In high income countries, low socioeconomic status (SES) is related to unhealthier dietary patterns, while evidence on the social patterning of diet in low and middle income countries is scarce.
In this study, we assess dietary patterns in the general population of a middle income country in the African region, the Republic of Seychelles, and examine their distribution according to educational level and income.
Data was drawn from two independent national surveys conducted in the Seychelles among adults aged 25–64 years in 2004 (n = 1236) and 2013 (n = 1240). Dietary patterns were assessed by principal component analysis (PCA). Educational level and income were used as SES indicators. Data from both surveys were combined as no interaction was found between SES and year.
Three dietary patterns were identified: “snacks and drinks”, “fruit and vegetables” and “fish and rice”. No significant associations were found between SES and the “snacks and drinks” pattern. Low vs. high SES individuals had lower adherence to the “fruit and vegetables” pattern [prevalence ratio (95% CI) 0.71 (0.60–0.83)] but a higher adherence to the traditional “fish and rice” pattern [1.58 (1.32–1.88)]. Income modified the association between education and the “fish and rice” pattern (p = 0.02), whereby low income individuals had a higher adherence to this pattern in both educational groups.
Citation: Mayén A-L, Bovet P, Marti-Soler H, Viswanathan B, Gedeon J, Paccaud F, et al. (2016) Socioeconomic Differences in Dietary Patterns in an East African Country: Evidence from the Republic of Seychelles. PLoS ONE 11(5): e0155617. https://doi.org/10.1371/journal.pone.0155617
Editor: Elizabeth W. Triche, St Francis Hospital, UNITED STATES
Received: September 23, 2015; Accepted: May 2, 2016; Published: May 23, 2016
Copyright: © 2016 Mayén et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data were obtained from health surveys conducted in the Seychelles Ministry of Health in collaboration with the University Institute of Social and Preventive Medicine in Lausanne, Switzerland. The data may be requested from Dr Jude Gedeon, Commissioner of Public Health, Ministry of Health, Victoria, Republic of Seychelles, Jude.Gedeon@health.gov.sc.
Funding: Ana-Lucia Mayén is supported by a Swiss Excellence Government scholarship awarded by the Swiss Confederation. Silvia Stringhini is supported by an Ambizione Grant (n° PZ00P3_147998) from the Swiss National Science Foundation (SNSF).
Competing interests: The authors have declared that no competing interests exist.
In high income countries (HICs), socioeconomic status (SES) is a major determinant of healthy eating: individuals with a high SES tend to have a higher intake of whole grains, lean meat and fish, low-fat dairy products, fresh vegetables and fruit , while individuals with a low SES have a higher intake of refined grains and fatty foods [2–4]. A western-like diet, low in fiber and rich in fat, salt and sugar, has been associated with an increased body mass index (BMI) and non-communicable diseases (NCD) development . Thus, the higher prevalence of NCDs in low SES individuals may be partly explained by unhealthier dietary patterns in this group. The effect of SES on diet may be related to several factors. For example, low SES individuals are more exposed to adverse psychosocial conditions such as stress at home or work settings, and lack social networks and support [6, 7], which may lead to the adoption of unhealthy behaviours (e.g. smoking, high fat diets, heavy drinking) [8–11]. Low SES individuals may also have less knowledge about healthy behaviours, health risks [8, 12], and be less willing to invest in future health . Additionally, low SES individuals may have a lower access to more expensive healthy foods because of financial and environmental constraints [14–16].
In low and middle income countries (LMICs), shifts in physical activity, and the structure and composition of diet are occurring along with demographic and socioeconomic changes. Fiber-rich foods are being substituted by foods high in fat and sugar while physical activity levels are decreasing. These changes are defined as the nutrition transition theory [17, 18]. Further, evidence on social differences in diet in LMICs is scarce. The few studies that have examined the social patterning of dietary intake have generally found that wealthier populations have an unhealthier diet, higher in cholesterol and saturated fats and lower in fiber . Possible reasons include a higher access to processed foods available in supermarkets and the adoption of “western” dietary patterns by the upper classes [17, 19]. In spite of this, there is evidence to suggest that the social patterning of diet may reverse over time along with socioeconomic development, with poverty-stricken populations adopting unhealthier diets at later stages of the nutrition transition [20, 21]. This reversal has been shown before for obesity  and cardiovascular risk factors .
Almost three quarters of worldwide NCD-related deaths occur in LMICs . As diet is one of the major modifiable risk factors for NCDs, several aspects of diet are included in the nine targets of the World Health Organization Global Action Plan for the Prevention and Control of NCDs 2013–2020 to reduce NCD mortality by 25% by 2025 . Evidence on the social distribution of dietary patterns is essential for an adequate implementation of this action plan.
Only few studies have assessed the social distribution of diet in African countries [25–30], and some were conducted among high SES individuals only [31, 32]. These studies have generally shown that individuals with high SES have a higher intake of fruit and vegetables  and a lower intake of traditional foods compared to individuals with low SES [25, 26]. Few studies have assessed diet in Seychelles and they show an increasing intake of calories, salt and vegetable oils , plus a decreasing intake of traditional foods . However, no study has assessed the social distribution of dietary patterns in Seychelles. In this study, we assessed dietary patterns in the general population of the Republic of Seychelles, a higher middle income country in the African region, and examine their distribution according to educational level and income. We hypothesize that, in the Seychelles, individuals with lower SES have an unhealthier western-like diet compared to those with higher SES, as observed in HICs.
Materials and Methods
The Republic of Seychelles is a country located in the Indian Ocean with a total population of 84,000 in 2004 and 93,000 in 2013 . The large majority of inhabitants are of African descent with minorities from Caucasian, Indian or Chinese descent . While based only 50 years ago on the exportation of coprah, cinnamon and other cash crops, Seychelles rising economy currently relies on tourism, industrial fishing, and services . The gross national income (GNI) per capita increased from 2,000 US$ in the 1980s to more than 13,000 US$ in 2014 . Since more than a decade ago, nearly all children have attended at least 10 years of free and compulsory education (kindergarten, primary and secondary) up to the age of 15–16 years . At least one third of the population attends post secondary vocational or academic education. The unemployment rate for men and women has been low for several years, e.g. <5% in 2014 . A high prevalence of cardiovascular risk factors (CVRFs) has been previously described, including a high and increasing prevalence of overweight and obesity .
Two cross sectional national surveys of CVRFs were conducted in 2004 (n = 1236) and 2013 (n = 1240), including participants aged 25–64 years. The sampling methods have been described previously [41, 42]. Briefly, for each survey, eligible individuals were selected from a database derived from population censuses, regularly updated by civil authorities. Random samples were drawn from the entire population. Eligible participants were invited to attend the survey centers and the ones who did not attend were traced and re-invited to attend. The participation rate in 2004 was 80.3% and in 2013 was 72.9%. As associations between SES and dietary patterns were similar in the two surveys (p for interaction between SES and year >0.05) and the sample size was rather small, we pooled data from the two surveys. The surveys were approved by the research and ethical board of the Ministry of Health of the Republic of Seychelles and all participants provided an informed written consent.
Socioeconomic status (SES).
We used two indicators of SES in this study: education and income. Education refers to the highest level of education attained. The variable was categorized in two groups: “high” (post secondary school: polytechnic school, university or similar schools, excluding post secondary vocational school) and “low” (secondary school and post secondary vocational school). Income refers to average personal earnings per month in the Seychelles Rupees. Income substantially differed between 2004 and 2013, partly because of an economic reform. Thus we harmonized the variable by creating two categories: low income (≤3,000 rupees in 2004 and ≤8,000 in 2013) and high income (>3,000 rupees in 2004 and >8,000 rupees in 2013) to have similar proportions of the population in each group for both years.
Food frequency questionnaires (FFQs) were used to assess dietary intake. These were prepared in collaboration with the Nutrition Unit of the Ministry of Health in Seychelles, taking into account local dietary habits. The questionnaires assessed the number of days per week certain food items were consumed (e.g. fruit, rice, salad, fish, chicken, etc) [43, 44]. Participants were asked about their retrospective weekly intake of selected food items and those with missing values for at least one food group (e.g. vegetables) were excluded from the analytical sample (n = 9 in 2004 and n = 0 in 2013). The FFQs included between 25 and 33 foods for both surveys and were grouped into 18 foods, which appear in S1 Table. As the FFQs for both years were slightly different, categories for fruit, fish, meat, drinks and juice were harmonized before merging dietary intake data.
Analyses were conducted using Stata version 13.1 (Stata Corp., College Station, TX, USA). Dietary patterns were obtained using principal component analysis (PCA) with varimax orthogonal rotation  to minimize correlation between components . This is a widely used statistical method to identify dietary patterns  and we expect a variance similar to other studies to explain variability of the data [48–50]. A Kaiser-Meyer-Olking (KMO) test was performed to assess the adequacy of the data for PCA and all food items with a value >0.5 were included. Two criteria were used to retain dietary patterns after running the PCA [51, 52]: an eigenvalue higher than one and the interpretability of each dietary pattern. Food items with an absolute factor loading greater than 0.2 were considered to characterize the dietary pattern. Dietary patterns were named after the food items with the highest factor loadings (i.e. those with the highest correlation to the dietary patterns). A dietary score was assigned to each participant for each of the three dietary patterns, to indicate the accordance between the participant’s usual intake and each food pattern. Scores were calculated using the “PCA” and “predict” commands in Stata. These commands create linear composites by standardizing variables to a zero mean and unit variance. Then factor score coefficients are weighted and summed for each factor . Dietary patterns were dichotomized as high adherence (being in the highest quartile of diet scores for each pattern) vs. low adherence (lower three quartiles) .
Analysis were performed to assess if there was a modifying effect by gender in the association between SES indicators and the dietary patterns. As there was no evidence of such an effect (p for interaction ≥0.05), analyses were performed for men and women together and adjusted by sex.
As the outcomes of interest (being in the highest quartile of each dietary pattern) were not a rare event (prevalence higher than 10%), the associations between SES and dietary patterns were assessed using Poisson regression models to avoid overestimation of associations [48, 54]. Four different models were built: 1) adjusted for age, sex, year and education; 2) adjusted for age, sex, year and income; 3) adjusted for age, sex, year, education and income and 4) same as Model 3 and including an interaction term for education and income. The latter model was aimed at assessing the modifying effect of income on the association between dietary patterns and education.
Participants with missing values on the variables used in this study were excluded (S1 Fig). Only 19 participants were excluded from the original sample due to missing information on income or dietary data. Analyses were based on 2476 participants, 56% of whom were women, 72% had low education and 69% had low income (S2 Table).
Three dietary patterns were identified: “snacks and drinks” (composed mainly of soft and energy drinks, salty and sweet snacks, and fruit juice), “fruit and vegetables” (composed mainly of fruit, vegetables and salad), and “fish and rice”. The total variance explained by the three patterns was 29% (Table 1).
Population characteristics by adherence to patterns
The baseline characteristics by adherence to the three defined dietary patterns are shown in Table 2. Men had a higher adherence to the “snacks and drinks” and “fish and rice” patterns compared to women, but a lower adherence to the “fruit and vegetables” pattern (p<0.001). Individuals with high SES had a higher adherence to the “snacks and drinks” and “fruit and vegetables” patterns, but a lower adherence to the “fish and rice” pattern (p<0.001). Older participants had a lower adherence to the “snacks and drinks” pattern, but a higher adherence to the “fish and rice” pattern.
Associations of socioeconomic indicators with dietary patterns
Age and sex-adjusted associations of SES indicators with the dietary patterns are shown in Table 3. As for the “snacks and drinks” pattern, education was not related to adherence to this pattern while individuals with low income tended to have a lower adherence, although associations were not significant [prevalence ratio (PR) 0.93 (0.81–1.07) for low vs. high income]. No evidence was found for an interaction between education and income. Older individuals and women had a lower adherence to this pattern. When adjusting for age, sex, education and/or income, individuals assessed in the 2013 survey had a higher adherence to this pattern compared to those assessed in the 2004 survey.
Model 1: adjusted for age, sex, year and education; Model 2: adjusted for age, sex, year and income; Model 3: adjusted for age, sex, year, education and income; Model 4: same as Model 3 and including an interaction term for education and income. Low education: secondary (obligatory), post secondary vocational or lower. High education: polytechnic and university. High income defined as income ≥3,001 Rupees in 2004 and ≥8,001 Rupees in 2013. Significant associations are indicated in bold.
Regarding the “fruit and vegetables” pattern, individuals with low SES had a lower adherence to this pattern in age and sex-adjusted analyses [PR 0.75 (0.64–0.87) for low vs. high education, and 0.67 (0.58–0.77) for low vs. high income]. The association between education and the “fruit and vegetables” pattern was partly attenuated after adjustment for income and became not significant. Income did not modify the association between education and adherence to the “fruit and vegetables” pattern (p for interaction = 0.25). Older individuals and women had a higher adherence to this pattern.
As to the “fish and rice” pattern, individuals with low SES had a higher adherence to the “fish and rice” pattern [PR 1.56 (1.29–1.90) for low vs. high education and 1.71 (1.45–2.01) for low vs. high income]. Associations remained after mutual adjustment for education and income. Income modified the association between the “fish and rice” pattern and education (p for interaction = 0.02), with low income individuals having a higher adherence to the pattern independently of education. Older individuals had a higher adherence to this pattern while women had a lower adherence. When adjusting for age, sex, education and/or income, individuals assessed in the 2013 survey had a lower adherence to this pattern compared to those assessed in the 2004 survey.
This study assessed dietary patterns and their social distribution in an upper middle income country of the African region. Individuals with low SES had a lower adherence to the “fruit and vegetables” pattern but were more likely to consume a traditional diet (“fish and rice” pattern) than their more advantaged counterparts. Moreover, low income individuals had a higher adherence to the “fish and rice” pattern across all education levels. Although associations were not significant, lower income tended to be related to a lower adherence to a western-like pattern characterized by “snacks and drinks”.
The patterns identified in this study were similar to those generally found in other low and middle income settings. For example, the “snacks and drinks” pattern shared several items with the “processed food” and “snacking” patterns identified in Brazil and Africa. These included salty snacks, soda, beer, processed foods , fried foods and sweetened products . The “fruit and vegetables” pattern is similar to the “healthy pattern” found in two Iranian studies [56, 57]. Finally, a “balanced” pattern identified in Bangladesh included the same components as the “fruit and vegetables” and “fish and rice” patterns . Our findings support the view that high SES individuals have a lower adherence to “traditional-like” diets, possibly related to the rapid economic development experienced by the Seychelles . We did not assess changes of dietary patterns over time in our population. However, our results suggest that the Seychelles is experiencing the nutrition transition and the population tends to substitute fiber-rich traditional foods with high-fat diets. This is confirmed by data on obesity over time in Seychelles .
Associations of socioeconomic indicators with dietary patterns
The low SES group had a lower adherence to the “fruit and vegetables” pattern, a common finding in both LMICs  and HICs . This may be explained by a lack of knowledge regarding the benefits of fruit and vegetable consumption, their high price, and their limited accessiblility and availability  due to low local production  and large reliance on imported produce.
In this study, low education and low income individuals had a higher adherence to the “fish and rice” pattern, which included two of the traditional staple foods in Seychelles. This result is in line with previous studies conducted in HICs [62, 63] and LMICs [48, 64], where low SES individuals adhere more to a traditional-like diet. This may be partly explained by the fact that high SES individuals tend to choose modern western-like “trendy” foods as a matter of prestige , while low SES individuals have lower access to expensive imported foods (other than essential subsidized food items). Thus, they tend to maintain their traditional diet, which generally includes low-cost staple foods, widely consumed and often subsidized to fulfill the minimum caloric needs of a household [65, 66].
In addition, our results suggest that income plays a major role in the substitution of traditional foods with new foods in all educated groups. This result is consistent with previous studies showing that increasing income is related to a substitution of traditional foods with animal-based high fat diets .
In parallel with a higher adherence to a traditional pattern, our study suggests that the low SES group had a lower adherence to the “snacks and drinks” pattern. Although non-significant, possibly due to a lack of statistical power, this result may indicate that the Seychelles is at a degenerative diseases stage of the nutrition transition, when foods high in added sugars, fats and soft drinks are more likely to be consumed by the most advtanged sections of society [17, 20, 68]. This is in contrast with our hypothesis that low SES groups would have shown a higher adherence to a western-like diet, due to the rapid economic development of the country. Our results thus confirm that the nutrition transition and the shift of unhealthy diets towards the low SES group occurs at a different pace in different countries. Moreover, our results are in line with the higher prevalence of overweight and obesity among high SES men in the Seychelles [59, 69], although not with the higher prevalence of obesity among low SES women. A possible explanation is that low SES women have a high access to traditional cakes, samoosa, banana or breadfruit chips while high SES individuals have a higher access to “modern” sweet snacks like chocolate bars and chips.
In our study, education and income seem to influence dietary patterns independently, as associations were only modestly attenuated by mutual adjustment. This has been reported previously [70, 71]. Moreover, individuals with a low income tended to stick to a traditional diet irrespective of their educational level, probably because of its lower cost, and it may be considered as a fair option to fulfill the daily caloric requirement although it may lack a number of nutrients (e.g. fiber).
A previous study showed that Seychelles is experiencing a nutrition transition characterized by a decreased consumption of traditional staple foods (fish, rice), beverages (tea), inexpensive home brews, and by an increased intake of meat, poultry and snacks . Our results suggest that this pattern is mostly affecting high SES individuals. However, as experienced in HICs and for other CVRFs in LMICs [72, 73], it is possible that this gradient reverses in the future, thus contributing to increase the burden of NCDs among low SES individuals.
Global efforts for the prevention of NCDs (e.g. the Global Action Plan of the World Health Organization) are proposing prevention strategies aimed at all social strata. However, as our results indicate, the degree of adherence to a “western-like” and “traditional” patterns might not be proportionate in all socioeconomic groups. Moreover, educational interventions alone might not be sufficient, especially in the Seychelles, where income seems to play a more important role to determine dietary consumption.
Strengths and limitations
This study is based on nationally representative data and represents one of the very few studies assessing socioeconomic differences in dietary intake in an African country. However, some limitations must be considered. First, in order to harmonize education and income across the two surveys, we used dichotomized variables thereby losing information on middle SES groups. Second, comparable dietary information was available for a limited number of foods only and this may limit the identification of precise dietary patterns. Still, the patterns identified were in agreement with those observed in other studies conducted in LMICs, so the loss of information might be small. Third, we combined data from 2004 and 2013 even though the nutrition transition may have progressed during this period. However, the same dietary patterns were found when data were analyzed separately for each year. Fourth, foods that were introduced in the more extensive 2013 questionnaire were not taken into account for this analysis as only food items that were common to the two years were examined. However, our study did not aim to assess changes in diet between years. Fifth, FFQs have been widely used in epidemiological studies since more than 2 decades . However, these instruments are prone to reporting biases, and their validity has been questioned regarding portions measurement and precision . As performed by others , we did not assess portion sizes to reduce the burden on participants and limit the lack of precision and number of missing values. We assessed the number of days per week a participant would consume a food item as performed previously , although such categorization (with responses ranging from 0 to 7) should not influence the results obtained by PCA to derive dietary patterns. Finally, the small sample size may have decreased statistical power in some of the analysis.
In summary, our study shows that, in a middle income country, low SES individuals have a lower consumption of fruit and vegetables but a higher consumption of traditional foods such as fish and rice, a fair option to fulfil the daily caloric needs. As the Seychelles seems to be at a degenerative diseases stage of the nutrition transition, it is possible that this pattern may reverse along with socioeconomic development and the progress of the nutrition transition. This underlies the need to develop and implement adequate policies to limit unhealthy foods (e.g. taxes) and promote healthy and locally produced foods (e.g. reformulation of foods items regarding fat, sugar or salt, healthy food programs in schools, etc.).
S1 Fig. Exclusion criteria and final sample included in the study.
S1 Table. Food items included in PCA analysis, by frequency of intake.
Conceived and designed the experiments: SS PB PMV ALM. Analyzed the data: ALM SS HMS. Wrote the paper: ALM. Has primary responsibility for the final content of the manuscript: SS. Data collection: PB BV JG. Critically reviewed the manuscript for important intellectual content and approved the final version: ALM PB HMS BV JG FP PMV SS.
- 1. Darmon N, Drewnowski A. Does social class predict diet quality? The American Journal of Clinical Nutrition. 2008;87(5):1107–17. pmid:18469226
- 2. Smith AM, Baghurst KI. Public health implications of dietary differences between social status and occupational category groups. J Epidemiol Community Health. 1992;46(4):409–16. Epub 1992/08/01. pmid:1431718; PubMed Central PMCID: PMC1059611.
- 3. McCartney DM, Younger KM, Walsh J, O'Neill M, Sheridan C, Kearney JM. Socio-economic differences in food group and nutrient intakes among young women in Ireland. Br J Nutr. 2013;110(11):2084–97. Epub 2013/06/01. pmid:23721781.
- 4. Mishra G, Ball K, Arbuckle J, Crawford D. Dietary patterns of Australian adults and their association with socioeconomic status: results from the 1995 National Nutrition Survey. Eur J Clin Nutr. 2002;56(7):687–93. Epub 2002/06/25. pmid:12080411.
- 5. Azizi F, Ghanbarian A, Momenan AA, Hadaegh F, Mirmiran P, Hedayati M, et al. Prevention of non-communicable disease in a population in nutrition transition: Tehran Lipid and Glucose Study phase II. Trials. 2009;10:5. Epub 2009/01/27. pmid:19166627; PubMed Central PMCID: PMC2656492.
- 6. Siegrist J, Marmot M. Health inequalities and the psychosocial environment-two scientific challenges. Soc Sci Med. 2004;58(8):1463–73. Epub 2004/02/05. pmid:14759690.
- 7. McLoyd VC. The impact of economic hardship on black families and children: psychological distress, parenting, and socioemotional development. Child Dev. 1990;61(2):311–46. Epub 1990/04/01. pmid:2188806.
- 8. Lantz PM, House JS, Mero RP, Williams DR. Stress, life events, and socioeconomic disparities in health: results from the Americans' Changing Lives Study. J Health Soc Behav. 2005;46(3):274–88. Epub 2005/11/02. pmid:16259149.
- 9. Layte R, Whelan CT. Explaining Social Class Inequalities in Smoking: The Role of Education, Self-Efficacy, and Deprivation. European Sociological Review. 2009;25(4):399–410.
- 10. Mayen AL, Marques-Vidal P, Paccaud F, Bovet P, Stringhini S. Socioeconomic determinants of dietary patterns in low- and middle-income countries: a systematic review. The American journal of clinical nutrition. 2014;100(6):1520–31. Epub 2014/11/21. pmid:25411287.
- 11. Hosseinpoor AR, Bergen N, Kunst A, Harper S, Guthold R, Rekve D, et al. Socioeconomic inequalities in risk factors for non communicable diseases in low-income and middle-income countries: results from the World Health Survey. BMC Public Health. 2012;12:912. Epub 2012/10/30. pmid:23102008; PubMed Central PMCID: PMC3507902.
- 12. Link BG. Epidemiological sociology and the social shaping of population health. J Health Soc Behav. 2008;49(4):367–84. Epub 2009/02/03. pmid:19181044.
- 13. Fuchs VR. Time Preference and Health: An Exploratory Study. Chicago: 1982.
- 14. Hawkes C. Uneven dietary development: linking the policies and processes of globalization with the nutrition transition, obesity and diet-related chronic diseases. Global Health. 2006;2:4. Epub 2006/03/30. pmid:16569239; PubMed Central PMCID: PMC1440852.
- 15. Drewnowski A, Darmon N. Food choices and diet costs: an economic analysis. J Nutr. 2005;135(4):900–4. Epub 2005/03/30. pmid:15795456.
- 16. Beaulac J, Kristjansson E, Cummins S. A systematic review of food deserts, 1966–2007. Prev Chronic Dis. 2009;6(3):A105. Epub 2009/06/17. pmid:19527577; PubMed Central PMCID: PMC2722409.
- 17. Popkin BM, Adair LS, Ng SW. NOW AND THEN: The Global Nutrition Transition: The Pandemic of Obesity in Developing Countries. Nutrition Reviews. 2012;70(1):3–21. pmid:PMC3257829.
- 18. Popkin BM. Global nutrition dynamics: the world is shifting rapidly toward a diet linked with noncommunicable diseases. The American journal of clinical nutrition. 2006;84(2):289–98. Epub 2006/08/10. pmid:16895874.
- 19. Popkin BM. Nutrition in transition: the changing global nutrition challenge. Asia Pac J Clin Nutr. 2001;10 Suppl:S13–8. Epub 2001/11/16. pmid:11708577.
- 20. Rogers EM. Diffusion of innovations. 5th ed. New York2003.
- 21. Monteiro CA, Moura EC, Conde WL, Popkin BM. Socioeconomic status and obesity in adult populations of developing countries: a review. Bull World Health Organ. 2004;82(12):940–6. Epub 2005/01/18. /S0042-96862004001200011. pmid:15654409; PubMed Central PMCID: PMC2623095.
- 22. Stringhini S, Bovet P. Commentary: The social transition of cardiovascular disease in low- and middle-income countries: wait and see is not an option. Int J Epidemiol. 2013;42(5):1429–31. Epub 2013/09/07. pmid:24008332.
- 23. World Health Organization. Noncommunicable diseases: Fact sheet 2015. Available from: http://www.who.int/mediacentre/factsheets/fs355/en/.
- 24. World Health Organization. Global Action Plan for the Prevention and Control of NCDs 2013–2020 2014. Available from: http://www.who.int/nmh/events/ncd_action_plan/en/.
- 25. Njelekela MA, Liu E, Mpembeni R, Muhihi A, Mligiliche N, Spiegelman D, et al. Socio-economic status, urbanization, and cardiometabolic risk factors among middle-aged adults in Tanzania. East Afr J Public Health. 2011;8(3):216–23. Epub 2011/09/01. pmid:23120960.
- 26. Sodjinou R, Agueh V, Fayomi B, Delisle H. Dietary patterns of urban adults in Benin: relationship with overall diet quality and socio-demographic characteristics. Eur J Clin Nutr. 2009;63(2):222–8. Epub 2007/09/14. pmid:17851458.
- 27. Pisa PT, Pedro TM, Kahn K, Tollman SM, Pettifor JM, Norris SA. Nutrient patterns and their association with socio-demographic, lifestyle factors and obesity risk in rural South African adolescents. Nutrients. 2015;7(5):3464–82. Epub 2015/05/20. pmid:25984738; PubMed Central PMCID: PMC4446762.
- 28. Landais E, Bour A, Gartner A, McCullough F, Delpeuch F, Holdsworth M. Socio-economic and behavioural determinants of fruit and vegetable intake in Moroccan women. Public Health Nutr. 2015;18(5):809–16. Epub 2014/08/29. pmid:25166305.
- 29. Ag Bendech M, Gerbouin-Rerolle P, Chauliac M, Malvy D. [An approach to food consumption in an urban environment. The case of west Africa]. Sante. 1996;6(3):173–9. Epub 1996/05/01. pmid:8764452.
- 30. Faber M, Laubscher R, Laurie S. Availability of, access to and consumption of fruits and vegetables in a peri-urban area in KwaZulu-Natal, South Africa. Matern Child Nutr. 2013;9(3):409–24. Epub 2011/12/23. pmid:22188599.
- 31. Phaswana-Mafuya N, Peltzer K, Chirinda W, Musekiwa A, Kose Z. Sociodemographic predictors of multiple non-communicable disease risk factors among older adults in South Africa. 2013. 2013;6. Epub 2013-01-02. pmid:24044582
- 32. El-Gilany AH, Elkhawaga G. Socioeconomic determinants of eating pattern of adolescent students in Mansoura, Egypt. Pan Afr Med J. 2012;13:22. Epub 2013/01/12. pmid:23308327; PubMed Central PMCID: PMC3527023.
- 33. Bovet P, Romain S, Shamlaye C, Mendis S, Darioli R, Riesen W, et al. Divergent fifteen-year trends in traditional and cardiometabolic risk factors of cardiovascular diseases in the Seychelles. Cardiovascular Diabetology. 2009;8:34–. pmid:PMC2719584.
- 34. Cardoso I, Bovet P, Viswanathan B, Luke A, Marques-Vidal P. Nutrition transition in a middle-income country: 22-year trends in the Seychelles. Eur J Clin Nutr. 2013;67(2):135–40. Epub 2012/12/20. pmid:23249880; PubMed Central PMCID: PMC3786399.
- 35. World Health Organization. Countries: Seychelles 2015. Available from: http://www.who.int/countries/syc/en/.
- 36. Faeh D, William J, Tappy L, Ravussin E, Bovet P. Prevalence, awareness and control of diabetes in the Seychelles and relationship with excess body weight. BMC Public Health. 2007;7:163. Epub 2007/07/21. pmid:17640380; PubMed Central PMCID: PMC1959189.
- 37. UNCTAD. Country Fact Sheets: 2015 [cited 2015 01 may]. Available from: http://unctad.org/en/Pages/DIAE/World%20Investment%20Forum/Search-by-countryeconomy.aspx?Co=151,15,,&FD=1,15,,.
- 38. The World Bank. GNI per capita, Atlas method (current USD) 2015 [cited 2015 14 September]. Available from: http://data.worldbank.org/indicator/NY.GNP.PCAP.CD/countries/SC—XR?display=default.
- 39. United Nations General Assembly. Report of the Special Rapporteur on the right to education, Kishore Singh. 2014.
- 40. Statistics NBo. Statistics Bulletin: Quarterly Unemployment Statistics 2014-Q1. 2014.
- 41. Bovet P, William J., Viswanathan B., Madeleine G., Romain S., Yerly P., et al. The Seychelles Heart Study 2004: methods and main findings. 2007.
- 42. Bovet P, B. V, Louange M, Gedeon J. National Survey of Noncommunicable Diseases in Seychelles 2013–2014 (Seychelles Heart Study IV): methods and main findings Victoria, Seychelles2015. Available from: http://www.who.int/chp/steps/Seychelles_2013_STEPS_Report.pdf
- 43. Shim J-S, Oh K, Kim HC. Dietary assessment methods in epidemiologic studies. Epidemiology and Health. 2014;36:e2014009. pmid:PMC4154347.
- 44. Byers T. Food Frequency Dietary Assessment: How Bad Is Good Enough? American Journal of Epidemiology. 2001;154(12):1087–8. pmid:11744510
- 45. P K . An easy guide to factor analysis. Routledge, editor. London1994.
- 46. Link LB, Canchola AJ, Bernstein L, Clarke CA, Stram DO, Ursin G, et al. Dietary patterns and breast cancer risk in the California Teachers Study cohort. The American Journal of Clinical Nutrition. 2013;98(6):1524–32. pmid:24108781
- 47. Varraso R, Garcia-Aymerich J, Monier F, Le Moual N, De Batlle J, Miranda G, et al. Assessment of dietary patterns in nutritional epidemiology: principal component analysis compared with confirmatory factor analysis. The American journal of clinical nutrition. 2012;96(5):1079–92. Epub 2012/10/05. pmid:23034967.
- 48. Arruda SP, da Silva AA, Kac G, Goldani MZ, Bettiol H, Barbieri MA. Socioeconomic and demographic factors are associated with dietary patterns in a cohort of young Brazilian adults. BMC Public Health. 2014;14:654. Epub 2014/06/28. pmid:24969831; PubMed Central PMCID: PMC4082487.
- 49. Olinto MTA, Gigante DP, Horta B, Silveira V, Oliveira I, Willett W. Major dietary patterns and cardiovascular risk factors among young Brazilian adults. European Journal of Nutrition. 2012;51(3):281–91. pmid:PMC3313034.
- 50. Marchioni DM, Latorre Mdo R, Eluf-Neto J, Wunsch-Filho V, Fisberg RM. Identification of dietary patterns using factor analysis in an epidemiological study in Sao Paulo. Sao Paulo Med J. 2005;123(3):124–7. Epub 2005/07/16. /S1516-31802005000300007. pmid:16021275.
- 51. Fernandez-Alvira JM, Bammann K, Pala V, Krogh V, Barba G, Eiben G, et al. Country-specific dietary patterns and associations with socioeconomic status in European children: the IDEFICS study. Eur J Clin Nutr. 2014;68(7):811–21. Epub 2014/05/16. pmid:24824009.
- 52. Kesse-Guyot E, Bertrais S, Peneau S, Estaquio C, Dauchet L, Vergnaud AC, et al. Dietary patterns and their sociodemographic and behavioural correlates in French middle-aged adults from the SU.VI.MAX cohort. Eur J Clin Nutr. 2009;63(4):521–8. Epub 2008/01/24. pmid:18212801.
- 53. Hamilton LC. Statistics with STATA. USA.: Richard Stratton; 2013.
- 54. Barros AJ, Hirakata VN. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol. 2003;3:21. Epub 2003/10/22. pmid:14567763; PubMed Central PMCID: PMC521200.
- 55. Becquey E, Savy M, Danel P, Dabiré HB, Tapsoba S, Martin-Prével Y. Dietary patterns of adults living in Ouagadougou and their association with overweight. Nutrition Journal. 2010;9:13–. pmid:PMC2848625.
- 56. Azizi H, Asadollahi K, Davtalab Esmaeili E, Mirzapoor M. Iranian Dietary Patterns and Risk of Colorectal Cancer. Health Promotion Perspectives. 2015;5(1):72–80. pmid:PMC4430700.
- 57. Safari A, Shariff ZM, Kandiah M, Rashidkhani B, Fereidooni F. Dietary patterns and risk of colorectal cancer in Tehran Province: a case–control study. BMC Public Health. 2013;13:222–. pmid:PMC3605096.
- 58. Chen Y, Factor-Litvak P, Howe GR, Parvez F, Ahsan H. Nutritional influence on risk of high blood pressure in Bangladesh: a population-based cross-sectional study. The American Journal of Clinical Nutrition. 2006;84(5):1224–32. pmid:17093178
- 59. Bovet P, Chiolero A, Shamlaye C, Paccaud F. Prevalence of overweight in the Seychelles: 15 year trends and association with socio-economic status. Obes Rev. 2008;9(6):511–7. Epub 2008/08/05. pmid:18673305.
- 60. Ball K, Lamb KE, Costa C, Cutumisu N, Ellaway A, Kamphuis CB, et al. Neighbourhood socioeconomic disadvantage and fruit and vegetable consumption: a seven countries comparison. Int J Behav Nutr Phys Act. 2015;12(1):68. Epub 2015/05/23. pmid:25997451; PubMed Central PMCID: pmid:PMC4456793.
- 61. IFAD. Enabling poor rural people to overcome poverty in Seychelles. 2013.
- 62. Hopping BN, Erber E, Mead E, Sheehy T, Roache C, Sharma S. Socioeconomic indicators and frequency of traditional food, junk food, and fruit and vegetable consumption amongst Inuit adults in the Canadian Arctic. J Hum Nutr Diet. 2010;23 Suppl 1:51–8. Epub 2011/01/05. pmid:21158962.
- 63. Perrin AE, Dallongeville J, Ducimetiere P, Ruidavets JB, Schlienger JL, Arveiler D, et al. Interactions between traditional regional determinants and socio-economic status on dietary patterns in a sample of French men. Br J Nutr. 2005;93(1):109–14. Epub 2005/02/12. pmid:15705232.
- 64. Bojorquez I, Unikel C, Cortez I, Cerecero D. The social distribution of dietary patterns. Traditional, modern and healthy eating among women in a Latin American city. Appetite. 2015;92:43–50. Epub 2015/05/16. pmid:25975967.
- 65. Hoffmann JF, Nunes MA, Schmidt MI, Olinto MT, Melere C, Ozcariz SG, et al. Dietary patterns during pregnancy and the association with sociodemographic characteristics among women attending general practices in southern Brazil: the ECCAGe Study. Cad Saude Publica. 2013;29(5):970–80. Epub 2013/05/25. pmid:23703002.
- 66. Menchu M. La canasta básica de alimentos en Centro America. Guatemala: Intituto de Nutricion de Centroamérica y Panamá (INCAP), Organizacion Panamericana de la Salud (OPS), 2002.
- 67. Kearney J. Food consumption trends and drivers. Philosophical Transactions of the Royal Society B: Biological Sciences. 2010;365(1554):2793–807. pmid:PMC2935122.
- 68. Mayén A- L, Marques-Vidal P, Paccaud F, Bovet P, Stringhini S. Socioeconomic determinants of dietary patterns in low- and middle-income countries: a systematic review. Am J Clin Nutr. 2014.
- 69. Stringhini S, Viswanathan B, Gedeon J, Paccaud F, Bovet P. The social transition of risk factors for cardiovascular disease in the African region: evidence from three cross-sectional surveys in the Seychelles. Int J Cardiol. 2013;168(2):1201–6. Epub 2012/12/05. pmid:23206526.
- 70. Araujo MC, Verly Junior E, Junger WL, Sichieri R. Independent associations of income and education with nutrient intakes in Brazilian adults: 2008–2009 National Dietary Survey. Public Health Nutr. 2014;17(12):2740–52. Epub 2014/01/31. pmid:24476935.
- 71. Monteiro CA, Conde WL, Popkin BM. Independent effects of income and education on the risk of obesity in the Brazilian adult population. J Nutr. 2001;131(3):881S–6S. Epub 2001/03/10. pmid:11238779.
- 72. Clark AM, DesMeules M, Luo W, Duncan AS, Wielgosz A. Socioeconomic status and cardiovascular disease: risks and implications for care. Nat Rev Cardiol. 2009;6(11):712–22. pmid:19770848
- 73. Danaei G, Singh GM, Paciorek CJ, Lin JK, Cowan MJ, Finucane MM, et al. The Global Cardiovascular Risk Transition: Associations of Four Metabolic Risk Factors with National Income, Urbanization, and Western Diet in 1980 and 2008. Circulation. 2013;127(14):1493–502. pmid:23481623
- 74. Mohammadifard N, Sajjadi F, Maghroun M, Alikhasi H, Nilforoushzadeh F, Sarrafzadegan N. Validation of a simplified food frequency questionnaire for the assessment of dietary habits in Iranian adults: Isfahan Healthy Heart Program, Iran. ARYA Atherosclerosis. 2015;11(2):139–46. pmid:PMC4568199.
- 75. Fallaize R, Forster H, Macready AL, Walsh MC, Mathers JC, Brennan L, et al. Online Dietary Intake Estimation: Reproducibility and Validity of the Food4Me Food Frequency Questionnaire Against a 4-Day Weighed Food Record. Journal of Medical Internet Research. 2014;16(8):e190. pmid:PMC4147714.