- Research article
- Open Access
Dietary quality and patterns and non-communicable disease risk of an Indian community in KwaZulu-Natal, South Africa
© Naicker et al. 2015
- Received: 29 May 2015
- Accepted: 26 June 2015
- Published: 8 August 2015
Limited data exist on the South African Indian diet despite their high prevalence of non-communicable diseases. This study attempted to determine the dietary quality and patterns of an Indian population in KwaZulu-Natal with reference to the high prevalence of non-communicable disease
Two-hundred-and-fifty apparently healthy Indians, aged 35–55 years participated in a cross-sectional study where diet was assessed using a validated quantitative food frequency questionnaire. Mean intakes were compared to the World Health Organization goals. Dietary quality was determined by index construction and dietary patterns by factor analysis.
The mean daily percentage of energy (%E) from n-3 fatty acids (0.24 %E), dietary fibre (18.4 g/day) and fruit and vegetable intakes (229.4 g/day) were below the World Health Organization goals. Total fat (36.1 %E), polyunsaturated fatty acids (11.8 %E), n-6 fatty acids (11 %E) and free sugars (12.5 %E) exceeded the goals. The means for the deficient index reflected a moderate diet quality whereas, the excess index reflected good diet quality. The Pearson partial correlation coefficients between the deficient index and risk markers were weak whilst, the excess index was inversely correlated with waist circumference for the whole sample. Two factors were identified, based on the percentage of fat that contributed to each food group: factor 1 (meat and fish versus legume and cereal pattern), which accounted for added fat through food preparation; and Factor 2 (nuts and seeds versus sugars and visible fat pattern), which accounted for obvious fat. The medians for waist circumference, blood glucose, cholesterol and triglyceride levels showed significant decreasing trends for factor 1 (p < 0.05). The medians for blood glucose and cholesterol showed significant decreasing trends for factor 2 (p < 0.01).
A shortfall of fruit and vegetable, fibre and n-3 fatty acid intake in the diet is highlighted. When assessing the diet quality and patterns, guidance on the prudent use of added fats may lead to a healthier lifestyle reducing the prevalence of non-communicable diseases.
- Diet quality
- Diet patterns
- Non-communicable diseases
Global diet is undergoing an alarming transition: staple foods are becoming more refined and processed; fat and meat intakes are increasing; more processed dairy products and other foods are consumed; and larger numbers of meals are consumed outside the home, making households more reliant on the food industry. Dietary habits have collectively been influenced by urbanisation, economic development, market globalisation and industrialisation where the move from a traditional to a Western-type diet is a characteristic feature . Consequently, this transition has been associated with an increase in the global prevalence of non-communicable diseases (NCDs) .
The role of diet in the aetiology of most NCDs is extremely important. A body of literature attests to the fact that diets and specific nutrient deficiencies and excesses influence the development of NCDs and that appropriate dietary changes may reduce the risk of NCDs . Little is known about the South African Indian diet, despite their high incidence of diabetes mellitus  and coronary heart disease (CHD) . The first migrant Indians were reported to arrive in South Africa in 1860 to work as sugarcane farm labourers. With equity and social transformation, their lifestyles have changed rapidly for the better, however, their risk for NCDs have reached epidemic proportions [5, 6].
Diet is considered a modifiable risk factor for NCDs . Over recent years dietary assessment has moved away from single nutrients to assessing the whole diet and interrelations between dietary factors in relation to NCDs [8, 9]. Therefore, a number of epidemiologists propose using a dietary pattern approach to investigate food in relation to NCDs [10, 11]. Most studies use either, a priori methods such as diet indices to assess the quality of the diet or a posteriori data-driven methods, such as factor analysis to derive food patterns in an attempt to describe the dietary behaviour of a population  and the associated risks thereof.
The primary hypothesis of this study was that dietary quality and patterns are related to risk markers for developing nutrition–related NCDs among Indians in KwaZulu-Natal (KZN), South Africa. Therefore, the aim of this paper is to compare dietary intakes with the World Health Organization (WHO) population nutrient goals for the prevention of death and disability from NCDs  and to describe the dietary quality and patterns in relation to the risk markers for NCDs in an Indian community in KZN, South Africa.
At the time of the 2011 South African census, the total population of Indians was 1 286 930, of whom the majority, totalling 757 000, resided in the province of KZN . Two-hundred-and-fifty apparently healthy individuals within the age range of 35–55 years were recruited from the homogenous Indian community, Stanger in the province of KZN. Once eligibility was established, participants who were willing to participate in the study were randomly selected. Individuals diagnosed previously with hypercholesterolaemia and/or diabetes mellitus and individuals using dietary food supplements were excluded. The other exclusion criteria were pregnancy, lactation, any acute or chronic illness, use of chronic medication and inability to communicate freely. Informed consent was sought and received from study participants. The study was approved by the Ethics Committee of North-West University (05 M15).
Habitual dietary intakes were measured by a validated quantitative food frequency questionnaire (QFFQ)  modified to include Indian foods and with a one-month recall period. It was acknowledged that the diet might be affected by seasonal variability. However, fresh fruits and vegetables are widely available throughout the year in KZN while other food products obtained via the retail food markets are not affected by seasonality . In a preliminary study, an inventory of foods commonly consumed by South African Indians was compiled from interviews conducted in 25 randomly selected households. Since the type and amount of ingredients and preparation methods of composite dishes vary according to differences in religious and economic status  all traditional recipes were collected. Final construction of the recipes was determined by averaging three recipes per dish. A prototype QFFQ was developed from the food list. Foods were grouped together, starting with the most frequently consumed staple foods and ending with the least common foods, totaling 285 listed foods. Thereafter, the QFFQ was piloted among ten randomly selected households from the preliminary study sample to determine the completeness and appropriateness of the food list.
The QFFQ was then tested for reproducibility and comparative validity against three 24-hour food recalls as the reference method in a sub-sample of 50 participants. The QFFQ, was shown to be relatively reproducible (range for Pearson r: 0.4 to 0.7 – total carbohydrate 0.4, energy 0.44, total fat 0.48, total protein 0.49, total polyunsaturated fatty acids 0.55, niacin 0.63, cholesterol 0.7) but to have moderate to poor validity (range for deattenuated Pearson r: 0.008–0.5 – iron 0.008, energy 0.39, total carbohydrate 0.42, total fat 0.44, total protein 0.46, cholesterol 0.5) in comparison to the reference method. Recommendations derived from the comparative validity and reproducibility results were taken into account in the main study to limit quantification, portion size estimation error and respondent fatigue. The person responsible for food preparation was required to be present at the interview to assist the participant to estimate the amount of foods consumed. Measures were put into place to limit participant fatigue due to the length of the QFFQ by arranging convenient interview times. The first 50 participants completed three 24-hour food recalls and two QFFQs (as part of the comparative reproducibility and validation study); thereafter, the remaining 200 participants completed one QFFQ and 24-hour food recall. In total, 250 participants were interviewed by trained field workers. A random selection of days was used. Weekends and days that were unusual in terms of normal dietary intake – e.g. dry, saltless and whole-food fasting, which are typical of a Hindu’s religious prescription – were avoided. In order to determine portion sizes, the Medical Research Council (MRC) Dietary Assessment Education Kit was used in combination with typical household utensils .
Physical and biochemical measurements
Anthropometric measurements [weight, height and waist circumference (WC)] were measured using standardised techniques and instruments . Participants were instructed to fast for 10 h before biochemical assessment. Blood pressure was recorded using a standard mercury sphygmomanometer according to standard guidelines. Only finger-prick blood samples were taken because of limited financial and infrastructure resources. Fasting blood glucose, triglycerides and total cholesterol were recorded using the Accutrend meter (Accutrend GCT kit, Accu-Check Softclix Pro Lancets) and strips (Accutrend Cholesterol, Glucose and Triglyceride Strips) supplied by Roche Products Pty Ltd, Randburg, Gauteng. All biochemical measurements were recorded by a registered nurse.
Score risk marker
Asian Indians have been shown to have more than a two-fold risk of dying from coronary artery disease (CAD) from any combination of risk factors as compared with Caucasians . For comparative purposes, two scores were developed for this study; namely “risk score”, which was an aggregate total of all risk markers defined by Asian standards  and “risk score 1” which was the aggregate of all risk markers, with fasting glucose counting twice as much as the other risk markers because of many and strong correlations with nutrients.
Dietary data analysis
The FoodFinder 3® dietary analysis software program  based on the South African MRC food composition tables was used to capture the dietary data. Indigenous Indian ingredients and recipes not coded in the MRC food composition tables were added, using the nutritive value of Indian foods as a guide .
In order to determine the relationship between food consumption and disease prevalence, dietary quality and patterns were determined by index construction and factor analysis. Two dietary quality indices, a deficiency index (Index 1) and an excess index (Index 2), were calculated from the unadjusted nutrient intakes, following the procedure of Thiele et al. . Calculation of the deficiency index (a summary score similar to the nutrient adequacy and mean nutrient adequacy ratios ) and the excess index provide a summary of both adequacy and the risk of excess of the total diet which might be lost in a single aggregate score .
In preparation for factor analysis, food items from the QFFQ were aggregated into 12 food groups based on the system of the International Network of Food Data Systems (INFOODS)  using the FoodFinder 3 program® . The percentages of total energy, total fat, saturated (SFA), PUFA, monounsaturated fat (MUFA), n-3 and n-6 fatty acids derived from each food group were calculated.
The Statistical Package for Social Science (SPSS)  version 17® was used to calculate frequency tables, means, medians, and standard deviations for anthropometric and biochemical measurements, nutrient intakes and diet quality. Pearson partial correlations, using BoxCox transformed data and controlled for age and smoking were used to detect associations between diet indices and clinical risk markers. Principal component analysis with Varimax rotation was performed on the percentage total fat derived from each of the food groups. Principal components derived from the percentage total fat intake with eigenvalues of >1 were used for further analyses . Factor scores were calculated by adding the percentage fat from each food group weighted by the factor loading for each participant , and divided into quintiles of the distribution. Medians of the risk markers were calculated for the first quintile, second to fourth quintiles combined, and fifth quintile. The Jonckheere-Terpstra test  was used to test for trends in the medians of the risk markers for the above-mentioned three groups of factor scores.
Anthropometric characteristics and clinical profile
N = 250
N = 111
N = 139
Total serum Cholesterol
WC (central obesity)
≥90 cm (men)
≥80 cm (women)
Obesity class I 25.0–29.9
Obesity class II 30.0>
Comparison of mean intakes of men and women with the WHO Population nutrient intake goals for prevention of death and disability from NCDs 
Mean reported intakes
N = 111
N = 139
Total fat %E
SAT FAT %E
n-6 PUFAs %E
n-3 PUFAs %E
Trans fatty acids %E
Total carbohydrate %E
Free sugars %E
Fruit and vegetables g/day
Dietary fibre g/day
Mean (SD) and median (25th, 75th) dietary quality indices scores
Total group N = 250
Men N = 111
Women N = 139
25th, 75th percentile
25th, 75th percentile
25th, 75th percentile
Pearson partial correlations* controlled for age and smoking between risk scores, clinical factors and diet index scores
bRisk Score 1
N = 250
Men N = 111
Women N = 139
Factor loadings for the first two principal components identified from the percentage total fat contributed by each food group
Factor 1: Meat and fish versus legume and cereal pattern
Factor 2: Nuts and seeds versus sugars and fat pattern
Cereals, cereal products
Nuts and seed
Full-fat milk, milk products
Low-fat milk, milk products
Meat and meat products
Fish and seafood
Visible fats and oils
Sugar and sweets
Median values of risk markers according to the 3 groups of quintiles for factor 1 and factor 2 scores
Factor 1: Meat and fish versus legume and cereal pattern
Factor 2: Nuts and seeds versus sugars and fat pattern
P for trend1
P for trend1
Blood glucose (mmol/L)
Blood Cholesterol (mmol/L)
Blood Triglyceride (mmol/L)
The role of diet in the aetiology of most NCDs is extremely important. A plethora of scientific publications has revealed – through ecological and epidemiological studies, interventions with specific nutrients and foods in placebo-controlled trials and through molecular and genetic research – that there is a consistent relationship between unhealthy diets and the emergence of NCDs , making nutrition a major modifiable determinant of NCDs. The investigation of diet quality and patterns in relationship to the high prevalence of NCDs in this population paves the way for ethno-sensitive prudent intervention.
Using Asian cut-offs, the assessment of anthropometric characteristics and clinical profile led to the identification of a number of risk markers for NCDs, which supports the current high prevalence of NCDs in the Indian community [2, 3]. Although HDL-C levels were not measured, the results of the anthropometric and clinical measurements suggest that the participants displayed characteristics typical of the metabolic syndrome that is common among Asian Indians .
The comparison of food and nutrient mean intakes with the WHO population nutrient goals for the prevention of death and disability from NCDs  showed that the mean percentage energy from total fat, free sugars and n-6 fatty acids were above the recommended goals. Similar percentages of energy from total fat have been reported from Indians living in the metropolitan area of Durban,  and from Asian Indian migrant studies, [34, 35] reaffirming the nutrition transition . We found that the mean percentage energy from n-3 fatty acids and from carbohydrate, and the mean daily intakes of fruit and vegetables and fibre were below the recommended WHO goals. The mean fruit and vegetable intake of the total sample in the current study was only slightly over half the recommended WHO intake . A study on the association of fruit and vegetable intake with CAD in urban South Indians reported slightly higher mean intakes (260 g/day and 270 g/day for men and women respectively) . Although Indians in South Africa consume a variety of fruit and vegetables in their diet, the apparent low intake of fruit and vegetables could be because vegetables are often added to a meat dish as an extender and for this reason, are classified under the meat group, masking the actual intake of vegetables.
In agreement with Wolmarans et al. , our results showed PUFAs provided more than 10 % energy. In this study, the mean percentage energy derived from n-6 fatty acids (11.7 % for men and 10.3 % for women) exceeded the recommended WHO goal of 5 to 8 %. In contrast, the percentage energy derived from n-3 fatty acids (0.25 % for men and 0.23 % for women) was lower than the WHO recommendation of 1 to 2 %, reflecting the Western diet trend. Expressing the intake of PUFAs as a mass is considered a better approach to n-6 and n-3 fatty acid balance than a ratio . The United Kingdom (UK) Food Standards Agency Workshop  recommended that the use of the ratio to estimate CVD risk should be abandoned. South Asians have been reported to consume significantly less n-3 fatty acids (0.08 %E versus 0.13 %E) but more n-6 fatty acids (5.4 %E versus 5.0 %E) than Caucasians in the UK . The n-3 fatty acid intake was possibly low in the current study because most sources rich in n-3 fatty acids do not feature in the South African Indian diet for several reasons. Oily fish intake is largely restricted because of lack of availability and high cost. Sunflower oil is the preferred cooking oil in terms of cost, taste and availability.
The percentage energy from carbohydrates (47.2 % for men and 45.4 % for women) was below the 55–75 % range as recommended by the WHO goals . The reported range for percentage energy from carbohydrates by Wolmarans et al. , on the dietary intake of Indians living in Durban was similar to our study for men and slightly higher for women. In contrast to the lower percentage energy from carbohydrates, the sugar intake in the current study (12.5 E% for both men and women) was higher than the WHO goals, mirroring similar results by Wolmarans et al.  with effects noted in the high prevalence of diabetes in this community .
With reference to the low fibre intake in this study, it is well recognised that low glycaemic index diets are associated with improved weight loss, decreased fasting glucose and triglyceride levels and lowering of blood pressure . Increased intake of legumes and pulses may contribute to lowering the glycaemic index of the South African Indian diet provided that the cooking methods are appropriate. The American Dietetic Association acknowledges that the consumption of viscous dietary fibre lowers blood cholesterol levels and helps to normalise blood glucose and insulin levels and therefore, is ideal for the treatment and prevention of obesity . Refined carbohydrates such as rice, bread and roti are staple foods for Indians in South Africa and may predispose individuals to atherogenic dyslipidaemia including hyper-triglyceridaemia . In light of this, it is recommended that refined carbohydrates be reduced to achieve optimum triglyceride levels in our study population.
The percentage energy from protein (12.8 % for men and 12.0 % for women) fell within the range of WHO goals. A possible reason explaining the prudent protein intake could be that animal protein intake is limited to certain days of the week because of religious worship, in the form of abstinence from meat. On days of abstinence from meat, the protein intake is balanced by the consumption of legumes (11 %E).
Diet quality and patterns can be used to assess associations between clinical markers and dietary behaviours. The deficient index (mean 2260 out of a possible 3000) and excess index (mean 518 out of 600) calculated in our study reflected moderate to good diet quality. The higher the risk score and risk score 1, the greater the risk towards CAD whereas a high index indicates a better quality of diet. Thus, an inverse correlation could be expected between the values of the risk markers and diet index. The Pearson partial correlation coefficients between the deficient index and risk markers were weak, with only the coefficients for glucose for the whole sample, men and women and waist circumference for the men giving significant positive correlations. The deficient index reflects the adequacy of micronutrients, fibre, protein, carbohydrate and essential fatty acid intakes in comparison to the Recommended Dietary Allowances . Thus, a diet with a high deficiency index score, while of good quality in terms of nutrient adequacy may have characteristics associated with dietary risk of NCDs as seen by the positive correlation with waist circumference and glucose for men (Table 4). The excess index is a measure of how closely the diet meets the prudent diet recommendations and was inversely correlated with WC for the whole sample and for men and with glucose for men, which could be expected: the better the quality of the diet the lower the waist circumference and blood glucose levels.
Two factors were identified from the percentage fat contributed by the food groups. The meat and fish versus legumes and cereal pattern contrasted in Factor 1 in two ways: one in which fat intake from meat, meat products, fish and shellfish was higher than average, and one in which the main contributors to fat intake were legumes, cereal and cereal products and vegetables. The significant decreasing trends in the medians of WC and blood glucose, cholesterol and triglycerides from Quintile 1 to Quintile 5 suggest that a dietary pattern in which a higher proportion of fat is derived from the cereal, legume and vegetable food groups is associated with lower levels of these risk markers. However, since the Quintile 5 medians for WC and blood triglycerides remained above the risk cut-off values, whilst those for blood glucose and cholesterol were only slightly lower than the cut-off values, and the remaining risk markers (SBP, DBP and BMI) were unchanged across quintiles, the cereal and legume pattern cannot be considered as protective against NCD risk. The high loadings for the cereal, vegetable and legume groups obtained in our study can be explained by the traditional food preparation practices of South African Indian cuisine. The fat content in cereals and cereal products and in legume and vegetable dishes is derived from “added” fat, used during preparation and cooking of these foods. For example, onions and spices may be tempered in butter or ghee and tossed into boiled rice or legumes; savoury pastries are fried in sunflower oil; and sweets made from cereals are cooked in a ghee base.
Dietary patterns vary by ethnicity and may change to a new environment . In an Indian Health study across three regions in India, the diet was characterized by dairy, fried snacks and sweets which were positively associated to abdominal obesity . It was also found that there was a continued preference for Indian sweets prepared with sugar and substantial amounts of saturated fat from ghee and butter . Furthermore, the cooking method of vegetables might alter some of their preventative properties due to added fat .
The nuts and seeds versus sugars and fats pattern reflects a food pattern with high intakes of inherent fats either in the form of nuts and seeds or as discretionary fat added to foods or in combination with sugars and sweets. Generally, studies examining the dietary patterns frequently choose to name the fruit and vegetable and whole grain loading, the “healthy/ prudent” pattern  whereas, dietary patterns characterized by higher intake of fat, meat and refined grain are referred to as the “Western pattern” . The dietary patterns derived from the percentage fat supplied by food groups in our study suggest that a high legume, cereal and cereal products and vegetable dietary pattern may not be a healthier’ dietary pattern versus the meat the meat and fish pattern for Indians in KZN, and could be identified as a “added” fat pattern. The nuts and seeds versus sugar and fat pattern may be the “healthier” pattern in this community as fat content comes from obvious sources, which may be easier to control. As pointed out by several authors the interpretation the results of factor analyses must be done with caution, as factor analyses may be influenced by several subjective decisions, [28, 47, 48] including: whether foods are classified into food groups, the food groups’ used and how foods are classified into food groups, the dietary component analysed, and the selection of factors included in the analyses.
The overwhelming burden of the nutrition transition in the Indian community is characterised by a shortfall of fruit and vegetable, fibre and n-3 fatty acid intake and the traditional use of added fat in food preparation. We based the factor analysis on the percentage fat intake derived from each food group, as fat intake was the primary focus of our study. Use of percentage energy intake, and/or different food groupings may have resulted in the emergence of different food patterns. Further research is needed for a better understanding of the diet as modifiable risk factor. However, the assessment of diet quality and patterns in this populace, allows for the translation of dietary goals into practical recommendations such as increasing fruit and vegetable, fibre and n-3 fatty acid intake, changing cooking methods, and limiting the use of added fats for a healthier lifestyle.
We thank Barbara English for her meticulous editing. This study was supported by grants from the South African Medical Research Council, National Research Foundation; Thuthuka and North-West University.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Drewnowski A, Popkin BM. The nutrition transition: new trends in the global diet. Nutr Rev. 1997;55:31–43.PubMedView ArticleGoogle Scholar
- Vorster HH, Gibney MJ. Food and nutrition-related diseases: the global challenge. In: Gibney MJ, Lanham SA, Cassidy A, Vorster HH, editors. Introduction to Human Nutrition. 2dth ed. South Africa, Oxford: Blackwell Publishing Ltd; 2009. p. 350–9.Google Scholar
- Norman R, Bradshaw D, Schneider M, Joubert J, Groenwald P, Lewin S, et al. A comparative risk assessment for South Africa in 2000: towards promoting health and preventing disease. S Afr Med J. 2007;97:637–41.PubMedGoogle Scholar
- Bradshaw D, Norman R, Pieterse D, Levitt NS. Estimating the burden of disease attributable to diabetes in South Africa in 2000. S Afr Med J. 2007;97:700–6.PubMedGoogle Scholar
- Ranjith N, Pegoraro RJ, Naidoo DP. Demographic data and outcome of acute coronary syndrome in the South African Asian Indian population. Cardiovasc J S Afr. 2005;16:48–54.PubMedGoogle Scholar
- Seedat YK. Diabetes mellitus in South African Indians. In: Mohan V, Rao GHD, editors. Type 2 Diabetes in South Asians: Epidemiology, Risk factors and Prevention. New Delhi: Jaypee Brothers; 2007. p. 40–4.Google Scholar
- Bradshaw D, Steyn K, Levitt N, Nojilana B. Non-Communicable Diseases- A race against time. Parow Valley: Medical Research Council South Africa; 2014. p. 1–4.Google Scholar
- Mertz W. Foods and nutrients. J Am Diet Assoc. 1984;84:769–70.PubMedGoogle Scholar
- Kant AK. Indexes of overall diet quality: a review. J Am Diet Assoc. 1996;96:785–91.PubMedView ArticleGoogle Scholar
- Kennedy EF, Ohls J, Carlson S, Fleming K. The healthy eating index: design and application. J Am Diet Assoc. 1996;95:1103–8.View ArticleGoogle Scholar
- Schwerin HS, Stanton JL, Smith JL, Riley AM, Brett BE. Food, eating habits, and health: a further examination of the relationship between food eating patterns and nutritional health. Am J Clin Nutr. 1982;35:1319–25.PubMedGoogle Scholar
- Kant AK. Dietary patterns and health outcomes. J Am Diet Assoc. 2004;104:615–35.PubMedView ArticleGoogle Scholar
- World Health Organization. Joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases, WHO Technical Report Series no 916. Geneva: World Health Organization; 2003.Google Scholar
- Statistics SA. Census 2011 Statistical release. Pretoria: Statistics South Africa; 2012. p. 17.Google Scholar
- MacIntyre UE, Venter CS, Vorster HH. A culture-sensitive quantitative food frequency questionnaire used in an African population: Relative validation by 7-day weighed records and biomarkers. Public Health Nutr. 2001;41:63–71.Google Scholar
- Wu HS. Global Trade Patterns in Fruits and Vegetables. Agriculture and trade report WRS-04-06. 2004.Google Scholar
- Messer E. Intra-household allocation of food and health care; current findings and understandings – introduction. Soc Sci Med. 1997;44:1675–84.PubMedView ArticleGoogle Scholar
- Steyn NP, Senekal M. A guide to use the Dietary Assessment Education Kit. Parow Valley: Medical Research Council: South Africa; 2002.Google Scholar
- Lee RD, Nieman DC. Nutritional Assessment. New York: McGraw-Hill; 2002.Google Scholar
- Forouhi NG, Sattar N, Tillen T, Mckeigue PM, Chaturvedi N. Do known risk factors explain the higher coronary heart disease mortality in South Indian compared with European men? Prospective follow-up of the Southall and Brent studies, UK. Diabetologia. 2006;49:2580–8.PubMedView ArticleGoogle Scholar
- WHO/IASO/IOTF. The Asia-Pacific perspective: redefining obesity and its treatment. Sydney: Health Communications Australia Pty Ltd; 2000.Google Scholar
- Medical Research Council. FoodFinder 3 version 1.1.3. Parow Valley: Medical Research Council South Africa; 2002.Google Scholar
- Gopalan C, Ramasastri BV, Balasubramanian SC. Nutritive value of Indian foods. National Institute of Nutrition. Hyderabad: Indian Council of Medical Research India; 2004.Google Scholar
- Thiele S, Mensink GBM, Beitz R. Determinants of diet quality. Public Health Nutr. 2004;7:29–37.PubMedView ArticleGoogle Scholar
- Institute of Medicine. Dietary Reference Intakes. Food and Nutrition Board. Washington DC: National Academy Press; 2003.Google Scholar
- Wolmarans P, Danster N, Dalton A, Rossouw K, Schonfeldt H, editors. Condensed food composition tables for South Africa. Parow Valley: Medical Research Council South Africa; 2010.Google Scholar
- Statistical Package for the Social Sciences for Windows. Version 17.0. Chicago: SPSS Inc; 2008.Google Scholar
- Newby PK, Tucker KL. Empirically derived eating patterns using factor or cluster analysis: a review. Nutr Rev. 2004;62:177–203.PubMedView ArticleGoogle Scholar
- Karamati M, Jessri M, Shariati-Bafghi SE, Rashidkhani B. Dietary patterns in relation to bone mineral density among menopausal Iranian women. Calcif Tissue Int. 2012;91:40–9.PubMedView ArticleGoogle Scholar
- Bewick V, Cheek V, Ball J. Statistics review 10: Further nonparametric methods. Crit Care. 2004;8:196–9.PubMed CentralPubMedView ArticleGoogle Scholar
- National Cholesterol Education Program (NCEP) Expert Panel on Detection. Evaluation and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143–421.Google Scholar
- Das UN. Metabolic syndrome X is common in Indians: but why and how? J Assoc Physicians India. 2003;51:987–98.PubMedGoogle Scholar
- Wolmarans P, Seedat YK, Mayet FG, Joubert G, Wentzel E. Dietary intake of Indians living in the metropolitan area of Durban. Public Health Nutr. 1999;2:55–60.PubMedView ArticleGoogle Scholar
- Yagalla MV, Hoerr L, Song WO, Enas A, Garg A. Relationship of diet, abdominal obesity, and physical activity to plasma lipoprotein levels in Asian Indian physicians residing in the United States. J Am Diet Assoc. 1996;96:257–61.PubMedView ArticleGoogle Scholar
- Lovegrove JA, Lovegrove SS, Lesauvage SVM, Brady LM, Saini N, Minihane AM. Moderate fish-oil supplementation reverses low-platelet, long-chain n-3 polyunsaturated fatty acid status and reduces plasma triacylglycerol concentrations in British Indo-Asians. Am J Clin Nutr. 2004;79:974–82.PubMedGoogle Scholar
- Radhika G, Sudha V, Sathya RM, Ganesan A, Mohan V. Association of fruit and vegetable intake with cardiovascular risk factors in urban South Indians. Br J Nutr. 2008;99:398–405.PubMedView ArticleGoogle Scholar
- Wijendran V, Hayes KC. Dietary n-6 and n-3 fatty acid balance and cardiovascular health. Ann Rev Nutr. 2004;24:597–615.View ArticleGoogle Scholar
- Stanley JC, Elsom RL, Calder PC, Griffin BA, Harris WS, Jebb SA. UK Food Standards Agency Workshop Report: the effects of the dietary n-6:n-3 fatty acid ratio on cardiovascular health. Br J Nutr. 2007;98:1305–10.PubMed CentralPubMedView ArticleGoogle Scholar
- Radulian G, Rusu E, Dragomir A, Stoina M, Vladica M. Metabolic effects of low glycaemic diets. Nutr J. 2009;8:5–13.PubMed CentralPubMedView ArticleGoogle Scholar
- American Dietetic Association. Position of the ADA: Health implications of dietary fibre. J Am Diet Assoc. 2002;102:993–1000.View ArticleGoogle Scholar
- Goyal A, Yusuf S. The burden of cardiovascular diseases in the Indian subcontinent. Indian J Med Res. 2006;124:235–44.PubMedGoogle Scholar
- Dl R, Gasevic D, Seidell J, Lear SA. Determinants of changes in dietary patterns among Chinese immigrants: A cross sectional analysis. Int J Behav Nutr Phys Act. 2011;42:1–8.Google Scholar
- Daniel CR, Prabhakaran D, Kapur K, Grabard BI, Devasenapathy N, Ramakrishnan L, et al. A cross-sectional investigation of regional patterns of diet and cardio-metabolic risk in India. Nutr J. 2011;10:1–13.View ArticleGoogle Scholar
- Shi Z, Hu X, Yuan B, Hu G, Pan X, Dai Y, et al. Vegetable-rich food pattern is related to obesity in China. Int J Obes. 2008;32(6):975–84.View ArticleGoogle Scholar
- Newby PK, Muller D, Tucker KL. Association of empirically derived eating patterns with plasma lipid biomarkers:a comparison of factor and cluster analysis methods. Am J Clin Nutr. 2004;80:759–67.PubMedGoogle Scholar
- Hu FB. Dietary patterns analysis: a new direction in nutritional epidiomology. Curr Opin Lipidol. 2002;13(1):3–9.PubMedView ArticleGoogle Scholar
- Roman-Vinas B, Barba LR, Ngo J, Martinez-Gonzalez MA, Wijnhoven TMA, Serra-Majem L. Validity of dietary patterns to assess nutrient intake adequacy. Br J Nutr. 2009;101(Suppl):S12–20.PubMedView ArticleGoogle Scholar
- Naja F, Nasreddine L, Itani L, Chameih MC, Adra N, Sibai AM. Dietary patterns and their association with obesity and sociodemographic factors in a national sample of Lebanese adults. Public Health Nutr. 2011;14:1570–8.PubMedView ArticleGoogle Scholar