Version classiqueVersion mobile

Alimentation et nutrition dans les départements et régions d’Outre-mer/Food and nutrition in the French overseas departments and regions

Caroline Méjean
Xavier Debussche
Yves Martin-Prével
et al.

English Synthesis

Introduction : Diverse transitions facing the French overseas departments and regions

Texte intégral

  • 1 Nutritional transition is a term used to describe the transition from a situation where deficiency- (...)

1This summary report is based on the referenced contributions of scientists and professionals gathered together over a period of several months in a multidisciplinary committee (representing epidemiology, public health, nutrition, economics and sociology). It reports on the current state of knowledge regarding determining factors affecting food and nutrition in the French overseas regions, with the aim of clarifying public policy in this regard. Apart from their varying geographic characteristics, Guadeloupe, French Guiana, Martinique, Mayotte and La Réunion have experienced more recent demographic, social and economic transformations than mainland France, each in its own specific way. The evolution of these French overseas departments and regions (referred to here by the French acronym DROM) occurs rapidly, as do transitions in their food-related and nutritional situations1, which operate at more or less advanced stages of development depending on the territory.

2In general the French overseas populations have found themselves in a state of “pressured modernisation” (Chang, 2016) over the course of just a few decades, with a rapid, direct shift from a predominantly agricultural society towards a largely tertiary economy in the making (Rochoux, 2003; Bertile, 2002). Profound social changes have appeared, including the emergence of a middle class and changes in family structure, a greater presence of audio-visual technology in the home and a modernisation of domestic tools. The demographic situation, socio-economic factors and food-and nutrition-related transitions vary depending on the overseas territory in question.

Demographic situation

  • 2 These cross-references correspond to the chapters written by the experts and attached to the digita (...)
  • 3 Source :

3With regard to demographic change, the five French overseas regions are not all in the same position (see I.1 and III.12). Life expectancy at birth is higher in Martinique (84.7 years for women and 78.6 years for men in 2018), similar to mainland France (85.4 years for women and 79.5 years for men). The indicators in Martinique are trailed by those of Guadeloupe (84.4 years for women and 77.3 years for men) and La Réunion (83.9 years for women and 77.8 for men), followed by French Guiana (83.3 years for women and 78.2 years for men), whereas in Mayotte they decrease (76 years for women and 75.6 years for men)3. In Guadeloupe and Martinique the demographic transformation terminated more than twenty years ago (decline in birth rate, negative immigration rate and an ageing population), albeit with a few differences between the two: Martinique does not have demographic growth either through the arrival of new immigrants or the return of retirees from mainland France, whereas Guadeloupe has a population growth rate that is still positive due to a birth rate higher than that of Martinique. The ageing of the population in the Antilles explains the rise in the number of childless couples. The situation is quite different in French Guiana and Mayotte, where the median age of the population is only 23 and 18 years respectively, with a high birth- and immigration rate, the latter mostly foreign in origin and economically insecure. In La Réunion the demographic transformation is not yet complete, sporting as it does natural growth and a young population, but the island is experiencing the beginnings of an ageing process. Although all of the DROM have a higher rate of single-parent families than mainland France (average rate in the DROM: 41 % compared to 22 % in mainland France), it is above the DROM average in French Guiana and Mayotte.

Socio-economic characteristics

4The DROM also have specific socio-economic profiles. The level of education in the DROM (between 39 and 68 % of people aged 15 and over with no or very few qualifications) is much lower than that of mainland France (30 %). Unemployment is an additional handicap with rates two to three times higher than those of continental France, especially in Mayotte and La Réunion. In fact, the poverty rate is more than double that of mainland France. Gender inequality is an aggravating factor and in that respect women are over-represented among the most disadvantaged populations.

5Overall, Mayotte is in the most disadvantageous position (84 % of the population below the poverty line, a precarious social situation for foreigners, an estimated illiteracy rate of around 35 %), along with La Réunion (poverty rate of 42 % and illiteracy rate of 21 %) and French Guiana (poverty rate of 44 %), whereas the population of the French Antilles are in a better position (poverty rate on both islands around 20 %), yet still worse off than mainland France (17 % in 2017, according to Insee). As a result, the French overseas departments and regions suffer more pronounced social inequalities than mainland France, with repercussions in terms of mortality and morbidity with regard to diet-related chronic diseases and their associated risk factors (see III.2).

Food and nutritional transition

6Nutritional transition is traditionally described as the shift by a society where malnutrition is predominant to a state where diseases related to excess consumption become the norm. The evolution in the prevalence of chronic diseases in the DROM over the last few decades correlates to profound dietary changes, changes in the environment, lifestyle, social structure, etc. In terms of diet, traditional dietary models co-exist with new models, with a decrease in traditional starchy foods and tubers, plants and vegetables and an increase in the consumption of animal protein, sugar and processed products. Dietary transition unfolds differently depending on the overseas territory in question. Thus in Mayotte it is less advanced than in the other DROM, with the local dietary model incorporating African (Bantu, Sahel), Madagascan and Creole influences. In the 1980s, the Mayotte diet was centred on coconut, rice, bananas, cassava, vegetables and a small quantity of proteins, fats and simple sugars (Fontaine, 1995), whereas from the 2000s there is an increase in the consumption of rice, vegetables and tubers, but still a limited consumption of animal protein (Vernay et al., 2009). In Guadeloupe and Martinique the traditional model, based primarily on starchy foods and a relatively small proportion of animal protein has been eroded with a reduction in the intake of complex carbohydrates and dried vegetables (Lombion, 2012) in favour of a model that includes more animal protein, simple carbohydrates and fats (Orsag, 2010a). In La Réunion we find signs of dietary transition with a decrease in starchy foods, plants and vegetables and an increase in the consumption of animal protein, sugars and new products as part of a diversification of the food supply (fast food, processed products). As far as French Guiana is concerned, what little information is available does not support any conclusions.

7Changes in shopping trends and supply also serve as proof of nutritional transition and are partly related to international trade and the increasing dependence on food imports.

8This expert appraisal is based on available sources: studies and surveys conducted since the 2000s, presented in a timeline chart for easier identification throughout the text (fig. 1). Although the value of these studies and surveys are clear, the absences or gaps in certain topic areas should be noted, as well as methodological differences which sometimes make it difficult to compare territories and interpret data (box 1).

9This updated body of knowledge on determining factors related to nutrition and food in the French overseas regions has been divided into three subject areas comprising the first three parts, followed by a final part including recommendations: nutrition and health (Part I); dietary models (Part II) and finally socio-economic determinants and the food supply (Part III). More specifically, Part I reports on the nutritional status of overseas populations: the nutritional factors (nutritional quality, physical activity and sedentary lifestyle, genetic factors) and health risks (food contamination and environmental pollution), prevention and the use of healthcare. Part II focuses on dietary models through dietary behaviour, the structure of food consumption and the sociability of food, followed by different social and cultural representations. Finally, Part III explores the socio-economic determinants and the food supply (distribution channels, food prices and socio-economic determinants of food consumption). The knowledge gathered on the nutritional status of French overseas populations shows that there are similarities between the five regions studied, but also at times major differences. Each part will therefore include graphs or summary tables on the main issues raised for each region.

Box 1: Data limitations and quality
This expert appraisal is mainly based on the data provided by surveys conducted between 2000 and 2017 (fig. 1). The results of the expert appraisal do, however, highlight several gaps and difficulties regarding the studies consulted and the data collected. The lack of information available for some DROM and rather out-dated information available for others have made it impossible to respond to the questions raised with regard to all of the territories. In French Guiana, for example, no studies on nutritional status, dietary behaviour and physical activity have yet been carried out among the general population. Nevertheless, it should be remembered that, although collecting information on the nutritional status of individuals is not particularly complicated (which does not mean that there are regular collections), collecting data on food consumption requires considerable resources, both in terms of finance and logistics and in terms of the technical level of the human resources employed.
The difficulties encountered have included the small number of studies conducted thus far, the age of much of the available data, the level of accuracy of the samples (varying from one study to the next), the phenomenon of under-reporting, the poor comparability of the variables and indicators, and also the desirability bias.
As a result, the available studies represent a certain heterogeneity. Although a number of them were carried out by experienced teams on representative samples chosen at random from the general population by applying subject-specific epidemiological standards – both upon selection and later in terms of sample adjustment – others involved specific populations only, with at times non-optimal random selection procedures and/or without a subsequent sample adjustment. Furthermore, a number of surveys were triggered by the existence of specific “health-related problems” (chlordecone, beriberi, etc.) and as a result concentrated on particular areas or on certain population categories. While most of the surveys took these factors into account and subsequently applied the necessary adjustments, their stratified sampling method led to some population groups being too small to produce sufficiently precise figures.
With regard to measuring food consumption, the studies have all had to contend with the classic phenomenon of under-reporting as previously observed in this type of survey. This phenomenon may be related to a desirability bias, but also to other factors such as the survey participants’ lack of time, apathy or poor understanding or simply the surveyors’ level of technical expertise, or imperfect data collection instruments. The Kannari study in the Antilles showed that people with a low level of education (primary or secondary school only) are more likely actively to under-report, irrespective of other socio-demographic characteristics (Colombet,
to be published). As a result, it is possible that the variations observed between populations sporting different levels of education with regard to consumption might be mitigated by overall under-reporting among the most underprivileged groups. Excessive under-consumption also generally leads to subjects being excluded from the analyses (although this depends on the case in question), which affects the sample’s representativeness. This problem is not unique to the DROM, but under-consumption, where present, was somewhat more prevalent in the studies analysed here than in surveys conducted in mainland France. In general, in addition to under-reporting specific to food consumption, we are sometimes confronted with a significant amount of missing data which, in much the same way, again raises the issue of representativeness bias.
Some additional difficulties were encountered in the analysis due to the poor comparability of the studies with one another. The most typical example of this is the fact that anthropometric measurements of weight and height were sometimes taken in the home during surveys, but at other times were obtained by self-declaration only. We know that in the latter case more corpulent individuals generally tend to under-estimate their weight and/or over-estimate their height, and slighter subjects sometimes over-estimate both their height and weight. In other cases the lack of comparability derived from a change in definition, threshold or standard – at national or international level – for one of the variables characterising nutritional status (for example the definition of the metabolic syndrome, thresholds for waist size considered as representing in individual “at risk”) or concerning a variable used to calculate an indicator (e.g. the French National Health and Nutrition Programme - PNNS - guidelines). In other cases, non-standard methodological choices were made by the authors without their justification always being presented. There may also be cases where the definition of variables or indicators, or their meanings, is not identical for all populations or for all survey teams.
A final point to be kept in mind regarding the interpretation of results is that we possess very few chronological series allowing for the estimation of trends within a particular population, and also that many of the studies examined are relatively dated. The age of many studies, in particular on the links between nutritional status and proximal determinants (conducted 10 years ago or more, except for Kannari), necessarily affects conclusions drawn within the context of very rapid changes in the nutritional situation, as well as modes and types of consumption.

Figure 1: Timeline chart of the dates and locations of the studies and surveys used for the joint scientific expert appraisal


1 Nutritional transition is a term used to describe the transition from a situation where deficiency-related nutritional problems are predominant to one where nutritional problems associated with excessive consumption are the norm; it is characterised by an increase in the prevalence of obesity and metabolic diseases such as Type 2 diabetes and cardiovascular disease, while diseases caused by malnutrition may persist simultaneously for longer or shorter periods (a phenomenon known as “double burden”). Nutritional transition is the result of a combination of dietary transition on the one hand, reflecting at times profound changes in diets and dietary models, including an increase in the consumption of animal fat, sugar and salt and, on the other hand, lifestyle changes related to economic development and urbanisation, including an increasingly sedentary lifestyle.

2 These cross-references correspond to the chapters written by the experts and attached to the digital version of the work.

3 Source :

Table des illustrations

Légende Figure 1: Timeline chart of the dates and locations of the studies and surveys used for the joint scientific expert appraisal
Fichier image/jpeg, 95k

© IRD Éditions, 2020

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search