AJCN EB Program 2010 Early Registration
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kamath, S. K
Right arrow Articles by Alekel, D L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kamath, S. K
Right arrow Articles by Alekel, D L.
Agricola
Right arrow Articles by Kamath, S. K
Right arrow Articles by Alekel, D L.
American Journal of Clinical Nutrition, Vol. 69, No. 4, 621-631, April 1999
© 1999 American Society for Clinical Nutrition

Original Research Communications

Cardiovascular disease risk factors in 2 distinct ethnic groups: Indian and Pakistani compared with American premenopausal women1,2,3

Savitri K Kamath, Erum A Hussain, Daxa Amin, Erica Mortillaro, Barbara West, Charles T Peterson, Faustina Aryee, Genoveva Murillo and D Lee Alekel


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Although people from the Indian subcontinent have high rates of cardiovascular disease (CVD), studies of such in Indian and Pakistani women living in the United States are lacking.

Objective: This study accounted for variability in serum lipid (total cholesterol and triacylglycerol) and lipoprotein [LDL cholesterol, lipoprotein(a), and HDL cholesterol] concentrations in Indian and Pakistani compared with American premenopausal women in the United States. Body composition, regional fat distribution, dietary intake, and energy expenditure were compared between groups.

Design: The 2 groups were 47 Indian and Pakistani and 47 American women. Health was assessed via medical history, physical activity, body composition (via anthropometry and dual-energy X-ray absorptiometry), dietary intake (via 7-d food records), and serum lipids.

Results: Serum total cholesterol, triacylglycerol, LDL cholesterol, lipoprotein(a), the ratio of total to HDL cholesterol, and the ratio of LDL to HDL cholesterol were greater (P <0.03), whereas HDL-cholesterol values were lower (P = 0.011) in Indians and Pakistanis than in Americans. Multiple regression analysis indicated that {approx}18% of the variance in total cholesterol (P = 0.0010) and LDL cholesterol (P = 0.0009) was accounted for by ethnicity, energy expenditure, and the ratio of the sum of central to the sum of peripheral skinfold thicknesses. Ethnicity, sum of central skinfold thicknesses, ratio of polyunsaturated to saturated fat, and monounsaturated fat intake accounted for {approx}43% of the variance in triacylglycerol concentration (P <= 0.0001). Monounsaturated fat, percentage body fat, and alcohol intake accounted for {approx}26% of variance in HDL cholesterol. Ethnicity contributed {approx}22% of the 25% overall variance in lipoprotein(a).

Conclusions: Results suggest that these Indian and Pakistani women are at higher CVD risk than their American counterparts, but that increasing their physical activity is likely to decrease overall and regional adiposity, thereby improving their serum lipid profiles.

Key Words: Ethnicity • lipids • lipoproteins • body composition • regional fat distribution • dietary intake • Indians • Pakistanis • premenopausal women • Americans • cardiovascular disease


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Numerous studies have reported that Asian Indians (defined as people from the subcontinent, which includes India, Pakistan, and Bangladesh) have one of the highest rates of coronary artery disease (CAD) of all the ethnic groups studied (1–8). The incidence of CAD is higher in India than in the Framingham Study (9) and is also high among the Asian Indians in countries to which they have migrated, such as Singapore, Trinidad, the United Kingdom, and the United States (3, 5, 10, 11).

The prevalence of CAD in India has more than doubled in the past 2 decades. This is so in both rural and urban populations, although it is higher in urban than in rural populations, with a greater prevalence in affluent groups (1, 2, 4, 6, 7). Regional differences have also been reported, with the CAD prevalence being higher in southern than in northern India (1, 7, 8). Much of this work was carried out in India, with very little research done in Pakistan (4) or Bangladesh. In Asian Indians, CAD occurs at an early age and in both sexes. For instance, the prevalence of CAD in women in New Delhi was found to be 10% (1). The standardized mortality ratio for CAD in Asian Indian males in the United Kingdom is twice the UK national average between the ages of 30 and 39 y and 3 times greater between the ages of 20 and 29 y (5), indicating premature disease development. Again, the standardized mortality ratio for CAD in Asian Indian immigrant women in the United Kingdom is much higher than that of UK or immigrant US women.

The Asian Indian population is said to have risk factors associated with cardiovascular disease (CVD), such as elevated serum total cholesterol, total triacylglycerol, and lipoprotein(a) [Lp(a)]; lower concentrations of HDL cholesterol; as well as central obesity, hyperglycemia, glucose intolerance, diabetes, and insulin resistance (5, 11–13). However, even those with acceptable serum total cholesterol concentrations have been shown to develop CAD, suggesting that the acceptable cutoff concentration for total cholesterol of 5.2 mmol/L in the United States (9) may actually be too high for the Asian Indian population (14). High total cholesterol in combination with high triacylglycerol and low HDL-cholesterol concentrations, as well as high ratios of total to HDL cholesterol and of LDL to HDL cholesterol, have been shown to be greater risk factors than total cholesterol concentrations alone (15). Furthermore, Lp(a) appears to be another important but often overlooked risk factor for CVD (16, 17). The Lp(a) concentration of Indians in general has been shown to be much higher than that of any other ethnic population, including Europeans or Americans (16, 17). This phenomenon of CAD in Asian Indians has been reviewed in detail (3, 10).

In view of the diet-CAD relation, it is surprising that few studies have considered dietary aspects. The few that have (11, 13, 18–20), showed that the dietary fat intake of Asian Indians in the United Kingdom and the United States is in the range of 30–32% of energy and that the ratio of polyunsaturated to saturated fat (P:S) is {approx}0.8. In general, diets of Asian Indians living in the United States seem to be high in total fat and polyunsaturated fatty acids (PUFAs). Furthermore, few researchers have studied both body composition and dietary risk factors in relation to serum lipids and lipoproteins in Asian Indians. Elevated adiposity has been associated with chronic diseases, particularly CAD, diabetes, and hypertension (21, 22). Yagalla et al (19) studied some of these variables in Asian Indian male physicians. However, there are virtually no studies of Asian Indian women living in the United States. The significance of this study is that there are no published studies on the relative contributions of ethnicity, body composition and size, regional fat distribution, nutritional factors, and physical activity to serum lipid and lipoprotein concentrations in premenopausal women originating from the Indian subcontinent compared with those of their American counterparts. The Asian Indian population in this country is large and growing, with numbers approaching 1 million—about one-half of whom are female (23). In view of the high rate of CAD in Asian Indian women and the tendency of increasing risk of disease as persons migrate, a study of CVD risk factors in premenopausal women is in order before the disease becomes clinically evident.

The overall objective of the present cross-sectional study was to compare known CVD risk factors, including serum lipids and lipoproteins, dietary energy, fat and alcohol intakes, body composition, regional fat distribution, and physical activity level in 2 distinct ethnic groups of premenopausal women—American women and Indian and Pakistani women now living in the United States. More specifically, we were interested in accounting for the variance in serum lipid (total cholesterol and triacylglycerol) and lipoprotein [LDL cholesterol, Lp(a), and HDL cholesterol] concentrations in these 2 ethnic groups.


   SUBJECTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Subject screening and selection
Subject recruitment began in October 1994 and continued until data collection was complete in July 1995. Graduate students recruited potential subjects via notices posted on and around the University of Illinois at Chicago campus, as well as by contacting student organizations on campus (ie, the Indian Student Association and the Pakistani Student Association). Graduate students screened interested individuals over the telephone and those who qualified initially were invited to participate in the study. Subjects were then required to pass a final health and medical screening and receive negative pregnancy test results.

Subjects
Premenopausal Indian and Pakistani and American women of European descent aged 20–40 y with a body mass index (BMI; in kg/m2) between 16 and 30 (to avoid extreme leanness or obesity) were included in the study. They were also required to be eumenorrheic (>=9 regular menstrual cycles/y, 21–35 d in length), nonpregnant, nonsmoking (<=3 cigarettes/d), and without a history of competitive athletics. In addition, exclusion criteria included a history of illnesses or medical conditions such as severe or chronic allergies or asthma, arthritis, bone disorders, cancer, CVD, diabetes mellitus, eating disorders, gastrointestinal or chronic digestive disorders, hypertension, hysterectomy or oophorectomy, liver disease, parathyroid disorders, renal disease, respiratory disease, or thyroid disorders. Women who consumed alcohol heavily or who were chronic users of medications known to affect bone or lipid metabolism (including oral contraceptives) were also excluded. Qualified subjects willing to participate were then asked to sign the consent form. The protocol and consent form were approved by the Institutional Review Board of the University of Illinois at Chicago. Forty-seven Indian and Pakistani and 47 American women met all the inclusion and exclusion criteria and participated in the study.

Data collection
During the afternoon of the first visit, data were collected on health and medical history, physical activity patterns, body composition, and nutrition. All of the questionnaires were administered by interviewers for consistency and to clarify the questions as needed. At the second visit, on day 8 ± 2 of the menstrual cycle (during the follicular phase), each subject brought in a 2-d (48 h) urine collection and 7-d food record, and a fasting blood sample was collected.

Health and medical history questionnaire
The health and medical histories (24) were used both as a secondary screening tool to exclude women not meeting the criteria of the study and to describe the overall health and medical characteristics of each ethnic group. General questions included those on cultural background (eg, country and state of origin if applicable, length of time in the United States), highest degree of education attained, occupation, weight history and eating disorders, personal and family history of CVD or CAD, personal history of cancer and osteoporosis, and history of drug, medication, cigarette, and alcohol use.

Physical activity assessment
Current physical activity was assessed by using the Five-City Project physical activity recall (25). This questionnaire, designed to elicit information on current sleep and physical activity patterns, was validated and used in cross-sectional studies on premenopausal women. It provides an estimate of daily energy expenditure based on the previous week's physical activity pattern.

Body-composition assessment
Body composition was assessed via both dual-energy X-ray absorptiometry (DXA) (QDR-2000+; Hologic, Waltham, MA) and anthropometry. From the whole-body bone density scans (using enhanced array, whole-body software version 5.67A), DXA gave values of lean body mass, fat mass, and percentage body fat, whereas anthropometry gave values of height, weight, waist-to-hip ratio, and skinfold thicknesses. Height was measured twice to the nearest 0.1 cm and the average recorded; weight was taken once by using a balance-beam scale and recorded to the nearest 0.1 kg. Abdominal and hip circumferences were taken twice and recorded to the nearest 0.1 cm. Each subject's abdominal and hip circumferences were used to calculate the waist-to-hip ratio. Skinfold-thickness measurements of subcutaneous fat were made by an experienced anthropometrist using the Harpenden caliper (British Indicators Ltd, St Albans, United Kingdom). Skinfold thicknesses were taken thrice on the right side of the body and recorded to the nearest 0.1 mm; mean values for each site were calculated. The central skinfold sites included the subscapular, suprailiac-waist, and abdomen; the peripheral sites included the tricep, bicep, and thigh. The ratio of the sum of central (Cent = subscapular + suprailiac-waist + abdomen) to the peripheral (Peri = tricep + bicep + thigh) skinfold thicknesses (Cent:peri) was calculated for each subject. Both the waist-to-hip ratio and the Cent:peri are indicative of regional fat distribution.

Nutrition, history, and 7-d food records
The nutrition history questionnaire (26) contained general questions about recent eating habits and was used to clarify items listed in the subjects' 7-d diaries. Questions included type of modified or specialized diet followed (ie, all Indian or Pakistani, all American, vegetarian, or semivegetarian) and whether the subject had modified her diet since arriving in the United States (if applicable). The 7-d food diary was completed in 2- and 3-d segments (ie, Monday, Tuesday, and Wednesday; Thursday and Friday; Saturday and Sunday; or Thursday, Friday, and Saturday; Sunday and Monday; Tuesday and Wednesday) extended over a 3.5-wk period to prevent recorder fatigue and bias due to short-term fluctuations in eating habits and to encompass most of the menstrual cycle (27). The food records were analyzed by 3 cross-trained Indian and Pakistani female nutrition graduate students (who had grown up on the Indian subcontinent) using the NUTRITIONIST IV computerized nutrient database program (28). Food items and mixed dishes were coded and entered according to their individual ingredients rather than by finding a similar American dish. A few recipes for the more common dishes (such as dahl, chapati, or rice dishes) were added to the NUTRITIONIST IV database; when encountered, the graduate students used the same codes for ingredients found in Indian and Pakistani diets. Dietary analysis was used to characterize typical 7-d intakes of energy, macronutrients, and caffeine of these 2 groups.

Blood collection and analyses
Each subject had a 12-h fasted blood sample (20 mL) drawn from the antecubital vein and serum or plasma was separated (centrifuged at 3500 x g for 15 min at 4°C) for measurement of lipids, lipoproteins, glucose, and insulin. Serum and plasma samples were frozen in microcentrifuge tubes at -20°C for {approx}8 mo. Serum lipids and lipoproteins were determined by using Sigma Diagnostics kits (Sigma, St Louis) and following manufacturer's instructions. Serum total cholesterol (29) and triacylglycerol (30) were determined enzymatically, whereas HDL was measured after separation from VLDL and LDL by using the dextran sulfate–Mg2+ procedure (31). The LDL-cholesterol concentration (in mg/dL) was calculated by the Friedewald formula (32): (total cholesterol) - (HDL cholesterol) - (triacylglycerol/5). The total cholesterol, triacylglycerol, and LDL-cholesterol and HDL-cholesterol values were converted from mg/dL to mmol/L. Lp(a) was measured by using the Apo-Tek Lp(a) kit (Perimmune, Rockville, MD) according to the manufacturer's instructions. Serum glucose was determined enzymatically in duplicate with a commercially available spectrophotometric assay (Sigma). Plasma insulin was measured in duplicate by radioimmunoassay (Diagnostic Products, Los Angeles).

Power analysis for sample size determination and statistical analyses of data
Power analysis was based on a multiple regression model for total cholesterol as the outcome variable, contrasting the 2 ethnic groups. The total sample size needed for a model that includes 6 independent variables (k), a directional hypothesis, a desired power of 0.80, a significance level set at 0.05, and a medium effect size (f 2 = 0.15) is equal to 98 (33). The effect size (f 2) for R2 is the ratio of proportion of variance accounted for by the independent variable or variables to residual variance, expressed as f 2 = R2/(1 - R2). The 6 independent variables included ethnic group [Indian and Pakistani (x1 = 0) versus American (x1 = 1)], the best body composition (lean body mass, fat mass, or percentage body fat) or size (total body weight or BMI) factor, the best index of regional fat distribution (waist-to-hip ratio, Cent, or Cent:peri), the 2 best dietary (total energy, total fat, type of fat, fiber, alcohol, or caffeine intake) factors, and daily energy expenditure (MJ/d or MJ•kg-1•d-1). Our multiple regression analyses, which included ethnicity and <=3 additional variables for each lipid and lipoprotein model, with 83 subjects, resulted in a power of 0.78. This was, slightly less than the desired power of 0.80, but nonetheless sufficient to account for a significant proportion of the variance in each lipid or lipoprotein.

Statistical analyses were performed with the SAS computer program, version 6.12 (34), at Iowa State University. Means (±SDs) were calculated on all normally distributed variables and independent t tests were calculated on the lipid and lipoprotein variables to characterize and contrast the 2 ethnic groups. Median, minimum, and maximum values are reported for the variables that were not normally distributed. Frequency distributions were determined separately for each group. Pearson product-moment correlation coefficients were computed among all dependent lipid and lipoprotein variables and independent predictors. The simple bivariate correlations between the predictors and serum lipid and lipoprotein concentrations were used to initially select the best variables for each regression model. An additional consideration in selecting the best predictor was its independent contribution or uniqueness to the model.

Entry and exit of variables for the backward stepwise regression models was P <= 0.10 (default for SAS). Ethnic group was specified as the sampling criterion and thus was not part of the variable selection process and remained in each model. To determine the proportion of variance in serum lipids (total cholesterol and triacylglycerol) and serum lipoproteins [LDL cholesterol, Lp(a), and HDL cholesterol] accounted for in these 2 ethnic groups, the 6 best independent variables specified previously were used for each serum lipid or lipoprotein model.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Subject characteristics: descriptive data
Age, sociodemographics, educational background, and health and medical histories
In the Indian and Pakistani group, 39 were Indian and 8 were Pakistani. The Indian and Pakistani women ranged in age from 19.8 to 38.7 y (x ± SD: 27.5 ± 5.6 y) and the American women from 21.4 to 39.9 y (29.7 ± 6.0 y).

Only 2 of the Indian and Pakistani women were born in the United States; more than one-third (18 subjects, 38%) originated from Gujarat and Bombay, which are in west central India. Thirteen (28%) subjects were from south India in the regions of Hyderabad (Andhrapradesh), Madras (now Chennai, Tamilnado), Bangalore, and Kerala; 10 (21%) were from the northwest region of Pakistan and Indian Punjab; 3 (6.4%) subjects came from Madhiapradesh; and the remaining 3 subjects were from the Indian capital of New Delhi in the north. More than one-quarter (25.5%) of the Indian and Pakistani subjects had lived in the United States for >15 y, 17% from 10 to 15 y, 13% from 5 to 10 y, 11% from 3 to 5 y, 13% from 1 to 3 y; and 17% for <1 y. All of the American women were US born, except one who had migrated 16 y earlier. The highest levels of education achieved in both groups were comparable, with 64% or 28% of Indian and Pakistani and 55% or 43% of American women having completed college or some graduate education, respectively. Nearly all subjects, with the exception of 4 (3 Indian and Pakistani and 1 American) women, were engaged in professions outside the household at the time of the study.

None of the subjects currently used medications known to affect serum lipids or lipoproteins and none had used such medications in the past. None of the Indians and Pakistanis were current smokers, whereas 3 (6%) of the Americans smoked during the study period, but <3 cigarettes/d. None of the subjects had any personal history of CVD or CAD or other related diseases as specified by the selection criteria. Twenty-five (53%) Indian and Pakistani and 19 (40%) American women had no known family history of CVD or CAD, 14 (30%) Indian and Pakistani and 17 (36%) American women had one family member or relative with a known history of CVD or CAD, whereas 8 (17%) Indian and Pakistani and 11 (23%) American women had more than one relative with a known history of CVD or CAD.

Body size, body composition, and physical activity patterns
Mean (±SD) height (P <= 0.0001) and weight (P = 0.0017), respectively, were significantly greater in the American (165.6 ± 6.1 cm and 62.3 ± 7.4 kg) than in the Indian and Pakistani (158.5 ± 5.6 cm and 56.8 ± 9.1 kg) women. However, mean BMI values were not significantly different (P = 0.78) between the Indian and Pakistani (22.6 ± 3.2) and American (22.7 ± 2.6) women. The dissimilar body compositions of these 2 ethnic groups are shown in Figure 1Go. Although mean waist-to-hip ratios were not significantly different (P = 0.34), percentage body fat (P <= 0.0001) and Cent:peri (P = 0.0003), were significantly greater in the Indians and Pakistanis (38.1 ± 6.9% and 1.35 ± 0.26) than in their American counterparts (30.9 ± 8.0% and 1.12 ± 0.33), respectively. In addition, lean body mass was significantly greater (P <= 0.0001) in the American (41.5 ± 4.8 kg) than in the Indian and Pakistani (33.7 ± 4.1 kg) women.



View larger version (29K):
[in this window]
[in a new window]
  		FIGURE 1. Overall body composition and regional fat distribution measures in premenopausal Indian and Pakistani (n = 47) and American (n = 47) women. A: Mean (±SD) percentage body fat and lean body mass determined by dual-energy X-ray absorptiometry; significant differences between groups, P <= 0.0001. B: Mean (±SD) waist-to-hip ratio and ratio of the sum of central to the sum of peripheral skinfold thicknesses (Cent:peri). Central skinfold thicknesses are subscapular, suprailiac-waist, and abdomen, and peripheral skinfold thicknesses are tricep, bicep, and thigh. Significant difference in Cent:peri between groups, P = 0.0003.

 
Most of our subjects were relatively inactive, but both the mean (±SD) total daily energy expenditure (P <= 0.0001) and the daily energy expenditure corrected for body weight (P = 0.0002), respectively, were greater in the American (9.534 ± 1.378 MJ and 0.152 ± 0.017 MJ/kg) than in the Indian and Pakistani (8.036 ± 1.387 MJ and 0.141 ± 0.005 MJ/kg) women; our value for the American subjects compares closely with that of other American women aged 20–35 y (9.623 MJ/d) (25). None of the Indian and Pakistani subjects were engaged in competitive athletics during the study, whereas 8 (17%) of the Americans were participating in recreational athletics, such as softball, volleyball, or bicycling.

Serum lipids and lipoproteins
Not all of the women provided a blood sample for the lipid and lipoprotein measurements and, hence, the sample size was smaller for these values (Table 1Go). Mean (±SD) serum concentrations of total cholesterol, triacylglycerol, and LDL cholesterol were significantly greater in the Indian and Pakistani than in the American women, whereas HDL cholesterol was significantly greater in the American women. Accordingly, the ratios of both total to HDL cholesterol and LDL to HDL cholesterol were greater in the Indian and Pakistani than in the American women. Median values for serum Lp(a) are reported because they were not normally distributed; these were significantly higher in the Indian and Pakistani women than in the American women (Table 1Go).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Serum lipid and lipoprotein profiles of premenopausal Indian and Pakistani and American women
 
Serum glucose and plasma insulin
Because the serum glucose and plasma insulin values for each group were not normally distributed (except glucose values for the Indian and Pakistani women), the median values and their ranges are presented. The values were based on 79 samples for glucose and 44 samples for insulin. For the Indian and Pakistani compared with the American women, respectively, median values (minimum, maximum) were as follows: glucose [5.7 (4.4, 7.7) compared with 5.8 (3.9, 10.3) mmol/L] and insulin [53.8 (28.7, 122.0) compared with 70.3 (46.6, 143.5) pmol/L]. With use of Wilcoxon's rank-sum test (34), the distributions of glucose values between ethnic groups were not significantly different (P = 0.54). However, the distributions of insulin values between groups were significantly different (P = 0.0054), suggesting that the median values in the Americans were greater than those for the Indians and Pakistanis. Insulin was positively correlated with group (r = 0.347, P = 0.021; consistent with the median values of Americans being higher) and positively correlated with glucose concentration (r = 0.420, P = 0.005). Otherwise, glucose and insulin were not correlated with any of the lipids or lipoproteins.

Dietary patterns and dietary intakes
A greater percentage of Indian and Pakistani (38%) than American women (19%) reported following some type of vegetarian diet. Of the Indians and Pakistanis, 11% (n = 5) were vegan, 21% (n = 10) were lactovegetarian, and 6% (n = 3) were lactoovovegetarian. Of the American women, 2% (n = 1) were vegan, 4% (n = 2) were lactovegetarian, and 13% (n = 6) were lactoovovegetarian. As expected, ethnic food consumption differed between the 2 groups. Nineteen percent of the Indian and Pakistani women indicated that they followed an "all-Indian/Pakistani" diet, whereas 53% of the Americans indicated that they followed an "all-American" diet. The remaining 81% of Indian and Pakistani and 47% of American women stated that they consumed foods from a variety of cultures.

Most of the dietary data were not normally distributed and thus median, minimum, and maximum values are reported. Because 5 Indian and Pakistani (n = 42) and 3 American (n = 44) women did not return their 7-d food diaries, the sample sizes are smaller for the dietary (Table 2Go) than for the body-composition data. The median intake of most nutrients, including energy (except when adjusted for body weight), appeared to be greater for the American than for the Indian and Pakistani women. Values reported for alcohol intake from the 7-d food records (daily alcohol in g or as a % of energy) or from the nutrition history reflecting "typical" intake (usual alcohol in g/wk) are based only on those subjects who indicated that they drank alcohol. We excluded abstainers to provide the reader with a true picture of alcohol consumption in each group because most of the Indian and Pakistani women reported no alcohol intake.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Daily nutrient intake from 7-d food diaries of premenopausal Indian and Pakistani and American women
 
Contribution of variables: regression analyses
The residual analysis indicated that the model assumptions of independence of residuals, normality of error terms, and homogeneity of residual variance were satisfied for most of the regression models, except that of the triacylglycerol and Lp(a) models. The residual plot (residual versus predicted values) for the triacylglycerol and Lp(a) models indicated the presence of nonconstant variance. After log-transforming the dependent variables triacylglycerol and Lp(a), the residual plots showed constant variance and were markedly improved. No further variable transformations were necessary to satisfy the regression model assumptions. In addition, there were no significant interactions between or among any of the predictor variables in any of the lipid or lipoprotein regression models. To determine multicollinearity among the variables, the variance inflation factor option in the regression analysis was used, which measures the inflation in the variances of the parameter estimates due to multicollinearities that exist among the regressor variables. The highest variance inflation factor was found for abdominal skinfold thickness (1.75) in the Lp(a) model, with most of the variance inflation factors in each of the regression models being closer to 1.2–1.4, which is considered very acceptable (34).

Serum lipids (total cholesterol and triacylglycerol) and lipoproteins [LDL cholesterol, Lp(a), and HDL cholesterol] were used as outcome variables and the predictors included ethnicity, the best index of body size (body weight or BMI) or composition (lean body mass, fat mass, or percentage body fat), the best index of regional fat distribution (waist-to-hip ratio, Cent, or Cent:peri), energy expenditure (per day or per kg/d), and the most highly correlated dietary indexes (total energy, total fat, saturated fat, monounsaturated fat, PUFAs, P:S, caffeine, or alcohol intake) with the outcome variable for each regression model.

After variable elimination was completed, >18% of the variance in serum total cholesterol (P = 0.0010) was accounted for by ethnicity, energy expenditure per day (strongest predictor of total cholesterol and LDL cholesterol), and Cent:peri (Table 3Go). Likewise, >18% of the variance in serum LDL cholesterol (P = 0.0009) was accounted for by these same variables. In contrast, {approx}43% of the variance in serum triacylglycerol (P <= 0.0001) was accounted for by ethnicity, Cent (strongest predictor of triacylglycerol), P:S, and monounsaturated fat intake (Table 4Go). The negative association of P:S (which was higher in the Indian and Pakistani than in the American women) with triacylglycerol concentration suggests that there was much variability in serum triacylglycerol concentrations among the Indian and Pakistani subjects (Table 1Go). The HDL-cholesterol model had an intermediate amount of the variance (>26%) accounted for by monounsaturated fat (strongest predictor of HDL cholesterol), percentage body fat, and usual alcohol intake (NS, but remained in the model because of inclusion criterion P <= 0.10), whereas ethnicity was not a significant predictor of HDL-cholesterol concentration (Table 4Go). In contrast, regression analysis indicated that ethnicity contributed >22% of the overall variance ({approx}25%) in Lp(a), with smaller contributions from hip circumference and abdominal skinfold thickness (Table 5Go). The negative association of ethnicity with serum total cholesterol, LDL cholesterol, triacylglycerol, and Lp(a) reflects higher circulating concentrations of these lipids and lipoproteins in the Indian and Pakistani than in the American group.


View this table:
[in this window]
[in a new window]
  		TABLE 3. Regression analyses: predictors of serum total cholesterol and LDL cholesterol in premenopausal Indian and Pakistani and American women
 

View this table:
[in this window]
[in a new window]
  		TABLE 4. Regression analyses: predictors of serum total triacylglycerol and HDL cholesterol in premenopausal Indian and Pakistani and American women
 

View this table:
[in this window]
[in a new window]
  		TABLE 5. Regression analysis: predictors of serum lipoprotein(a) in premenopausal Indian and Pakistani and American women
 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The lipid-lipoprotein profile differed significantly between ethnic groups, with the Indian and Pakistani group having a higher risk profile than the American group. However, the mean serum total cholesterol, HDL-cholesterol, and LDL-cholesterol values of both groups were within normal range according to the National Cholesterol Education Program (NCEP) guidelines (9). In particular, the ratios of total to HDL cholesterol and LDL to HDL cholesterol, respectively, were higher in the Indian and Pakistani than in the American women, and higher (4.46 and 2.95) than the values recommended (4.27 and 2.26, calculated from recommended total cholesterol, HDL-cholesterol, and LDL-cholesterol values) for Indians in general (14).

Numerous studies have reported a higher-risk lipid profile in Asian Indians than in western Europeans (11, 12, 18, 35). For example, Indian-born male physicians had significantly higher reported mean total cholesterol [193 ± 31.8 compared with 177 ± 25.3 mg/dL (4.99 ± 0.82 compared with 4.58 ± 0.65 mmol/L)], triacylglycerol [174 ± 181 compared with 86 ± 46 mg/dL (1.96 ± 2.04 compared with 0.97 ± 0.52 mmol/L)], and LDL cholesterol [127 ± 28.5 compared with 120 ± 22.3 mg/dL (3.28 ± 0.74 compared with 3.10 ± 0.58 mmol/L)], but lower HDL-cholesterol [36.2 ± 7.0 compared with 40.3 ± 8.2 mg/dL (0.94 ± 0.18 compared with 1.04 ± 0.21 mmol/L) concentrations, with higher ratios of LDL to HDL cholesterol (3.56 ± 0.96 compared with 3.09 ± 0.88) and total to HDL cholesterol (5.53 ± 1.42 compared with 4.55 ± 1.05) than physicians of European descent, respectively, both living in the United States (35). However, this high-risk lipid profile has not been documented consistently in Asian Indians. A study in west London (21) compared European and Asian women, respectively: the European women had higher total cholesterol (6.29 and 5.96 mmol/L) and HDL-cholesterol (1.58 and 1.38 mmol/L), but lower triacylglycerol (1.21 and 1.38 mmol/L) concentrations.

Although our Indian and Pakistani subjects had, on average, acceptable lipid concentrations as recommended by the NCEP, these values may not be optimal for this population. For populations originating from the Indian subcontinent, concentrations <4.40 mmol/L for total cholesterol, 2.33 mmol/L for LDL cholesterol, and 1.58 mmol/L for triacylglycerol, and >1.03 mmol/L for HDL cholesterol should be considered optimal (14, 36). However, these recommendations are based on prevalence studies and may be biased. More work is needed before definitive conclusions can be made, particularly in view of the association of low cholesterol concentrations with cancer (14). In this study, the American women had lower concentrations of total cholesterol (4.63 and 5.30 mmol/L), LDL cholesterol (2.92 and 3.21 mmol/L), and HDL cholesterol (1.37 and 1.42 mmol/L) than 20–74-y-old women in the third National Health and Nutrition Examination Survey (37), probably because of the younger age range of the women in this study.

The Indian and Pakistani women on average had significantly higher (almost 3-fold) concentrations of Lp(a) than the American women, which is consistent with findings from other studies (16, 17). Asian Indians had a mean concentration 3 times higher (20 mg/dL, or 48 nmol/L) than that of the Chinese (7 mg/dL, or 16.8 nmol/L) (17). Particularly in the presence of high LDL cholesterol and a high ratio of LDL to HDL cholesterol or total to HDL cholesterol, Lp(a) has been suggested to be a major or perhaps the single most important factor contributing to the high prevalence of CAD in Indians (17), even in those with low total cholesterol concentrations. As expected, ethnicity was the strongest contributor (22%) to Lp(a), indicating the importance of genetics in determining Lp(a) concentrations. Indeed, racial differences in Lp(a) have been documented (38). Interestingly, hip circumference also contributed significantly to Lp(a), as was shown previously (39). We have no explanation for this relation. Diet or physical activity did not appear to have any effect on circulating Lp(a). These results agree with published studies reporting that 90% of Lp(a) concentration is genetically determined and not affected by lifestyle factors, including diet (18, 40). On the other hand, some studies have reported that factors such as age, physical exercise, and diet may influence Lp(a) concentrations (38–41). However, the literature on Lp(a) differs from that of other lipids or lipoproteins with respect to CVD risk factors. For instance, the Oslo Study (41) found that physical exercise, a negative modulator of CVD, mediated in part through increased HDL-cholesterol concentrations, actually increases Lp(a) concentrations, a positive modulator of the disease. Thus, at this point, genetics appears to be the major determinant of Lp(a), with other factors being more equivocal.

Both ethnic groups had glucose concentrations lower than the upper recommended values of 6.1 mmol/L (110 mg/dL) (14). This is in contrast with the earlier findings of Miller et al (13), who reported that fasting plasma glucose in European men was significantly lower than that of Indian immigrants in London (5.5 ± 1.6 compared with 6.2 ± 2.7 mmol/L). Glucose intolerance and diabetes have been reported to be higher in Asian Indians than in Europeans (11, 21). The higher median insulin concentration in the American than in the Indian and Pakistani women in our study is in contrast with the findings of McKeigue et al (21), who found that Indian men and women living in London had higher concentrations of fasting and postglucose insulin than European men and women. Hyperinsulinemia, particularly after a glucose load, has been shown frequently in Indians in India as well as in immigrant Indians in the United Kingdom (21, 42). On the basis of the World Health Organization recommendations (43), Singh et al (14) suggested a fasting insulin concentration not more than 144 pmol/L (20 µU/mL) for Indian men and women. The anomaly observed here cannot be explained. The sample size for insulin measurements was small and more studies are needed to confirm or disclaim this finding.

Although the Indian and Pakistani women were smaller in size, which is consistent with the findings of earlier studies, the BMIs of the 2 ethnic groups were similar and also within the range of the World Health Organization recommendation of 18.5–23 for those in developing countries (14, 43). The BMI of Indians in India and in the United Kingdom, on average, has been shown to be less than that of Britons (44). The higher BMI (22.6 ± 3.2) of our immigrant subjects than of the Indians in India (21.1) may be attributable to either higher energy and fat intake, lower energy expenditure, or both.

Some studies (45–47) have reported that dietary fat intake may be a strong determinant of body fat. However, only percentage of energy from fat (r = 0.283, P = 0.0084) and not total fat (r = 0.152, P = 0.16) was correlated with percentage body fat in these premenopausal women. Although waist-to-hip ratio was not different between groups and was below the cutoff point of 0.85 suggested for Indian women (14), the Cent:peri was different and a better indicator of body fat distribution, probably because circumferences include bone and muscle mass as well as fat. Skinfold thicknesses estimate subcutaneous fat and are independent of lean body mass. Because subcutaneous abdominal fat is typically highly correlated with metabolically active visceral fat, we presumed that these Indian and Pakistani women had greater intraabdominal fat than their American counterparts and thus may be at greater risk for the metabolic derangements associated with abdominal adiposity. The greater adiposity, despite relatively low BMIs, in ethnic groups who move to more industrialized cultures is of current interest in view of its association with CVD and diabetes. Studies have shown that Asian Indians have a lower BMI and higher percentage body fat than whites (48, 49). Upper body fat and abdominal fat with its accompanying high lipolytic activity, rather than total body fat, have been associated with high serum lipids and CVD risk. Overall, the body composition of these Indian and Pakistani women indicated higher CVD risk than in the American women.

The median dietary intake of 30% of energy from fat in the Indian and Pakistani subjects appeared higher than that of the American subjects (26%) and also higher than that reported in a study from India (14): 14.8% of energy as fat in a rural population compared with 24.7% in an urban population. In contrast, dietary fat intake of Asian Indians in the United Kingdom and the United States is in the range of 30–32% of energy (19, 21), comparable with that found in our preliminary study of a group of Asian Indian men (20). The recommended intake of energy from fat is 21% for Indians (14). It is surprising that the median fat intake of the American women (26% of energy) was considerably lower than that reported in the third National Health and Nutrition Examination Survey (33.7% and 34.5% of energy for white women 20–29 and 30–39 y of age, respectively) (50) but may be explained by the health and fat consciousness of women represented in this study. Dissimilar cooking methods between the groups may have contributed to the difference in fat intake. Lip et al (51) reported that Indians in the United Kingdom, who mainly use frying as the method of cooking, purchased more fat per capita than did whites and blacks, who used boiling and broiling as methods of cooking. In addition, Indians and Pakistanis usually consume snacks fried in vegetable oil. The higher P:S (0.81), however, is consistent with that observed by other investigators in Asian Indian immigrants (11, 19, 20). We expected to find a higher fiber intake in the Indian and Pakistani women than in their American counterparts, considering the greater percentage of Indian and Pakistani vegetarians. However, Asians consider vegetables side dishes and thus consume them in small servings (52).

The regression models indicated that dietary energy did not contribute to the variance in any lipid or lipoprotein, nor did total fat intake, as has been reported in other studies (11, 19). Dietary monounsaturated fat was the only and strongest positive dietary contributor to HDL cholesterol, but a negative contributor to triacylglycerol, as was the P:S. Other studies have reported an association of dietary constituents with lipids and lipoproteins (53, 54), with a high fat intake being one of the risk factors for CVD. Yagalla et al (19) in their physicians' study observed a significant contribution of carbohydrate intake to the variance in serum triacylglycerol. The slight association of dietary constituents in our observational study might be because our subjects were young and had acceptable serum total cholesterol and LDL-cholesterol concentrations. Previous studies have examined the relation of diet to serum lipids in subjects with multiple risk factors as well as the effects of high-fat diets in clinical trials (53, 55). Whereas saturated fatty acids are known to raise and PUFAs to lower LDL cholesterol and HDL cholesterol, monounsaturated and short-chain fatty acids have been thought to be neutral. However, Mattson et al (54) suggested that monounsaturated fats do lower total and LDL cholesterol (and perhaps triacylglycerol) when substituted for saturated fat and may raise HDL-cholesterol concentrations. This is illustrated by the health-promoting effects of Mediterranean-style diets that are high in monounsaturated fats.

Alcohol intake was positively associated with HDL cholesterol in the regression model and with Lp(a) on the basis of simple correlations (r = -0.246, P = 0.027); the latter finding has not yet been documented in the literature. However, numerous studies have reported a strong association between alcohol intake and HDL cholesterol in women, with drinkers having HDL-cholesterol concentrations 6–18% higher than those of nondrinkers (56, 57). The higher HDL-cholesterol concentration in the American group may be due in part to their higher intake of alcohol, but also to their lower percentage body fat, higher energy expenditure, or both. Note that only 16 Indian and Pakistani, compared with 38 American women, reported typical intakes of alcohol.

Physical activity, as reflected by energy expenditure, significantly influenced total and LDL cholesterol, but not triacylglycerol, Lp(a), or HDL cholesterol. Although many studies have found an association of physical activity with serum HDL cholesterol, these were mostly intervention studies using intensive exercise (58). Again, ours was a cross-sectional study.

Abdominal fat, as indicated by either Cent:peri or Cent, remained in the models of total cholesterol, LDL cholesterol, and triacylglycerol, contributing modestly to the variance of each lipid variable. Central obesity has been shown to correlate strongly with CVD risk factors (21, 59). This is not surprising because abdominal fat with high lipolytic activity has been implicated in increasing serum lipids and augmenting CVD risk (59). Besides monounsaturated fat intake, percentage body fat also contributed significantly to HDL-cholesterol concentrations, but not to the same extent. Ethnicity contributed significantly to total cholesterol, triacylglycerol, LDL cholesterol, and Lp(a), being a particularly strong predictor for Lp(a) and triacylglycerol. In fact, ethnicity turned out to be the single most important predictor for Lp(a), as was shown by other investigators (17). Despite the significant simple correlation between ethnicity and HDL cholesterol (r = 0.279, P = 0.01), ethnicity did not contribute significantly to HDL cholesterol in the regression model.

The unfavorable lipid profile (higher total cholesterol, triacylglycerol, ratio of total to HDL cholesterol, ratio of LDL to HDL cholesterol, and Lp(a), and lower HDL cholesterol), more centrally deposited fat, and lower physical activity in the Indian and Pakistani compared with the American women indicate that this group of Indian and Pakistani women is at a higher risk for CVD than their American counterparts. Whether the CVD risk factors of centrally deposited fat and lipid profiles are attributable to ethnicity or lower energy expenditure in the Indian and Pakistani women cannot be conclusively stated on the basis of this cross-sectional study. The best way to test the effect of exercise and energy expenditure on body composition and fat deposition is under experimental conditions. This observational study did not test the effect of exercise per se. Furthermore, in this study we saw a similar difference in energy balance (energy intake - energy expenditure) in both the Indian and Pakistani ({approx}15%) and American ({approx}17%) women, suggesting that energy balance may not explain the differences in body composition or serum lipids and lipoproteins between these 2 ethnic groups.

The findings of this study indicate that there are significant differences between these 2 ethnic groups with respect to the examined CVD risk factors. Some of the differences may be attributable to ethnicity. However, some of the differences in body composition and regional fat distribution could be explained by the greater energy expenditure of the American compared with that of the Indian and Pakistani women. Likewise, a large proportion of the higher lipid risk profile of the Indian and Pakistani women could be attributed to their greater percentage body fat and centrally deposited fat. This, in turn, may be related to their lower physical activity level. Although dietary factors played some role in determining serum total cholesterol and HDL-cholesterol concentrations, they were not as important as body composition and regional fat distribution or energy expenditure. There is general consensus among international agencies that for Asians, including those from the Indian subcontinent, fat intake should not exceed 21% of energy, serum total cholesterol should be <=4.40 mmol/L, and BMI should not exceed 23. The Indian and Pakistani women in this study had a higher than recommended serum total cholesterol concentration of 5.15 mmol/L and a higher median fat intake than the American women. Additionally, although both groups were relatively sedentary, these Indian and Pakistani women had a lower energy expenditure than their American counterparts.

Thus, to improve the overall CVD risk among Indian and Pakistani women, the primary goal should be to decrease percentage body fat and centrally deposited fat and increase lean body mass, thereby favorably altering the serum lipid profile. The results from this study suggest that increasing physical activity is likely the single best approach for changing the body composition and regional fat distribution of these premenopausal women. However, lowering dietary fat intake to favorably modify the serum lipid and lipoprotein profile should also be a recommendation. An important aspect of these results is that they suggest the need for a clinical trial examining the effect of physical exercise, body composition, regional fat distribution, dietary factors (particularly type of fat), and perhaps other potential risk factors on serum lipids and lipoproteins in Indian and Pakistani women. To examine these CVD risk factors and their influence on disease development, it would be valuable to evaluate both pre- and postmenopausal women in both ethnic groups.


   ACKNOWLEDGMENTS
 
We thank Noreen Ahmed and Noopur Goyal, graduate students in the Department of Human Nutrition and Dietetics, University of Illinois at Chicago, for their help with testing of subjects and dietary analysis, and Larry Brace, Department of Pathology, University of Illinois at Chicago, for determining the serum Lp(a) concentrations.


   FOOTNOTES
 
1 From the College of Health and Human Development Sciences, University of Illinois at Chicago; Appalachian Regional Health Care, South Williamson, KT; the American Academy of Physical Medicine and Rehabilitation, Chicago; Doctor's Data, Inc, West Chicago; the Department of Statistics, Iowa State University, Ames; the Department of Biochemistry, National College of Chiropractic, Lombard, IL; and the Department of Food Science and Human Nutrition, Iowa State University, Ames.

2 Supported in part by a grant from the Campus Research Board at the University of Illinois at Chicago.

3 Address reprint requests to SK Kamath, College of Health and Human Development Sciences, University of Illinois at Chicago, 808 South Wood Street (mail code 518), Chicago, IL 60612. E-mail: savitri.kamath{at}uic.edu.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Chadha SL, Radhakrishan S, Ramachandran K, Gopinath N. Epidemiological study of coronary heart diseases in urban population of Delhi. Indian J Med Res 1990;92:424–30.[Medline]
  2. Beegom R, Singh RB. Prevalence of coronary artery disease and its risk factors in the urban population of south and north India. Acta Cardiol 1995;50:227–40.[Medline]
  3. Enas EA, Yusuf S, Mehta JL. Prevalence of coronary artery disease in Asian Indians. Am J Cardiol 1992;70:945–9.[Medline]
  4. Hameed K, Kadir M, Gibson T, Sultana S. The frequency of known diabetes, hypertension and ischemic heart disease in affluent and poor urban populations of Karachi, Pakistan. Diabet Med 1995;12:500–3.[Medline]
  5. Balarajan R. Ethnic differences in mortality from ischaemic heart disease and cerebrovascular disease in England and Wales. BMJ 1991;302:560–4.
  6. Sarvotham SG, Berry JN. Prevalence of coronary heart disease in an urban population in northern India. Circulation 1968;37:939–53.[Abstract/Free Full Text]
  7. Singh RB, Naiz AM, Agarwal P, Rastogi SS, Beegum R, Singh NK. Epidemiologic study of central obesity, insulin resistance, and associated disturbances in an urban population of north India. Acta Cardiol 1995;3:215–25.
  8. Raman Kutty V, Balakrishnan KG, Jayasree AK, Thomos J. Prevalence of coronary artery disease in the rural population of Thiruanan Thapuram, Kerala, India. Int J Cardiol 1993;39:59–70.[Medline]
  9. Wilson PWF, Christianson JC, Anderson RM, Kannel WB. Impact of national guidelines for cholesterol screening: The Framingham Offspring Study. JAMA 1989;262:41–4.[Abstract/Free Full Text]
  10. McKeigue PM, Miller GJ, Marmot MG. Coronary heart disease in South Asians overseas: a review. J Clin Epidemiol 1989;42:597–609.[Medline]
  11. McKeigue PM, Marmot MG, Adelstein AM. Diet and risk factors for coronary artery disease in Asians in north-west London. Lancet 1985;2:1060–90.[Medline]
  12. McKeigue PM, Marmot MG. Mortality from coronary heart disease in Asian communities in London. BMJ 1988;297:903.
  13. Miller GJ, Kotecha S, Wilkinson WH, et al. Dietary and other characteristics relevant for coronary heart disease in men of Indian, West Indian and European descent in London. Atherosclerosis 1988;70:63–72.[Medline]
  14. Singh RB, Sainani GS, Rastogi SS, et al. Diet and lifestyle guidelines and desirable levels of risk factors for prevention of coronary heart disease in Indians. J Intern Med 1996;7:46–9.
  15. Bush TL, Fried LP, Barrett-Connor E. Cholesterol, lipoprotein, and coronary heart disease in women. Clin Chem 1988;34:B60–70.
  16. Sandholzer C, Halman DM, Saha N, Sirgurdsson G, Ccaszar A. Effect of apolipoprotein(a) polymorphism on the lipoprotein(a) concentration in seven ethnic groups. Hum Genet 1991;86:607–14.[Medline]
  17. Enas EA, Mehta J. Malignanleftt coronary artery disease in young Asian Indians: thoughts on pathogenesis, prevention, and therapy. Clin Cardiol 1995;18:131–5.[Medline]
  18. Enas EA, Garg A, Davidson MA, Nair VM, Huet BA, Yusuf S. Coronary heart disease and its risk factors in first-generation immigrant Asian Indians to the United States of America. Indian Heart J 1996;4:343–53.
  19. Yagalla M, Hoerr S, Song W, Enas E, Garg A. Relationship of diet, abdominal obesity, and physical activity to plasma lipoprotein levels in Asian Indian physicians residing in the US. J Am Diet Assoc 1996;96:257–61.[Medline]
  20. Kamath SK, Ravishanker C, Broones E, Chen EH. Macronutrient intake and blood cholesterol level of a community of Asian Indians living in the United States. J Am Diet Assoc 1997;97:299–301.[Medline]
  21. McKeigue PM, Shah B, Marmot MG. Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet 1991;337:382–6.[Medline]
  22. Hartz A, Grubb B, Wild R, Van Nort JJ, Freedman D, Rimm A. The association of waist-to-hip ratio and angiographically determined coronary heart disease. Int J Obes 1990;14:657–65.[Medline]
  23. Bureau of the Census. We, the Asian Americans. Washington, DC: US Government Printing Office, 1993.
  24. Alekel L, Clasey JL, Fehling PC, et al. Contributions of exercise, body composition, and age to bone mineral density in premenopausal women. Med Sci Sports Exerc 1995;27:1477–85.[Medline]
  25. Sallis JF, Haskell WL, Wood PD, et al. Physical activity assessment methodology in the Five-City Project. Am J Epidemiol 1985;121:91–106.[Abstract/Free Full Text]
  26. Alekel L. Contributions of physical activity, body composition, age, and nutritional factors to total and regional bone mass in premenopausal aerobic dancers and exercisers. PhD dissertation. University of Illinois, Urbana, 1993.
  27. Block G. A review of the validations of dietary assessment methods. Am J Epidemiol 1982;115:492–505.[Free Full Text]
  28. N-Squared Computing. Nutritionist IV. Version 6.0 for IBM and compatible computers. San Bruno, CA: First Data Bank Inc, 1993.
  29. Allain CA, Poon LS, Chan CSG, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem 1974;20:470–5.[Abstract]
  30. Buoccolo G, David H. Quantitative determination of serum triglycerides by the use of enzymes. Clin Chem 1973;19:476–82.[Abstract]
  31. Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantitation of high density lipoprotein cholesterol. Clin Chem 1982;28:1379–88.[Free Full Text]
  32. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low density lipoprotein cholesterol in plasma without the use of preparatory ultracentrifugation. Clin Chem 1972;18:499–502.[Abstract]
  33. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Laurence Erlbaum Association, Inc, 1988.
  34. SAS Institute Inc. SAS/STAT user's guide, version 6. 4th ed. Vol 1. Cary, NC: SAS Institute Inc, 1989.
  35. Thomas I, Gupta S, Sempos C, Cooper R. Serum lipids of Indian physicians living in the US compared to US born physicians. Atherosclerosis 1986;61:99–106.[Medline]
  36. Gopinath N, Chadha SL, Sehgal A, Shekhawat S, Tandon R. What is a desirable lipid profile? Indian Heart J 1994;46:325–7.[Medline]
  37. Ernst ND, Sempos CT, Briefel RR, Clark MB. Consistency between US dietary fat intake and serum total cholesterol concentrations: the National Health and Nutrition Examination Surveys. Am J Clin Nutr 1997;66(suppl):965S–72S.[Abstract/Free Full Text]
  38. Guyton JR, Dahlen GH, Patsch W, Kautz JA, Gotto A. Relationship of plasma Lp(a) levels to race and apolipoprotein B. Arteriosclerosis 1985;5:265–72.[Abstract/Free Full Text]
  39. Slunga L, Asplund K, Johnson O, Dahlen GH. Lipoprotein(a) in a randomly selected 25–64 year old population: the Northern Sweden Monica Study. J Clin Epidemiol 1993;46:617–24.[Medline]
  40. Mackinnon LT, Hubinger L, Lepre F. Effects of physical activity and diet on lipoprotein(a). Med Sci Sports Exerc 1997;29:1429–36.[Medline]
  41. Holme I, Urdal P, Anderson S, Hjermann I. Exercise-induced increase in lipoprotein(a). Atherosclerosis 1996;122:97–104.[Medline]
  42. Singh RB, Ghosh S, Niaz AM, et al. Epidemiologic study of diet and coronary risk factors in relation to central obesity and insulin levels in rural and urban populations of North India. Int J Cardiol 1995;47:245–55.[Medline]
  43. World Health Organization Study Group. Diet, nutrition and prevention of chronic disease. Geneva: WHO, 1990.
  44. Potts J, Simmons D. Sex and ethnic group differences in body fat distribution in young United Kingdom South Asians and Europids. J Clin Epidemiol 1994;47:837–41.[Medline]
  45. Danforth E Jr. Diet and obesity. Am J Clin Nutr 1985;41(suppl):1132–45.[Free Full Text]
  46. Flatt JP. The biochemistry of energy expenditure. In: Bray GA, ed. Obesity research II. London: Newman, 1978:211–28.
  47. Schutz Y, Flatt JP, Jequier E. Failure of dietary fat intake to promote fat oxidation: a factor favoring the development of obesity. Am J Clin Nutr 1989;50:307–14.[Abstract/Free Full Text]
  48. Wang J, Thornton JC, Russell M, Burastero S, Heymsfield S, Pierson RN Jr. Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am J Clin Nutr 1994;60:23–8.[Abstract/Free Full Text]
  49. Dodson DJ, Hooten TM, Buchwald D. Prevalence of hypercholesterolemia and coronary heart disease risk factors among Asian refugees in a primary care clinic. J Clin Pharm Ther 1995;20:83–9.[Medline]
  50. McDowell M, Briefel R, Alaimo K, et al. Energy and macronutrient intake of persons ages 2 months and over in the Third National Health and Nutrition Examination Survey, Phase I, 1988–91. Hyattsville, MD: National Center for Statistics, 1994. [Publication no. DHHS(PHS)95-1250.]
  51. Lip GY, Malik I, Luscombe C, McCarry M, Beevers G. Dietary fat purchasing habits in whites, blacks and Asian peoples in England—implications for heart disease prevention. Int J Cardiol 1995;48:287–93.[Medline]
  52. Kittler PG, Sucher K. Food and culture in America. A nutrition handbook. Reinhold, NY: Van Nostrand, 1989:333–61.
  53. Katan MB, Zock PL, Mensink RP. Dietary oils, serum lipoproteins, and coronary heart disease. Am J Clin Nutr 1995;61(suppl):1368S–73S.[Abstract]
  54. Mattson R, Grundy SM, Vega GL. Two different views of the relationship of hypertriglyceridemia to coronary heart disease: implications for treatment. Arch Intern Med 1992;152:28–34.[Abstract/Free Full Text]
  55. Anderson JW, O'Neal DS, Riddell-Mason S, Floore TL, Dillon DW, Oetgen PR. Postprandial serum glucose, insulin, and lipoprotein responses to high and low fiber diets. Metabolism 1995;44:848–54.[Medline]
  56. Castelli WP, Doyle JT, Gordon T. High density lipoprotein cholesterol and other lipids in coronary heart disease. The Cooperative Lipoprotein Phenotyping Study. Circulation 1977;55:767–72.[Abstract/Free Full Text]
  57. Haarbo J, Hassager C, Schlemmer A, Christiansen C. Influence of smoking, regional fat distribution, and alcohol consumption on serum lipids, lipoproteins, and apolipoproteins in early postmenopausal women. Atherosclerosis 1990;84:239–44.[Medline]
  58. Perry AC, Applegate EB, Allison ML, Miller PC, Signorile JF. Relation between anthropometric measures of fat distribution and cardiovascular risk factors in overweight pre- and postmenopausal women. Am J Clin Nutr 1997;66:829–36.[Abstract/Free Full Text]
  59. Fujioka S, Matsuzama Y, Tokunaga K, Tarui S. Contribution of intra-abdominal fat accumulation to the impairment of glucose and lipid metabolism in human study. Metabolism 1987;36:54–9.[Medline]
Received for publication January 29, 1998. Accepted for publication November 23, 1998.




This article has been cited by other articles:


Home page
 J. Clin. Endocrinol. Metab.Home page
A. Misra and L. Khurana
Obesity and the Metabolic Syndrome in Developing Countries
J. Clin. Endocrinol. Metab., November 1, 2008; 93(11_Supplement_1): s9 - s30.
[Abstract] [Full Text] [PDF]

Home page
 Diabetes CareHome page
A. Misra, J. S. Wasir, and R. M. Pandey
An Evaluation of Candidate Definitions of the Metabolic Syndrome in Adult Asian Indians
Diabetes Care, February 1, 2005; 28(2): 398 - 403.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
M. Chandalia, A. V. Cabo-Chan Jr, S. Devaraj, I. Jialal, S. M. Grundy, and N. Abate
Elevated Plasma High-Sensitivity C-Reactive Protein Concentrations in Asian Indians Living in the United States
J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3773 - 3776.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
M. Chandalia, N. Abate, A. V. Cabo-Chan Jr., S. Devaraj, I. Jialal, and S. M. Grundy
Hyperhomocysteinemia in Asian Indians Living in the United States
J. Clin. Endocrinol. Metab., March 1, 2003; 88(3): 1089 - 1095.
[Abstract] [Full Text] [PDF]

Home page
 J. Am. Coll. Nutr.Home page
S. S. Jonnalagadda and S. Diwan
Nutrient Intake of First Generation Gujarati Asian Indian Immigrants in the U.S.
J. Am. Coll. Nutr., October 1, 2002; 21(5): 372 - 380.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Chem.Home page
M. Haidari, E. Javadi, A. Sanati, M. Hajilooi, and J. Ghanbili
Association of Increased Ferritin with Premature Coronary Stenosis in Men
Clin. Chem., September 1, 2001; 47(9): 1666 - 1672.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Clin. Nutr.Home page
R. B Singh
Coronary artery disease risk factors in south Asian and American premenopausal women
Am. J. Clinical Nutrition, December 1, 1999; 70(6): 1112 - 1113.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kamath, S. K
Right arrow Articles by Alekel, D L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kamath, S. K
Right arrow Articles by Alekel, D L.
Agricola
Right arrow Articles by Kamath, S. K
Right arrow Articles by Alekel, D L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS