation Programme. Low density lipoprotein cholesterol (LDL-C) concentration was estimated using a modification of the Friedewald formula [11]. A questionnaire adapted from the Australian National Heart Foundation Risk Factor Survey [12], and translated into Vietnamese was administered to each subject. This questionnaire addressed the areas of: past medical history; participation in regular physical exercise (measured in hours per 2-week period); history and duration of smoking, and dietary habits such as dairy intake.
Biochemistry
Data analyses were carried out using SPSS, Minitab 8.21 and Statview 5 software packages for Macintosh. The distribution of continuous variables was checked and, if skewed, a log transformation was used before comparing groups by Student’s ttest. Adjustment for confounding on mean levels between groups was achieved by using the beta coefficients of the covariates in a regression equation. Odds ratios were calculated and adjusted for confounding factors (age and BMI) using logistic regression techniques in SPSS [13]. Before the study commenced we calculated, based on pilot data that our sample size of 159 versus 57 and 40, respectively was more than sufficient to detect a difference of 0.1 of a blood profile or WHR unit or one BMI point (alpha 0.05, beta 0.8).
Fasting venous blood samples were collected from the patient group and finger-prick samples from the community group. All samples were centrifuged immediately and plasma was used to determine total cholesterol (TC), HDL-C and triglyceride (TG) concentrations on a Reflotron reflectance photometric analyser (Boehringer Mannheim Diagnostics, Germany) within 24 h. In the community group, sufficient blood was obtained only for TC determination. Coefficients of variation for repeated measurements were 2.9% for TC, 3.2% for HDL-C and 1.4% for TG. The accuracy of the Reflotron used in this study is regularly assessed through participation in the Australian College of Pathologist’s
Statistical analysis
Table 1 Demographic, behavioural, anthropometric and biochemical characteristics of Vietnamese women living in Australia
Age (years) Years resident in Australia Years in refugee camp BMI (kg/m2 ) WHR Waist (cm) Hips (cm) Height (cm) Vigorous exercisea Non-vigorous exercisea Cigarettes/day Drinks/week Dairy intake (L/week) ApoA1/B TC (mmol/L) Age adjusted HDL-C (mmol/L) TC/HDL Age adjusted a b
Hours/2 weeks. n = 10.
Newly arrived residents (n = 159)
Longer-stay residents (n = 97)
Significance (p value)
35.5 ± 6.5 0.3 ± 0.2 3.8 ± 1.0 21.1 ± 3.1 0.80 ± 0.06 71.4 ± 7.6 89.1 ± 5.6 154 ± 4.4 0.56 ± 1.5 4.7 ± 5.8 0.07 ± 0.8 0.01 ± 1.4 0.67 ± 1.35 1.4 ± 0.55 4.91 ± 0.86 4.80 1.17 ± 0.33 4.58 ± 2.0 4.68
33.5 ± 11 11.0 ± 3.7 0.5 ± 0.7 21.5 ± 3.5 0.76 ± 0.06 69.3 ± 7.5 91.1 ± 7.4 155 ± 5.1 1.7 ± 3.7 4.7 ± 5.8 0.17 ± 0.9 0.10 ± 0.7 0.94 ± 1.35 1.8 ± 0.43b 4.67 ± 1.03 4.69 1.74 ± 0.57 3.01 ± 2.0 2.93
0.0001 0.0001 0.0001 0.20 0.0001 0.83 0.009 0.23 0.006 0.96 0.106 0.24 0.039 0.382 0.001 0.003 0.0001 0.006 0.001
270
K.E. Brock et al.
Results The response rate for refugees was 85%, and over 70% for the longer-stay patient and community samples. As can be seen from Table 1 these two groups of Vietnamese women differ not only in length of time resident in Australia but also in their internment in refugee camps (almost exclusively outside Australia), in addition there was a slight age difference, the longer-stay group being on average younger than the refugee group. The proportion of life spent in Australia was calculated for both groups. Most refugees (>95%) had lived in camps in Asia before coming to Australia where the average period of stay was 3.8 ± 1.1 years. Almost all were examined within 2 weeks of their arrival, hence, the mean proportion of life spent in Australia was very small, average, 0.04 ± 0.04%, range 0.001—0.27%. The longer-stay residents had a much longer mean proportion of life in Australia, 19.5 ± 11.8%, range 1.9—54%. The mean BMI (21.5 ± 3.5 kg/m2 ) of longer-stay residents was low by Western standards but similar to that of the newly arrived women and slightly above that reported for Vietnamese women in their home country (19.7 kg/m2 for women in the 20—40-year-old age group [7]). However, the proportion of women with any classification of CED (BMI < 18.4 kg/m2 ) was 15% for the longerstay residents compared with 20% for the newly arrived women. The number of women in the most severe category of CED (BMI < 16 kg/m2 ) was 2% for the longer-stay residents compared with 3% for the newly arrived women. Only one case of self reported diabetes was recorded in each group,
and similarly smoking rates were very low in both groups, but appeared to be marginally increasing, which is a known effect of acculturation. A clear difference in the levels of vigorous exercise and dairy intake was evident between longer-stay residents and newly arrived women. There were also differences in hip measurements, which were reflected in WHR; however, these differences in body fat distribution were not reflected in a difference in BMI. Total cholesterol was significantly lower and HDL-C significantly higher, in the longerstay residents, and these data were reflected in a significantly lower TC/HDL-C ratio in the longerstay residents. These differences remained on adjustment for age (as cholesterol levels are known to increase with age in Europeans). The odds of attaining an ‘‘at risk’’ profile for the longer-stay residents compared with the newly arrived women was examined by determining the odds ratios of their having larger waist, WHR, hips, and an adverse blood lipid profile (Table 2). These odds were adjusted for age and BMI where appropriate. The risk of having a higher waist and WHR was significantly less in longer-stay residents, while the risk of having larger hips was doubled. The risk of having adverse TC (above the desirable limit of 5.2 mmol/L) was not altered, but, although not significant, it is interesting that the risk of not having adverse HDL-C (less then 1.0 mmol/L) increased OR = 2 (95% CI = 0.8—4.9). Similarly, the risk of having a high atherogenic index as estimated by the TC/HDL-C ratio was halved for the longer-stay residents (p = 0.04). These residents were also three times more likely to spend more than 1 h per week in vigorous exercise (p = 0.01) and have a non signif-
Table 2 Risk estimates of anthropometric, blood lipid and lifestyle profiles of longer-stay Vietnamese women compared to newly arrived Vietnamese women living in Australia Risk indexa
Odds ratio (95% CI) longer-stay residents vs. newly arrived women
Odds ratio (95% CI) adjusted for age and BMI
BMI > 21 kg/m2 (mean) Waist > 71 cm (median) Hips > 92 cm (mean) WHR > 0.80 (median) TC > 5.2 mmol/L HDL > 1.0 mmol/L HDL > 1.2 mmol/L TC/HDL > 5.0 Vigorous exercise >1 h/week Dairy intake > I L/week
1.1 0.6 2.1 0.3 0.9 1.9 4.2 0.5 2.9 1.8
0.4 2.6 0.2 0.9 2.0 4.2 0.4 2.9 1.7
(0.6—1.8) (0.3—1.0) (1.2—3.7) (0.15—0.52) (0.5—1.5) (0.8—4.6) (1.8—9.9) (0.2—1.3) (1.4—5.9) (0.9—3.5)
(0.2—0.8) (1.3—5.5) (0.1—0.5) (0.5—1.6) (0.8—4.9) (1.7—9.9) (0.1—0.9) (1.3—6.2) (0.8—3.7)
p value 0.04 0.01 0.02 0.00 0.86 0.15 0.00 0.04 0.01 0.16
Abbreviations: BMI, body mass index; WHR, waist to hip ratio; TC, total cholesterol (blood); HDL, high density lipoprotein (see Methods for definitions). a The risk was calculated as the odds of the longer-stay residents having this attribute compared to newly arrived women (baseline). For anthropometry, the cut-off values are the mean or median of all women in both samples; for TC and HDL-C, the cut-off risk levels are those recommended by the National Cholesterol Education Campaign.
Vietnamese women in Australia icant trend towards consuming more dairy products per week OR = 1.7 (CI = 0.8—3.7).
Discussion Before any conclusions are drawn from this crosssectional comparison it must be noted that the two groups compared have come by different routes to Australia and this may be a confounding factor for these data. Almost all of the refugees had spent long periods of time in camps in Asia and entered Australia legally between 1992 and 1994, they were among the last large groups of Vietnamese to come to Australia. The longer-stay group had arrived in Australia somewhat earlier, mostly in the decade 1980—1990. Some had spent time in refugee camps, but this was generally of shorter duration and they may have been boat people who entered Australia illegally. This may limit the extent to which the results can be generalised to the Australian Vietnamese population as a whole. In addition the different sampling methods used will probably have introduced selection bias. The recruitment by snowball sampling of the relatives and friendship group of a young Vietnamese researcher (n = 40) could be the reason for the longer-stay women being younger and healthier. However the GP sample (n = 57) did had a wider age range. These samples were recruited after various attempts at more random approaches such as telephone book and electoral role access failed due to the politically sensitive nature of this immigrant population. A definitive conclusion as to the progression of risk among Vietnamese women can only come from a cohort study which follows the same group of women over time. With these limitations in mind the results of this present comparison show that the heart risk profile of longer-stay Vietnamese women is not worse than that of a newly arrived group in regard to both measures of body fatness and blood lipids. A tendency toward a higher WHR or waist girth is less common in the longer-stay residents and there is no difference in age adjusted BMI between the two groups. The longer-stay group are also more likely to engage in exercise and have a trend towards consumption of dairy products, which from other studies of Vietnamese we have found to be indicators of acculturation [5,3]. In this study, a tendency toward abdominal obesity was estimated from anthropometric measurements such as waist or WHR, rather than using direct methods such as computed tomography of visceral fat volume. As in most field studies, this was a necessary limitation due to time and expense.
271 Both waist and WHR have been used extensively to predict visceral fat mass [14]. In this study, WHR was significantly lower in the longer-stay women, while waist measurement was not different. While waist is often considered as a better correlate of visceral fat than WHR [14], WHR has been shown to be closely correlated with direct measurements of visceral fat in Asian women [15], and WHR is a good predictor of diabetes. In addition, WHR may not only reflect visceral fat mass; a large European study has shown that the combination of smaller hips and larger waist than expected for BMI, has been strongly associated with diabetes [16]. The authors have speculated that narrow hips are indicative of relatively small muscle leg mass, which in turn is also associated with peripheral insulin resistance. Larger hip circumference has been shown to independently predict health and longevity in a Swedish female cohort [17], and, in a separate study, length of residence in Sweden was inversely associated with central adiposity in immigrants [18]. Although these studies investigated other immigrant groups than Vietnamese their findings are thought provoking. Thus, the WHR for a given BMI may be very important for women of low BMI in regard to risk of insulin resistance and diabetes. The newly arrived Vietnamese women have relatively narrow hips and large waist for their still low BMI, while the women who have lived in Australia longer appear to have a different shape with smaller waist and larger hips for the same BMI. It is possible to speculate that this may reflect an improvement in insulin resistance, but this would have to be confirmed by direct biochemical measures of glucose tolerance. While elevated TC is strongly associated with increased risk of CHD, a more precise lipid indicator of CHD risk is an atherogenic index of the ratio of TC to HDL [8]. Low HDL-C is an important risk factor for CHD, especially in women, and abdominal obesity is an important correlate of dyslipidaemia [19]. Some interesting trends occur when the data was analysed in risk categories; although not statistically significant, the risk having a high TC/HDL-C ratio was halved (p = 0.049) for the women in the longer-stay group. The trend in more reported exercise (p = 0.001) and consumption of dairy produce (p = 0.16) may also a marker of transition to Australian lifestyle, as exercise for its own sake is not widely practised in Vietnam, and intake of dairy produce is limited. In conclusion despite the obvious limitations of such a comparison this comparison shows that a sample of Vietnamese women when they arrive in Australia have relatively high indicators of abdominal obesity for their low BMI and relatively low HDL, while other Vietnamese women who have lived in
272
K.E. Brock et al.
Australia for 2—15 years have the same BMI, lower WHR and a clinically better TC to HDL ratio, and thus do not appear to be at increased heart disease risk as a result of living in Australia. If Vietnamese women maintain their healthy weight range, continue to avoid smoking and increase exercise levels, Australian living may contribute to positive health behaviour and reduced burden of not only heart disease but also diabetes and even colon and breast cancer.
Acknowledgments This research was supported by grant monies from: • Cumberland College, University of Sydney Research Grants and • a NHMRC Public Health Research and Development Grant.
References [1] DIMA. Australia’s population trends and prospects 1996. Canberra: Australian Government; 1997. [2] Yano K, Maclean CJ, Reed DM, Shimizu Y, Sasaki H, Kodama K, et al. A comparison of the 12-year mortality and predictive factors of coronary heart disease among Japanese men in Japan and Hawaii. Am J Epidemiol 1998;127(3):476—87. [3] Rissel C. The development and application of a scale of acculturation. Aust N Z J Public Health 1997;21(6):606—13. [4] Todd R, Gelbier S. ‘Eat more food, get more health’— –attitudes and food habits of a group of Vietnamese refugees. Health Educ J 1988;47(4):149—53. [5] Brock K, Lockwood E, Cant R, Bermingham M, Tran DH. An investigation of health behaviour change in Vietnameseborn living in Sydney. Ethn Dis 2001;11(3):385—90. [6] Bjorntorp P. Portal adipose tissue as a generator of risk factors for cardiovascular disease and diabetes. Arteriosclerosis 1990;10(4):493—6.
[7] Bermingham M, Brock K, Nguyen D, Tran-Dinh H. Body mass index and body fat distribution in newly-arrived Vietnamese refugees in Sydney, Australia. Eur J Clin Nutr 1996;50(10):698—700. [8] Despres JP. Dyslipidemia and obesity. Baillieres Clin Endocrinol Metab 1994;8(3):629—60. [9] Doery J, Edis K, Healy D, Bishop S, Tippet C. Very high prevalence of gestational diabetes in Vietnamese and Cambodian women in Australia. Med J Aust 1989;151(2): 111. [10] Ferro-Luzzi A, Sette S, Franklin M, James W. A simple approach to assessing adult chronic energy deficiency. Eur J Clin Nutr 1992;46(3):173—86. [11] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18(6):499—502. [12] AIHW. Risk factor prevalence study: survey no. 3. Canberra: National Heart Foundation of Australia Committee: Australian Institute for Health and Welfare; 1989. [13] Kleinbaum D, Kupper L, Muller K. Applied regression analysis. Boston: Duxbury Press; 1988. [14] Despres JP, Prud’homme D, Pouliot MC, Tremblay A, Bouchard C. Estimation of deep abdominal adipose-tissue accumulation from simple anthropometric measurements in men. Am J Clin Nutr 1991;54(3):471—7. [15] Masuda T, Imai K, Komiya S. Relationship of anthropometric indices of body fat to cardiovascular risk in Japanese women. Ann Physiol Anthropol 1993;12(3):135— 44. [16] Seidell JC, Han TS, Feskens EJ, Lean ME. Narrow hips and broad waist circumferences independently contribute to increased risk of non-insulin-dependent diabetes mellitus. J Intern Med 1997;242(5):401—6. [17] Lissner L, Bjorkelund C, Heitmann BL, Seidell JC, Bengtsson C. Larger hip circumference independently predicts health and longevity in a Swedish female cohort. Obes Res 2001;9(10):644—6. [18] Lahmann PH, Lissner L, Gullberg B, Berglund G. Differences in body fat and central adiposity between Swedes and European immigrants: the Malmo Diet and Cancer Study. Obes Res 2000;8(9):620—31. [19] Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol 1998;81(4A):7B—12B.
Available online at www.sciencedirect.com
ORIGINAL ARTICLE
Australian residence and heart disease risk in Vietnamese-born women Kaye E. Brock ∗, Margaret A. Bermingham, Dieu Tran, Aline Nguyen, Hoang Tran Dinh Faculty of Health Sciences, Lidcombe Campus, The University of Sydney, Australia Received 19 March 2007 ; received in revised form 13 May 2007; accepted 6 October 2007
KEYWORDS Vietnamese; Obesity; Body mass index; Body composition; Australia; Heart disease
Summary The aim of this study was to compare the disease risk profile of Vietnamese women who have lived in Australia for 2—15 years with a newly arrived group of Vietnamese women. The design was a comparison of two cross-sectional surveys (n = 256); one newly arrived (n = 159) and one (n = 97) who had lived in Australia for 2—15 years. The main outcome measures were body mass index (BMI), waist to hip ratio (WHR), total cholesterol (TC), high density lipoprotein cholesterol (HDL-C) and TC/HDL ratio (atherogenic index). The longer-stay residents had similar BMI (21.5 ± 3.5 kg/m2 vs. 21.1 ± 3.1 kg/m2 , p = 0.2); lower waist (69.3 ± 7.5 cm vs. 71.4 ± 7.6 cm, p = 0.8), WHR (0.76 ± 0.06 vs. 0.80 ± 0.06, p = 0.0001), TC (4.7 ± 1.0 mmol/L vs. 4.9 ± 0.9 mmol/L, p = 0.001), TC/HDL (3.0 ± 2.0 vs. 4.7 ± 2.0, p = 0.006) and higher hip measurement (91.1 ± 7.4 cm vs. 89.1 ± 5.6 cm, p = 0.009) than newly arrived Vietnamese women. After adjustment for BMI and age the odds of having a higher waist and WHR was significantly less for longer-stay residents, while the odds of having larger hips was doubled. The odds of having a high atherogenic index as estimated by the TC/HDL ratio was halved for the longer-stay residents (p = 0.15). We conclude that Vietnamese women we surveyed who have lived in Australia for 2—15 years have the same BMI, but lower levels of abdominal obesity and lower atherogenic index than newly arrived Vietnamese women surveyed. © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
Introduction Australia is a multiethnic country of rapidly changing demography. It is estimated that people of Asian ∗ Corresponding author. Tel.: +61 2 9351 9124; fax: +61 2 9351 9540. E-mail address: [email protected] (K.E. Brock).
origin will comprise up to 16% of Australia’s population by 2025 [1]; about 40% of this group will be Vietnamese-born. Descriptive migration studies have shown that migrants often change their diet and lifestyle to that of the host country [2]. While there are positive effects of migration from third world countries, such as improvement in life expectancy, migrants often subsequently develop
1871-403X/$ — see front matter © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2007.10.003
268 the chronic disease patterns of the new country [3]. Unfavourable changes in diet and lifestyle and lack of screening have been documented among Vietnamese migrants in the US [4]. We have already reported that Vietnamese who had lived for a time in Australia had started to eat more take away foods compared with new arrivals [5]. Increase in take away foods has been linked to obesity. This factor, together with potential exposure to an abundance of energy dense food, has the potential to lead to weight gain in any immigrant population, especially if exercise is low [2]. Obesity and overweight are defined by body mass index (BMI, weight in kg/height in m2 ) of greater than 30 and 25 kg/m2 , respectively. Generalised obesity is linked to risk of coronary heart disease (CHD), and, if the body fat is predominantly centrally distributed (abdominal obesity). This is a particularly potent risk factor for diabetes and CHD [6]. In a previous study [7] we found that Vietnamese females who were newly arrived in Australia (average time from arrival, 2 weeks), were of low BMI compared with the Australian national average, but had a high level of abdominal obesity as measured by waist to hip ratio (WHR), and a high prevalence of WHR above a recommended risk threshold [8]. These Vietnamese women also had low mean levels of high density lipoprotein cholesterol (HDL-C), the fraction of blood cholesterol which is anti-atherogenic, in comparison with an Australian national sample. A study of gestational diabetes in Australia found that women of Vietnamese ethnicity were among the highest risk group [9]. If Vietnamese women are at risk when they arrive first in Australia and are still of low BMI, the question arises, does this risk change with time of residence in Australia? As data on detailed blood lipid levels along with anthropometric measures in Vietnamese in both Australia and USA is very sparse we took the opportunity to compare these measures in Vietnamese women who have lived in Australia for up to 15 years and compare their profiles to a newly arrived group of Vietnamese women on whom we have already reported [7].
K.E. Brock et al.
Methods
through sequential sampling of patients at the surgery of a general practitioner with a relatively high proportion of Vietnamese patients (n = 57). Since separate analyses comparing these groups with new arrivals yielded similar results, the two groups were combined giving a total of 97 Vietnamese women aged 24—66 years who had been resident for at least 2 years in Australia (called longer-stay residents). These were compared with 159 Vietnamese women; this population was recruited via the auspices of the Refugee Health Screening Program, which all newly arriving immigrants are invited to attend soon after their arrival in Australia (usually within 2 weeks). A Vietnamese-speaking researcher attended screening services in Sydney from 1992 to 1994 and invited all Vietnamese subjects over age 18 to participate in the study [7] (newly arrived women). A Vietnamese-speaking researcher invited subjects to participate in the study, which had the approval of the Human Ethics Committee of the University of Sydney. Subjects had the study explained to them individually in Vietnamese and their voluntary participation was made clear. They were also given a subject information sheet in Vietnamese, the translation of which had been approved by an official Department of Health translator. All documents which were translated into Vietnamese and given to subjects were back translated by our Vietnamese assistant. All subjects signed informed consent before being accepted into the study. Subjects’ date of birth was recorded and their Vietnamese ethnicity confirmed by one of the authors (HTD) who is familiar with Vietnamese language and culture (Vietnamese people of Chinese origin were not included in these studies in order to exclude confounding by genetics and occupation). A questionnaire was administered which sought basic information on smoking and alcohol intake, time spent in vigorous (sports and heavy housework) and non vigorous (walking and light housework) physical activity, some dietary practices such as dairy intake and addition of salt to food, years of living in Australia and duration of residence in refugee camps before coming to Australia. Any subjects had clinical events associated with cardiovascular disease were excluded from the study.
Subjects
Measurements
Two groups of Vietnamese women who had lived in Australia for between 2 and 15 years were recruited. One group was recruited from friends and relatives of a researcher of Vietnamese origin and background (DT) by snowball community sampling (n = 40). The second group was accessed
Anthropometric measurements were taken with the subjects dressed in light indoor clothing with shoes removed. Weight was measured to the nearest 0.5 kg using a calibrated bathroom scales. Height was measured to the nearest 0.5 cm using a wall-mounted tape; BMI, weight in kg/height
Vietnamese women in Australia
269
in m2 , was calculated. The prevalence of chronic energy deficiency (CED) was determined on the BMI classification proposed by Ferro-Luzzi et al. [10]. BMI > 18.5 kg/m2 is classified as normal, 17—18.4 kg/m2 as grade 1, 16—16.9 kg/m2 as grade 2 and less than 16 kg/m2 as grade 3, CED. Waist and hip circumferences were measured using a flexible tape measure at the smallest circumference between the ribs and the iliac crest and the largest circumference in the buttock/gluteal area, respectively, and WHR calculated. All measurements were taken by a Vietnamese-speaking trained investigator.
Lipid Standardisation Programme. Low density lipoprotein cholesterol (LDL-C) concentration was estimated using a modification of the Friedewald formula [11]. A questionnaire adapted from the Australian National Heart Foundation Risk Factor Survey [12], and translated into Vietnamese was administered to each subject. This questionnaire addressed the areas of: past medical history; participation in regular physical exercise (measured in hours per 2-week period); history and duration of smoking, and dietary habits such as dairy intake.
Biochemistry
Data analyses were carried out using SPSS, Minitab 8.21 and Statview 5 software packages for Macintosh. The distribution of continuous variables was checked and, if skewed, a log transformation was used before comparing groups by Student’s ttest. Adjustment for confounding on mean levels between groups was achieved by using the beta coefficients of the covariates in a regression equation. Odds ratios were calculated and adjusted for confounding factors (age and BMI) using logistic regression techniques in SPSS [13]. Before the study commenced we calculated, based on pilot data that our sample size of 159 versus 57 and 40, respectively was more than sufficient to detect a difference of 0.1 of a blood profile or WHR unit or one BMI point (alpha 0.05, beta 0.8).
Fasting venous blood samples were collected from the patient group and finger-prick samples from the community group. All samples were centrifuged immediately and plasma was used to determine total cholesterol (TC), HDL-C and triglyceride (TG) concentrations on a Reflotron reflectance photometric analyser (Boehringer Mannheim Diagnostics, Germany) within 24 h. In the community group, sufficient blood was obtained only for TC determination. Coefficients of variation for repeated measurements were 2.9% for TC, 3.2% for HDL-C and 1.4% for TG. The accuracy of the Reflotron used in this study is regularly assessed through participation in the Australian College of Pathologist’s
Statistical analysis
Table 1 Demographic, behavioural, anthropometric and biochemical characteristics of Vietnamese women living in Australia
Age (years) Years resident in Australia Years in refugee camp BMI (kg/m2 ) WHR Waist (cm) Hips (cm) Height (cm) Vigorous exercisea Non-vigorous exercisea Cigarettes/day Drinks/week Dairy intake (L/week) ApoA1/B TC (mmol/L) Age adjusted HDL-C (mmol/L) TC/HDL Age adjusted a b
Hours/2 weeks. n = 10.
Newly arrived residents (n = 159)
Longer-stay residents (n = 97)
Significance (p value)
35.5 ± 6.5 0.3 ± 0.2 3.8 ± 1.0 21.1 ± 3.1 0.80 ± 0.06 71.4 ± 7.6 89.1 ± 5.6 154 ± 4.4 0.56 ± 1.5 4.7 ± 5.8 0.07 ± 0.8 0.01 ± 1.4 0.67 ± 1.35 1.4 ± 0.55 4.91 ± 0.86 4.80 1.17 ± 0.33 4.58 ± 2.0 4.68
33.5 ± 11 11.0 ± 3.7 0.5 ± 0.7 21.5 ± 3.5 0.76 ± 0.06 69.3 ± 7.5 91.1 ± 7.4 155 ± 5.1 1.7 ± 3.7 4.7 ± 5.8 0.17 ± 0.9 0.10 ± 0.7 0.94 ± 1.35 1.8 ± 0.43b 4.67 ± 1.03 4.69 1.74 ± 0.57 3.01 ± 2.0 2.93
0.0001 0.0001 0.0001 0.20 0.0001 0.83 0.009 0.23 0.006 0.96 0.106 0.24 0.039 0.382 0.001 0.003 0.0001 0.006 0.001
270
K.E. Brock et al.
Results The response rate for refugees was 85%, and over 70% for the longer-stay patient and community samples. As can be seen from Table 1 these two groups of Vietnamese women differ not only in length of time resident in Australia but also in their internment in refugee camps (almost exclusively outside Australia), in addition there was a slight age difference, the longer-stay group being on average younger than the refugee group. The proportion of life spent in Australia was calculated for both groups. Most refugees (>95%) had lived in camps in Asia before coming to Australia where the average period of stay was 3.8 ± 1.1 years. Almost all were examined within 2 weeks of their arrival, hence, the mean proportion of life spent in Australia was very small, average, 0.04 ± 0.04%, range 0.001—0.27%. The longer-stay residents had a much longer mean proportion of life in Australia, 19.5 ± 11.8%, range 1.9—54%. The mean BMI (21.5 ± 3.5 kg/m2 ) of longer-stay residents was low by Western standards but similar to that of the newly arrived women and slightly above that reported for Vietnamese women in their home country (19.7 kg/m2 for women in the 20—40-year-old age group [7]). However, the proportion of women with any classification of CED (BMI < 18.4 kg/m2 ) was 15% for the longerstay residents compared with 20% for the newly arrived women. The number of women in the most severe category of CED (BMI < 16 kg/m2 ) was 2% for the longer-stay residents compared with 3% for the newly arrived women. Only one case of self reported diabetes was recorded in each group,
and similarly smoking rates were very low in both groups, but appeared to be marginally increasing, which is a known effect of acculturation. A clear difference in the levels of vigorous exercise and dairy intake was evident between longer-stay residents and newly arrived women. There were also differences in hip measurements, which were reflected in WHR; however, these differences in body fat distribution were not reflected in a difference in BMI. Total cholesterol was significantly lower and HDL-C significantly higher, in the longerstay residents, and these data were reflected in a significantly lower TC/HDL-C ratio in the longerstay residents. These differences remained on adjustment for age (as cholesterol levels are known to increase with age in Europeans). The odds of attaining an ‘‘at risk’’ profile for the longer-stay residents compared with the newly arrived women was examined by determining the odds ratios of their having larger waist, WHR, hips, and an adverse blood lipid profile (Table 2). These odds were adjusted for age and BMI where appropriate. The risk of having a higher waist and WHR was significantly less in longer-stay residents, while the risk of having larger hips was doubled. The risk of having adverse TC (above the desirable limit of 5.2 mmol/L) was not altered, but, although not significant, it is interesting that the risk of not having adverse HDL-C (less then 1.0 mmol/L) increased OR = 2 (95% CI = 0.8—4.9). Similarly, the risk of having a high atherogenic index as estimated by the TC/HDL-C ratio was halved for the longer-stay residents (p = 0.04). These residents were also three times more likely to spend more than 1 h per week in vigorous exercise (p = 0.01) and have a non signif-
Table 2 Risk estimates of anthropometric, blood lipid and lifestyle profiles of longer-stay Vietnamese women compared to newly arrived Vietnamese women living in Australia Risk indexa
Odds ratio (95% CI) longer-stay residents vs. newly arrived women
Odds ratio (95% CI) adjusted for age and BMI
BMI > 21 kg/m2 (mean) Waist > 71 cm (median) Hips > 92 cm (mean) WHR > 0.80 (median) TC > 5.2 mmol/L HDL > 1.0 mmol/L HDL > 1.2 mmol/L TC/HDL > 5.0 Vigorous exercise >1 h/week Dairy intake > I L/week
1.1 0.6 2.1 0.3 0.9 1.9 4.2 0.5 2.9 1.8
0.4 2.6 0.2 0.9 2.0 4.2 0.4 2.9 1.7
(0.6—1.8) (0.3—1.0) (1.2—3.7) (0.15—0.52) (0.5—1.5) (0.8—4.6) (1.8—9.9) (0.2—1.3) (1.4—5.9) (0.9—3.5)
(0.2—0.8) (1.3—5.5) (0.1—0.5) (0.5—1.6) (0.8—4.9) (1.7—9.9) (0.1—0.9) (1.3—6.2) (0.8—3.7)
p value 0.04 0.01 0.02 0.00 0.86 0.15 0.00 0.04 0.01 0.16
Abbreviations: BMI, body mass index; WHR, waist to hip ratio; TC, total cholesterol (blood); HDL, high density lipoprotein (see Methods for definitions). a The risk was calculated as the odds of the longer-stay residents having this attribute compared to newly arrived women (baseline). For anthropometry, the cut-off values are the mean or median of all women in both samples; for TC and HDL-C, the cut-off risk levels are those recommended by the National Cholesterol Education Campaign.
Vietnamese women in Australia icant trend towards consuming more dairy products per week OR = 1.7 (CI = 0.8—3.7).
Discussion Before any conclusions are drawn from this crosssectional comparison it must be noted that the two groups compared have come by different routes to Australia and this may be a confounding factor for these data. Almost all of the refugees had spent long periods of time in camps in Asia and entered Australia legally between 1992 and 1994, they were among the last large groups of Vietnamese to come to Australia. The longer-stay group had arrived in Australia somewhat earlier, mostly in the decade 1980—1990. Some had spent time in refugee camps, but this was generally of shorter duration and they may have been boat people who entered Australia illegally. This may limit the extent to which the results can be generalised to the Australian Vietnamese population as a whole. In addition the different sampling methods used will probably have introduced selection bias. The recruitment by snowball sampling of the relatives and friendship group of a young Vietnamese researcher (n = 40) could be the reason for the longer-stay women being younger and healthier. However the GP sample (n = 57) did had a wider age range. These samples were recruited after various attempts at more random approaches such as telephone book and electoral role access failed due to the politically sensitive nature of this immigrant population. A definitive conclusion as to the progression of risk among Vietnamese women can only come from a cohort study which follows the same group of women over time. With these limitations in mind the results of this present comparison show that the heart risk profile of longer-stay Vietnamese women is not worse than that of a newly arrived group in regard to both measures of body fatness and blood lipids. A tendency toward a higher WHR or waist girth is less common in the longer-stay residents and there is no difference in age adjusted BMI between the two groups. The longer-stay group are also more likely to engage in exercise and have a trend towards consumption of dairy products, which from other studies of Vietnamese we have found to be indicators of acculturation [5,3]. In this study, a tendency toward abdominal obesity was estimated from anthropometric measurements such as waist or WHR, rather than using direct methods such as computed tomography of visceral fat volume. As in most field studies, this was a necessary limitation due to time and expense.
271 Both waist and WHR have been used extensively to predict visceral fat mass [14]. In this study, WHR was significantly lower in the longer-stay women, while waist measurement was not different. While waist is often considered as a better correlate of visceral fat than WHR [14], WHR has been shown to be closely correlated with direct measurements of visceral fat in Asian women [15], and WHR is a good predictor of diabetes. In addition, WHR may not only reflect visceral fat mass; a large European study has shown that the combination of smaller hips and larger waist than expected for BMI, has been strongly associated with diabetes [16]. The authors have speculated that narrow hips are indicative of relatively small muscle leg mass, which in turn is also associated with peripheral insulin resistance. Larger hip circumference has been shown to independently predict health and longevity in a Swedish female cohort [17], and, in a separate study, length of residence in Sweden was inversely associated with central adiposity in immigrants [18]. Although these studies investigated other immigrant groups than Vietnamese their findings are thought provoking. Thus, the WHR for a given BMI may be very important for women of low BMI in regard to risk of insulin resistance and diabetes. The newly arrived Vietnamese women have relatively narrow hips and large waist for their still low BMI, while the women who have lived in Australia longer appear to have a different shape with smaller waist and larger hips for the same BMI. It is possible to speculate that this may reflect an improvement in insulin resistance, but this would have to be confirmed by direct biochemical measures of glucose tolerance. While elevated TC is strongly associated with increased risk of CHD, a more precise lipid indicator of CHD risk is an atherogenic index of the ratio of TC to HDL [8]. Low HDL-C is an important risk factor for CHD, especially in women, and abdominal obesity is an important correlate of dyslipidaemia [19]. Some interesting trends occur when the data was analysed in risk categories; although not statistically significant, the risk having a high TC/HDL-C ratio was halved (p = 0.049) for the women in the longer-stay group. The trend in more reported exercise (p = 0.001) and consumption of dairy produce (p = 0.16) may also a marker of transition to Australian lifestyle, as exercise for its own sake is not widely practised in Vietnam, and intake of dairy produce is limited. In conclusion despite the obvious limitations of such a comparison this comparison shows that a sample of Vietnamese women when they arrive in Australia have relatively high indicators of abdominal obesity for their low BMI and relatively low HDL, while other Vietnamese women who have lived in
272
K.E. Brock et al.
Australia for 2—15 years have the same BMI, lower WHR and a clinically better TC to HDL ratio, and thus do not appear to be at increased heart disease risk as a result of living in Australia. If Vietnamese women maintain their healthy weight range, continue to avoid smoking and increase exercise levels, Australian living may contribute to positive health behaviour and reduced burden of not only heart disease but also diabetes and even colon and breast cancer.
Acknowledgments This research was supported by grant monies from: • Cumberland College, University of Sydney Research Grants and • a NHMRC Public Health Research and Development Grant.
References [1] DIMA. Australia’s population trends and prospects 1996. Canberra: Australian Government; 1997. [2] Yano K, Maclean CJ, Reed DM, Shimizu Y, Sasaki H, Kodama K, et al. A comparison of the 12-year mortality and predictive factors of coronary heart disease among Japanese men in Japan and Hawaii. Am J Epidemiol 1998;127(3):476—87. [3] Rissel C. The development and application of a scale of acculturation. Aust N Z J Public Health 1997;21(6):606—13. [4] Todd R, Gelbier S. ‘Eat more food, get more health’— –attitudes and food habits of a group of Vietnamese refugees. Health Educ J 1988;47(4):149—53. [5] Brock K, Lockwood E, Cant R, Bermingham M, Tran DH. An investigation of health behaviour change in Vietnameseborn living in Sydney. Ethn Dis 2001;11(3):385—90. [6] Bjorntorp P. Portal adipose tissue as a generator of risk factors for cardiovascular disease and diabetes. Arteriosclerosis 1990;10(4):493—6.
[7] Bermingham M, Brock K, Nguyen D, Tran-Dinh H. Body mass index and body fat distribution in newly-arrived Vietnamese refugees in Sydney, Australia. Eur J Clin Nutr 1996;50(10):698—700. [8] Despres JP. Dyslipidemia and obesity. Baillieres Clin Endocrinol Metab 1994;8(3):629—60. [9] Doery J, Edis K, Healy D, Bishop S, Tippet C. Very high prevalence of gestational diabetes in Vietnamese and Cambodian women in Australia. Med J Aust 1989;151(2): 111. [10] Ferro-Luzzi A, Sette S, Franklin M, James W. A simple approach to assessing adult chronic energy deficiency. Eur J Clin Nutr 1992;46(3):173—86. [11] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18(6):499—502. [12] AIHW. Risk factor prevalence study: survey no. 3. Canberra: National Heart Foundation of Australia Committee: Australian Institute for Health and Welfare; 1989. [13] Kleinbaum D, Kupper L, Muller K. Applied regression analysis. Boston: Duxbury Press; 1988. [14] Despres JP, Prud’homme D, Pouliot MC, Tremblay A, Bouchard C. Estimation of deep abdominal adipose-tissue accumulation from simple anthropometric measurements in men. Am J Clin Nutr 1991;54(3):471—7. [15] Masuda T, Imai K, Komiya S. Relationship of anthropometric indices of body fat to cardiovascular risk in Japanese women. Ann Physiol Anthropol 1993;12(3):135— 44. [16] Seidell JC, Han TS, Feskens EJ, Lean ME. Narrow hips and broad waist circumferences independently contribute to increased risk of non-insulin-dependent diabetes mellitus. J Intern Med 1997;242(5):401—6. [17] Lissner L, Bjorkelund C, Heitmann BL, Seidell JC, Bengtsson C. Larger hip circumference independently predicts health and longevity in a Swedish female cohort. Obes Res 2001;9(10):644—6. [18] Lahmann PH, Lissner L, Gullberg B, Berglund G. Differences in body fat and central adiposity between Swedes and European immigrants: the Malmo Diet and Cancer Study. Obes Res 2000;8(9):620—31. [19] Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol 1998;81(4A):7B—12B.
Available online at www.sciencedirect.com
