1The prevalence of obesity and chronic disease has increased rapidly in recent decades, and currently represents the leading cause of death in the DROM and in mainland France alike. Obesity, diabetes, high blood pressure (HBP) and nutritional deficiencies are all elements included in the determination of a nutritional status, and nutritional factors (food consumption and physical activity) are recognised as major risk or protection factors with regards to chronic conditions such as obesity, cardiovascular disease, certain cancers and diabetes.

2The issue of nutrition and health will be addressed as part of the definition of a nutritional status (including the role of obesity and diet-related chronic diseases, nutritional deficiencies and social nutritional inequalities), and then the factors affecting nutritional status will be analysed (nutritional quality of diets, physical activity and sedentary behaviour4, genetic factors, specific contexts of each DROM with regard to food contamination or environmental pollution that can affect the nutritional status of populations). Finally, within the context of social, economic and geographical vulnerability, the disparity between the provision and use of healthcare for nutritional diseases, treatment plans and prevention methods will be successively examined as determinants of health.

1. Health and nutritional status

3Chronic diseases (high blood pressure, diabetes, cardiovascular diseases, etc.) have increased rapidly over the last few decades in industrialised, developing and emerging countries alike, and now constitute the leading cause of death in the world. Yet nutritional status indicators are generally worse in the DROM than in mainland France.

1.1. Obesity and food-related chronic diseases

4The main chronic problems related to food in the DROM can be identified as obesity, diabetes and high blood pressure, given their high prevalence (see I.1).

5In addition to the BMI (Body Mass Index) used to estimate the prevalence of overweight and obesity5, waist size is used to assess abdominal obesity6.

6In Martinique, the Escal survey (Quenel et al., 2008; Merle et al., 2008) showed that more than half of the adult population fell into the “excess weight” category, including 33 % in the overweight category and 20 % in the obesity category, and which included nearly twice as many women as men (26 % women compared to 14 % men). This trend had increased over the course of a decade, as stated in the Kannari survey conducted in 2013, with an eight-point rise in obesity. Among adults aged 16 and over, 52 % of men versus 64 % women fell into the “excess weight” category in 2013 (fig. 2). In Martinique almost six in ten adults (59 %) suffered from abdominal obesity, an increase of six percentage points since 2004 (53 %, Escal survey). The prevalence of abdominal obesity increases with age. Among children (aged 3-15), only one in four fell into the “excess weight” category and less than 8 % were obese according to the Escal survey, whereas in 2013 the Kannari survey estimated the prevalence of excess weight in children at 35 %, of which nearly 10 % were obese (fig. 3).

7According to the Kannari survey, the results in Guadeloupe were very similar to those of Martinique, with excess weight reported among more than half of individuals aged 16 and over (57 %) and which increased with age. The prevalence varies from 25 % among adults aged 16 to 24 to 75 % for those aged 65 or over. The abdominal obesity results for Guadeloupe were similar to Martinique (six in ten adults). In Guadeloupe, women are more affected than men, with more than twice as many women classified as obese (31 % women compared to 12 % men). For abdominal obesity there is also a very wide gap (79 % women compared to 37 % men). For the child population, the overall prevalence of obesity and overweight was 22 % of children (aged 5-14) in Guadeloupe, according to the Kannari survey.

8In French Guiana there are very few studies on the population’s weight status and the available data is old. The prevalence of excess weight in children affected 18 % of 5-14 year olds according to the Podium study conducted in 2008 and 19 % of 7-11 year olds according to the Nutriel study of 2009-2010. Just like in the other DROM, obesity affected a higher proportion of women than men in 2008 (23 % compared to 13 %) (Daigre et al., 2012).

9In La Réunion, the prevalence of excess weight and obesity is also higher than the average for mainland France. For adults, almost one in four people (38 %) falls into the “excess weight” category (27 % overweight and 11 % obese). The reported prevalence of obesity reached 14 % among women and 8.3 % among men (Richard et al., 2014a) (see I.1). In addition, one third of the adults presented with abdominal obesity (Favier et al., 2002). No studies were conducted on children. If we compare the four overseas territories where the DOM Health Barometer study was conducted (Richard et al., 2014a, 2014b, 2014c, 2014d), La Réunion reported the least instances of overweight and obesity (Richard et al., 2014a).

10In Mayotte, the studies are few in number, but those that exist show that this region sports the greatest prevalence of obesity: Nutrimay found an occurrence of overweight measured at 45.7 % and a prevalence of obesity at 20.4 %– and once again, there is a greater occurrence of female obesity, with almost one in every two women affected (47 %) compared to less than one in every five men (17 %) (Solet and Baroux, 2009). According to the national perinatal study conducted in 2016, obesity before pregnancy affected a quarter of the island’s mothers, i.e. double the number observed in the national sample. In addition, excess weight affected 4.7 % of children under the age of 5 and 8.5 % of 5-14 year olds.

11Lastly, the prevalence of metabolic syndrome7 is 18 % in Guadeloupe, and 28 % in Martinique (Colombet, to be published). It is 31 % in Mayotte (Solet and Baroux, 2009) and, in La Réunion, 37.6 % for women and 29.8 % for men (Roddier, 2001).

12Obesity and excess weight are among the risk factors (among others) of Type 2 diabetes, which represents more than 90 % of diabetes cases. Out of the five DROM, the most concerning situation is that of La Réunion (the prevalence there is twice as high for women than the national average).

13The DROM are characterised by a difference in the distribution of the prevalence of diabetes by gender compared to mainland France. Unlike mainland France, where diabetes affects more men than women, in the overseas departments and regions women are more affected and twice as many women are treated for diabetes in comparison to the mainland average (8 % compared to 16-20 %). In the Antilles, the prevalence of treated diabetes is almost 9 % in Guadeloupe and 8 % in Martinique (Santé Publique France data, 2016), whereas the average for the whole of France (except Mayotte) is close to 5 %. In French Guiana, the prevalence of treated diabetes is almost 8 %. La Réunion has the highest prevalence of treated diabetes, with a rate of 10 %. In Mayotte, the overall prevalence of diabetes was 10 % (Solet and Baroux, 2009).

14Regarding current developments, the latest Santé Publique France data show an overall trend of a slowdown in the increase in diabetes in the DROM (unlike France as a whole) and a reduction in the annual incidence of pharmacologically treated diabetes8 (see I.1). This trend is more prevalent among people over the age of 45 and women. Although the change in the annual prevalence rate of treated diabetes in males is 0.9 % in France as a whole and in Guadeloupe, it is only 0.4-0.5 % in French Guiana and Martinique and 0.3 % in La Réunion. This trend is even more marked in women over the age of 45 with an annual change rate of 0.3 % in Guadeloupe, 0.9 % in Martinique and - 1.1 % in La Réunion (+ 0.4 % in France as a whole). The evolution in incidence of occurrence is even more indicative: although in France as a whole the annual occurrence of treated diabetes in women over the age of 45 was 6.2 cases per 1,000 in 2017, compared to 7.2 in 2012, at the same time it changed from 21.5 to 15.0 in French Guiana, 16.8 to 12.6 in Guadeloupe, 13.8 to 10.8 in Martinique and 16.4 per 1,000 to 10.7 in La Réunion.

15The data from the Podium survey9 in 2008 show that the prevalence of high blood pressure was 17.9 % in French Guiana, 27.6 % in Martinique and 29.2 % in Guadeloupe (Atallah et al., 2012) (see I.1). Among other pathologies, the Kannari survey showed that 39 % of adults from Guadeloupe and 42 % of those from Martinique were suffering from high blood pressure in 2013. In Mayotte, more than 15 % of adult women had HBP (Vernay et al., 2009). High blood pressure and high cholesterol are frequently associated with diabetes: one in every three diabetics has high cholesterol and seven out of ten –and one in two adults in the Antilles –have high blood pressure. The prevalence of diabetes is sometimes also associated with a higher incidence of chronic terminal kidney failure (CTKF) than in mainland France, especially in La Réunion.

16To conclude, we will highlight that obesity (especially among women) is much more common in the DROM, and especially in Mayotte, with higher levels than the mainland France averages. This contributes to the development of conditions such as diabetes and high blood pressure. Although Type 1 diabetes is not more common in the French overseas territories, Type 2 diabetes is one of their public health priorities, especially in La Réunion. The high prevalence of diabetes, obesity, metabolic syndrome and HBP contributes to the high prevalence of complications, which often occur earlier, such as CTKF, coronary heart disease and strokes.

1.2. Nutritional deficiencies

17For the five French overseas regions a calcium deficiency is noted due to the very low consumption rate of fruit, vegetables and dairy products.

18Mayotte’s nutritional status appears marked by, on the one hand, dietary imbalances due to profound social changes resulting in increased nutritional risks (deficiencies, overweight or obesity) and, on the other hand, significant risks of infection. The study of the 2004 infant beriberi epidemic in Mayotte (Quatresous et al., 2004) put forward several assumptions about its aetiology: a narrowly-based diet at the end of pregnancy, exclusive or predominant breast-feeding with no Vitamin B1 supplementation, supplementary baby food low in thiamine and Vitamin B1. In 2006, the results of the Nutrimay study still included the threat of beriberi even though it had not re-emerged since 2004, probably due more to the provision of supplemental resources than changes in eating habits. At the same time, the same survey revealed the thinness of children under the age of 5 (7 %) that could be explained by non-optimal weaning, while a correlation emerged between child malnutrition and maternal obesity. The issue of malnutrition in children in its acute and chronic forms is confirmed by a study conducted by the education authority (2012) which cites the figure of 26 % of primary school pupils being under-nourished.

19Other biological deficiencies10 were updated by the Kannari survey in Guadeloupe and Martinique (2016) (Castetbon et al., 2016a, 2016b): serum folate deficiency and anaemia with low iron reserves could have as their source an inadequate dietary intake of vegetables not offset in other ways, while the shortage of vitamin D reserves could be explained by low intake of dairy products. It should be noted that the Vitamin D benchmark is difficult to achieve through diet, both in the DROM and in mainland France alike.

20Among nutritional problems, inadequate nutrition among the elderly is less documented than malnutrition through excess. In La Réunion, in 2016-2017, the Gramoune Care study revealed that elderly people living at home present a significant prevalence in malnutrition (Vincenot, 2019) (14 % of people aged 65 or over with a BMI less than 21 kg/m²), with no primary care screening and a delay in the treatment of those at risk. Women were more at risk of inadequate nutrition than men.

1.3. Social inequalities in nutrition

21Today, serious social inequalities in health, particularly in relation to diabetes, obesity and high blood pressure persist between the most socioeconomically advantaged individuals and those living in less advantageous conditions in France, and particularly in the DROM (Inserm, 2014). In the French overseas departments and regions there is fairly limited information on social inequalities in health, however, it is all the more crucial to reduce social inequalities in health in these territories as the general socio-economic situation is less advantageous for the DROM populations than for those of mainland France.

22As is the case for studies conducted in mainland France and conforming to international literature (Vigneron, 2011; Inserm, 2014), information from studies on the general DROM populations shows that the prevalence of excess weight, obesity, metabolic syndrome, diabetes and, to a lesser extent, high blood pressure, is higher among people with a lower socioeconomic status than those with the highest (see III.2). These variations are significant, especially when the level of education is taken into account. For the weight status, the differences in prevalence according to socio-economic position (SEP) are as high as 20 points for excess weight and 20 points for obesity in all regions (except in French Guiana, for which no information is available). The difference for diabetes is particularly high for La Réunion (20 points difference); followed by Guadeloupe (15 points) and Martinique (10 points). For high blood pressure and metabolic syndrome in the Antilles, the prevalence rates differ by 20 and 25 points respectively between the lowest and the highest socio-economic categories. In the Antilles, the overall increase in prevalence over the last two decades could have mitigated the social inequalities had it affected a larger cross-section of the general population, but instead the socio-economic inequalities remain with regard to weight status, diabetes and high blood pressure.

23The Reconsal study in La Réunion showed that the prevalence of excess weight and abdominal obesity is higher in people with few or no qualifications and in people without employment (abdominal obesity only), compared to more advantaged groups. In Mayotte, the risk of excess weight was greater for women with no schooling at all, and abdominal obesity was greater among women who were not of French nationality or did not have a good command of the French language. In Guadeloupe, a number of studies showed that the prevalence of excess weight, obesity and abdominal obesity is clearly higher only among women with few or no qualifications (Carrère et al., 2018; Yacou et al., 2015; Atallah et al., 2012). The more recent Kannari study (2014) highlighted an inverse association between excess weight and level of education, again among women only. Regarding the other SEP indicators, few variations were found in Guadeloupe for employment status and socio-professional category, but an inverse association was highlighted between obesity and living standards (income, affluence, minimum social benefits) (Orsag, s.d. a). In Martinique, the prevalence of excess weight, obesity and abdominal obesity appeared inversely associated with the level of education among women only (Quenel et al., 2008; Merle et al., 2008; Castetbon et al., 2016a, 2016b). Furthermore, the risk of being in the “excess weight” category is higher among women at the lowest overall socio-economic levels, whereas the reverse is true for men. Lastly, in Martinique, women receiving minimum social benefits were more at risk of excess weight and obesity than non-beneficiaries.

24Among children, the few studies available in the DROM do not show a clear relationship between child weight status and parental SEP. However, the lack of studies and the surveys’ methodological problems make it impossible to conclude that there are no socio-economic disparities regarding children’s weight status.

25The difference in prevalence for diabetes among people with the lowest socio-economic status and those with the highest is particularly great for La Réunion (20 points), but it is also noteworthy for Guadeloupe (15 points) and Martinique (10 points). In La Réunion, the prevalence of diabetes is inversely associated with the level of education and income and with being actively employed, whereas in Martinique it varies according to an overall socio-economic level. In Guadeloupe, there is only an association between SEP and diabetes for women, with a higher prevalence among women without qualifications and those receiving minimum social benefits (Carrère et al., 2018).

26For high blood pressure and metabolic syndrome in the Antilles, the prevalence rates differ by 20 points and 25 points respectively between the lowest and the highest socio-economic categories. In Martinique (Escal and Kannari studies), the association between high blood pressure and SEP disappears once it has been adjusted to BMI, showing a mediating effect of a greater prevalence of obesity among people of low socio-economic status. However, among those with high blood pressure, the most disadvantaged are more likely to be treated, compared to those who are more advantaged. In Guadeloupe, those with a low level of education, especially women, showed a risk of high blood pressure twice as high as those with a higher level of education, despite the BMI adjustment. The prevalence of metabolic syndrome appears to be higher among inhabitants of the Antilles who have a low level of education and those who receive welfare (RSA) (Colombet et al., 2019a). For high cholesterol, no significant relationship with SEP has been identified in the Antilles.

27In conclusion, in the five French overseas departments and regions there are significant social inequalities regarding excess weight, obesity, metabolic syndrome, diabetes and, to a lesser extent, high blood pressure, inequalities greater than in mainland France, as shown by the differences in prevalence according to socio-economic position. Given the much higher prevalence among women in all regions, these disparities according to SEP are also coupled with the gender effect: the inequalities for overweight and obesity are more pronounced for women – a phenomenon also seen in mainland France – which leads to the identification of the most disadvantaged women as a particularly at- risk group (de Saint Pol, 2007). There are several explanations for this. First of all, a causal effect: these women are the most corpulent because they find themselves in more precarious situations, particularly through reduced economic accessibility to a high-quality diet or expensive sports activities. Secondly, discrepancies in body image and representation depending on the SEP among women, with an under-assessment of weight status among the most disadvantaged women.

28When several socio-economic indicators are used in surveys, the level of education seems to be the most systematically associated with the risks of excess weight, obesity, diabetes, high blood pressure and metabolic syndrome, whereas associations with employment status and socio-professional category are generally insignificant, with results similar to those obtained in mainland France (Vernay et al., 2009). These relationships are explained, for the more educated people, by a greater ability to understand and integrate health messages, enabling knowledge to be translated into health-promoting behaviour (Yen et Moss, 1999; Galobardes et al., 2006; Davey et al., 1998). Precarious populations, which represent a significant proportion of the overall DROM population, are a high risk group with regards to nutritional status, as is also the case in mainland France. Therefore proposing appropriate measures for this population is a matter of urgency for public health.

2. Factors influencing nutritional status

2.1. Nutritional quality of diets

2.1.1. Compliance with recommendations

29Nutritional intake and diet quality are determining factors in the population’s nutritional status. An overview of food consumption for the DROM can be obtained through measuring their rate of compliance with the recommendations for adequate nutrients and food groups for adults and then for children and adolescents. However, there are no representative data on nutritional intake among adults in French Guiana.

30With regard to nutrients, the DROM are characterised by similarities in fibre, iron and sodium, and differences in energy and macronutrient intake.

31Fibre intake is slightly lower than that of mainland France, except for La Réunion, and is far from compliant with the RNP (repères nutritionnels pour la population - nutritional guidelines for the population) especially among women in Martinique. Iron intake does comply with the RNP. Sodium intake is slightly lower than in mainland France. Energy intake is low, which may be due to under-reporting (as the low intake does not correspond to prevalence of obesity). The reported energy intake is lower in the Antilles than in the other DROM, and intake here is about 10 % lower than that of mainland France, which is about 1,800 kcal/day. This is similar to statistics for Mayotte and La Réunion. However, the contribution to energy is very different in Mayotte compared to the other DROM. In this region, fat intake is about 25 % of the energy intake without alcohol (EIWA) (compared to 35-40 % of the EIWA according to the guidelines) and carbohydrate intake is 60 % of the EIWA (compared to 40-55 % according to the guidelines). Protein intake is also lower in Mayotte than in the other DROM, but in line with the guidelines (12-20 % of the EIWA).

32The data on micronutrients is incomplete. Vitamin B1 intake is below the guideline for men and women in Mayotte, whereas it is just about compliant in La Réunion. Vitamin C intake is compliant in the Antilles, but not in La Réunion.

33In addition to compliance with the recommendations for nutrients, there is also the question of compliance with the recommendations for food groups. Although half of the adult population shows an intake of meat, poultry, fishery products and eggs (MPFE) which is in line with the PNNS consumption guideline in mainland France, the situation differs greatly in the DROM with a larger percentage in Guadeloupe (82 %), compared to about 38 % for Martinique and Mayotte. For foods, MPFE consumption in the DROM is one and a half portions, in keeping with the guidelines, and that of red meat is very moderate (50 % of the population is below the new limited weekly consumption guideline of 500 g). Red meat consumption is a little higher in the Antilles than in mainland France and lower in Mayotte and La Réunion. With regard to meat, eggs and fishery products, almost four in every ten adults in Mayotte were below the recommended consumption levels.

34As for starchy foods, the percentages of compliance with the PNNS 3 guideline are similar for the Antilles and Mayotte. There is no data on wholegrain cereals. Consumption of the “bread, cereals, potatoes and pulses” food group complies with the guidelines for a little over half of the population of Mayotte. Today, rice is present in 40 % of the island’s food intake, alongside animal protein, bata11 (Brasset, 2014a), bread, pastries and soft drinks. In the Antilles, the consumption of starchy foods is equally diverse, though rice not being the staple food and bread being the most eaten instead. In La Réunion, cereals, starchy foods and pulses are the products most consumed, 1.7 times more than in mainland France. Among this group of products rice serves most often as foundation for traditional cooking on La Réunion, ten times more than in mainland France (Ricquebourg, 2018). Consumption of pulses, which is a feature of Creole dietary models, is still high, but with a sharp decline nevertheless occurring over the past ten to twenty years.

35Consumption of dairy products is lower in the DROM than in mainland France. Dairy products are insufficiently consumed by virtually all adults in Mayotte, leading to problems linked to calcium deficiency. For the other regions, compliance with the guidelines is poor in the Antilles (14 %) and very poor in Mayotte (2 %), whereas half the population of La Réunion sports an intake in line with the PNNS 3 consumption guideline, which is higher than in mainland France (29 %). The percentage of consumers complying with the PNNS 3 consumption guideline for dairy products varies greatly from one DROM to the next: from the 2 % mentioned in Mayotte to 59 % in La Réunion. Since the PNNS 4 has proposed reducing the guideline to two dairy products per day, these low percentages of consumers following the guideline should be reassessed.

36For fruit and vegetables, consumption is also lower for all of the DROM than for mainland France. In Mayotte, only 18 % of adults comply with the recommendations of the French National Nutrition Health and Nutrition Plan (PNNS) for fruit and vegetables consumed, which explains a vitamin deficiency on the island (Vernay et al., 2009) (see III.1). Consumption of fruit juice is high (almost half in Martinique). As for the percentage of consumers in compliance with the PNNS guideline, these amount to only 8 % for La Réunion and 11 % for Martinique, 18 % for Mayotte and Guadeloupe.

37Salt consumption is less than that of mainland France. The PNNS score for all the DROM and mainland France is very similar.

38The average daily consumption of sugary drinks in Guadeloupe and Martinique is almost three times higher than that of mainland France. In French Guiana it is almost double. Consumption in Mayotte and La Réunion is slightly higher (see I.2) than that of mainland France. In the case of Martinique, if consumption of fruit juice is added, the daily consumption limit for sugary drinks is almost reached. Three out of ten inhabitants of La Réunion consume sugary drinks at least four times a week, one in five of which every day and more among young people aged 18 to 30. As for sugary products, compliance with the recommendations is lower in Guadeloupe (42 %) compared to Martinique, where compliance is very similar to that of mainland France (73 %). Compliance with the guideline for sugary products is the highest in Mayotte (90 %). Overall, the Guadeloupe and Martinique scores for compliance with the PNNS recommendations12 are very similar to those of mainland France.

39In terms of differences in adult consumption by gender, calcium deficiencies in Mayotte particularly affect young women (aged 15-29). In the “bread, cereals, potatoes and pulses” food group, more than twice as many women than men had a lower intake than the recommended threshold and more men than women had an intake which exceeded the guideline. More women in Mayotte are non-compliant with the recommendations for the consumption of meat, eggs and fishery products. However, 70 % of both men and women ate seafood products at least twice a week. Another point the DROM have in common with each other, as well as with mainland France, is that men and those under fifty consume the most fat, starchy foods, alcoholic drinks (box 2) and foods from the “meat, fish and eggs” group. However, there is a discrepancy in Guadeloupe and Martinique, where men consume more fish and seafood than women. In all contexts, women and older people consume more fruit and vegetables and dairy products, unlike men. In this regard Mayotte is an exception, as fewer men than women there consume fruit and vegetables in limited quantities. As regards sugary drinks, a much larger number of men than women consume them, and in Mayotte the consumption of sugary flavoured drinks has doubled since 2005. During the last three months of pregnancy, almost one third of the women in the Nutrimay study (Vernay et al., 2009) claimed that they were no longer eating certain forbidden foods, mainly dried fish and seafood, while almost one in five breast-feeding women no longer ate seafood, insects or fresh fish. As for salt intake, it remains low for both genders. In the French Antilles, Guadeloupe and Martinique are quite similar in their consumption of the major food groups, according to gender and age.

40Also in the Antilles, the consumption of fruit juice decreases with age, while people under the age of 45 eat less fruit and vegetables, unlike those over 60. A larger number of people under the age of 45 also eat poultry, cured meats, cakes, sugary drinks, snacks and fast-food products. In particular, snacks13 and aperitif products are frequently consumed by young adults. In La Réunion, consumption of fruit and vegetables is significantly more common among older people.

41For children and adolescents, consumption data is only available for the Antilles and Mayotte. There are noteworthy similarities in consumption for Guadeloupe and Martinique. As regards nutrients, fibre intake is similar in Guadeloupe, Martinique and Mayotte for the 11 to 14/15 age range, and slightly lower than the guidelines of 20 g for a 15 year old adolescent, but in comparison Mayotte exhibits distinct differences in energy-, macronutrient-and calcium intake. Overall energy intake is lower than the RNP - nutritional guidelines for the population (from 1,400 kcal/day for 5 year olds to 2,500 kcal for 13-15 year olds), but it is much lower in Mayotte (1,260 kcal/day for children aged 5 to 14) compared to the Antilles (1,750 kcal/day). In Mayotte, as for the adults, children’s fat intake is very low (24 % of the EIWA) in comparison with the RNP (35 % of the EIWA) and carbohydrate intake is 60 % of the EIWA in comparison with the RNP (40-55 % of the EIWA). As for protein intake, it is also lower in Mayotte than in the Antilles, but it is within the range of the guidelines (12-20 % of the EIWA). Mayotte also has a specific characteristic regarding Vitamin B1 deficiency: the intake of Vitamin B1 is lower for both 5-10 and 11-14 age groups.

42In terms of compliance with the recommendations per food group for children, children from the Antilles do not consume very much yoghurt or cheese (only one in five children are in compliance with the PNNS), and do not consume much fruit or vegetables. Furthermore, almost three out of ten 11-15 year olds are below the PNNS recommendations for meat, fish and eggs, but six out of ten of them eat fish at least twice a week. In French Guiana, one study (Carbunar et al., 2014) shows that more than a quarter of pupils in Grades 2 to 4 consumed fruit and vegetables once or twice a day, and the same number ate these three times a day or more. Dairy products and protein were eaten once or twice a day by a third of children aged 6-10. Nearly all of the pupils ate starchy foods “from time to time” and most drink water at mealtimes. Consumption of fatty, salty or sugary products was low for more than half of the children in the study.

43To conclude this section on compliance with the nutritional recommendations, we note that the available data indicate trends similar to those of mainland France, with some differences. Fruit, vegetables and dairy products are not consumed as much in all of the overseas regions. Starchy foods are on the decline, whereas consumption of animal protein, simple carbohydrates and new products (food industry products and fast food) is increasing. In all cases children, adolescents and the under-30s are generally below the consumption thresholds recommended by the PNNS for fruit and vegetables and below for the MPFE group: sugary drinks and products, as well as for dairy products. In this latter regard milk is the main source of consumption, and at a higher percentage for boys. Generally speaking, adolescents, especially boys, consume more sugary drinks and biscuit-type products, fatty, sugary and salty foods, as well as snacks.

2.1.2. Relationship to nutritional status

44The analysis of the relationship between the nutritional status of individuals and their diet in the DROM is made difficult firstly due to the small number of studies that have explored it and secondly due to desirability-or reverse causality bias – well-known and not specific to the DROM – given that the studies are cross-sectional and individuals are aware of having been diagnosed with a nutritional pathology (diabetes, high blood pressure) or are simply aware that they are very corpulent. Most studies which have led to this type of analysis were identified in the Antilles and a clear picture of the biases mentioned was painted by the Kannari study: a better overall quality of diet, measured by the Diet Quality Index International (DQI-I), was paradoxically associated with a greater frequency of metabolic syndrome (p=0.04), but this relationship would disappear when the fact that the subjects were, or were not, aware of a diagnosis of chronic diet-related illness was taken into consideration (Colombet et al., 2019b). A single cohort study was identified in the report, focusing on the nutrition of pregnant women in Guadeloupe between 2004 and 2007 (Saunders et al., 2014). It showed that a diet closer to the traditional Mediterranean diet was associated with a lower risk of premature birth in very corpulent women (OR = 0.7, p < 0.01) and a lower risk of intra-uterine growth restriction in other women (OR = 0.8; p < 0.05). For some diet indicators less sensitive to the biases of cross-sectional surveys, studies have shown a relationship more in line with what might be expected: in La Réunion, the Reconsal study showed that people who were more corpulent based on their waist size had a priori a more energy-dense diet (exceeding the fat and protein recommendations), but less varied from a nutritional point of view (not compliant with the dietary guidelines for calcium, magnesium or Vitamin B2). In French Guiana, the Nutriel study showed a positive association between childhood obesity and buying meals from street vendors for Year 4 pupils. Finally, when studies focused on malnutrition, very few associations with food consumption were highlighted but dietary behaviour appeared to be more consistent with the pathologies observed – as in Mayotte, where the restrictive dietary practices of mothers during the end of the pregnancy and the breastfeeding period, poor dietary diversity and low intake of calcium and vitamins, shown by the Nutrimay survey in 2006, were very likely causes of the beriberi epidemic that occurred in 2004.

45The apparent compliance for energy can be explained by under-reporting. There is no sufficiently reliable data for French Guiana.

2.2. Physical activity and sedentary behaviour

46Physical activity means “any form of activity that improves health and operational capacity without any damage or excessive risk” (HEPA Europe, 2010) (see II.3).

47Sedentary behaviour is characterised by activities with energy expenditure close to that of resting (see note 4). It differs from physical inactivity, which is defined as an insufficient level of physical activity, i.e. below recommended levels. Societal transformations play a role in changing the population’s relationship to physical activity and sedentary behaviour, with changes in lifestyle specifically leading populations to reduce their level of physical activity. Socio-demographic factors such as gender, age (with any generational dynamics), professional status and family structure are also correlated with dietary and physical activity behaviour.

48Physical activity and sedentary behaviour will be addressed through several points. First, we describe the levels of physical activity and sedentary behaviour in relation to the recommended activity levels, then we will detail the socio-demographic and socio-economic determinants of activity and sedentary behaviour, comparing the DROM with mainland France. Next the distribution of infrastructure (sports amenities and facilities) will be presented according to the territories, along with a discussion of the possibility of improving them. Finally the connection between activity and nutritional status and diet-related chronic diseases will be developed.

2.2.1. Compliance with recommendations

49A large proportion of the overseas populations do not achieve the level of physical activity recommended by the national physical activity guidelines issued by the French Agency for Food, Environmental and Occupational Health & Safety (Anses) of at least 30 minutes per day. The rare data available for describing the level of physical and sports activity and sedentary behaviour in the French overseas departments are mostly declarative, measured by questionnaire.

50Roughly one third of the DROM adult populations performs regular physical activity14 (at least five times per week) i.e. a level that is fairly similar to that of mainland France (fig. 7-8). According to the DOM Health Barometer (Richard et al., 2014b), 36 % of French Guineans reported performing regular physical activity and 25 % seldom or never. La Réunion has the specific characteristic of boasting the highest level of regular physical activity in the DROM, higher even than mainland France: 44 % of La Réunion inhabitants perform regular physical activity and 12 % less than once a week. 28 % of Martinique inhabitants perform regular physical activity compared to 25 % less than once a week. As for Mayotte, the data from the Nutrimay study conducted in 2006 show that 26 % of the population aged over 15 perform a moderate level of physical activity and 29 % a high level. In 2014, 31 % of Guadeloupe inhabitants reported performing physical activity regularly, 44 % fairly regularly (one to four times per week) and 25 % said they rarely or never did any (less than once a week). In Guadeloupe, the Aphyguad survey, carried out in 2007 by the regional health observatory, concludes that 59 % of the adult population performs physical activity in line with national recommendations.

51For children, the Aphyguad survey (2010) in Guadeloupe shows that almost one third of children aged 6-14 walk or cycle to school, but only 56 % of them walk for ten minutes or more per day. Furthermore, less than half of them are active during breaks and this inactivity increases at high school level. On the other hand, almost half of them are involved in a sporting activity outside of school. In Martinique, a study (Caïus and Bénéfice, 2002) shows that practising sport is common among adolescents: more than 66 % engage in a weekly sporting activity outside of school; the weekly average is four hours for boys and two hours for girls. In French Guiana, a nutritional programme for pupils monitored from Year 2 to Year 4 highlights that for over a third of the pupils walking is not recognised as a physical activity, and neither is cycling (Carbunar et al., 2014). Data is not available for La Réunion and Mayotte on this subject for children and adolescents.

52As regards sedentary behaviour, on non-school days over one-third of 5-6 year-olds in Guadeloupe (36 %), Martinique and French Guiana (35 %) spend more than three hours a day in front of the television or playing video games (Guignon et al., 2010). Among 15-year-olds, a quarter of high school students in La Réunion report that they spend four hours or more on the Internet on school days, and 55 % per day on weekends (Guignon et al., 2010).

2.2.2. Socio-demographic and socio-economic determinants

53From the point of view of socio-demographic variables, in all contexts men are more physically active than women, though the level of activity decreases with age. According to Kannari, in Martinique men performed an activity of moderate to high intensity 3.5 times more often than women.

54With regard to older people, the survey entitled “State of health and living standards of people aged 65 and over in La Réunion” (2018) established that people aged 65 and over still engaged in regular weekly physical activity, 28 % of them every day, but there were still 31 % who performed no activity. In Guadeloupe, according to the study carried out by the Orsag, 43 % of people aged 65 and over comply with national recommendations. For Mayotte and French Guiana, the information is incomplete for older people.

55As for sedentary behaviour, age is a factor when it comes to time spent in front of a screen in Guadeloupe, as young adults (aged 15-24) are more likely to watch television for more than two hours a day than adults and older people.

56According to the data obtained from studies in Guadeloupe, the overall level of physical activity varies according to the adult’s SEP: a positive relationship is recorded between a high SEP and practising sport, while a high level of physical activity at work is associated with a lower SEP (labourers, technicians). The percentage of people practising a sport is higher when the level of education increases and is also higher among non-working (but not unemployed) people and people working full-time compared to the unemployed, a result comparable to that of mainland France (Inserm, 2014). Furthermore, the relationship between the practice of a sport and income level constitutes a bell curve, with a greater proportion of individuals occupying middle-income categories involved in sport. The Kannari study in Martinique (Castetbon et al., 2016a, 2016b) revealed that the level of physical activity was not associated with either the level of education or employment status, or the fact of receiving RSA (Colombet, to be published).

57If we look at the socio-economic disparities in sedentary behaviour, the results are paradoxical depending on the sedentary behaviour indicator under consideration. For example, there is no significant relationship in Guadeloupe between the time spent in front of the television and the level of academic qualification, but those who have qualifications better than the high school diploma or A levels are more likely to spend more than two hours a day in front of a computer and spend more time sitting down than those of a lower qualification level, a result which could be explained by the fact that better-qualified workers tend to gravitate towards more sedentary employment. In the Kannari study in Martinique the prevalence of sedentary behaviour was higher among the unemployed or those receiving RSA, but also among executives compared to labourers and lower-level employees (Colombet, to be published). In Mayotte, according to the Nutrimay survey (2006), the risk of a sedentary lifestyle among women was lower among those who had no schooling at all and those without a good command of French (Vernay et al., 2009), thus identical to the situation in mainland France (Inserm, 2014).

58For children, a low SEP appears to be negatively associated with the level of physical activity and a sedentary lifestyle, but the studies are not sufficient for drawing any conclusions. According to the Aphyguad survey in Guadeloupe (Orsag, 2010a), the proportion of children going to school on foot or by bicycle was higher among those belonging to households of which the primary breadwinner was unemployed. However, children from households where the primary breadwinner had a full-time job were more likely to practise a sporting activity outside the school environment and spent less time in front of a screen at lunchtime on school days and in front of a computer on non-school days. In terms of family structure, the presence within the family of a child under the age of 15 in Guadeloupe appears to confirm this trend with regards to both the performance of physical activity and the fact of being employed. The data from the Kannari study in Martinique have shown that the percentage of children exhibiting a low level of activity is higher among children whose role model possessed a low level of education and was a labourer or a low-level employee, or did not work at all (Colombet, to be published). The same applies to the prevalence of sedentary behaviour: it is higher when the children’s role model possessed a low level of education and was a labourer or a low-level employee, or did not work at all (Colombet, to be published). A study in La Réunion has shown that fewer students in the professions practise sport than those in the general and technological fields (Bardot, 2018).

2.2.3. Territorial provision of sporting equipment and facilities

59In all territories (except in French Guiana), outdoor activities are most popular including, among aquatic activities, sailing, surfing or “yole” boat sailing in Martinique, as is hiking on land. With regard to team sports, football is primary, followed by basketball and handball. Athletics and swimming are the individual sports most practised, along with martial arts (especially in French Guiana and La Réunion). In La Réunion, so-called “nature” sports (paragliding, mountain biking, hiking, trails, etc.) have become extremely popular. For its part Martinique boasts many “yole” schools, “yole” being a traditional Martinique sailing boat.

60Sporting activities (to be clearly distinguished from simple physical activity) require facilities and equipment, and these are unequally distributed between the DROM as well as internally within each region. Overall the proportion of sports facilities in the DROM is less than that in mainland France: in 2010 the national average was 39.9 sports facilities for 10,000 inhabitants, compared to 28.6 in La Réunion, 23.6 in Martinique, 23.6 in Guadeloupe and 27.4 in French Guiana. In 2019, Mayotte had 12.9 facilities for 10,000 inhabitants. Mayotte and French Guiana are generally behind as regards the establishment of sports facilities compared to the other DROM and mainland France, but French Guiana does boast a lot of swimming pools, although mainly along the coast.

61On the island of La Réunion, there is a discrepancy between “land” sports facilities (hiking, climbing, cycling) and water sports facilities, developed more recently. The island is currently shifting towards a joint use model of school sporting infrastructure, i.e. shared usage by pupils and external users. These sporting facilities are unequally distributed within the territory and located on the coasts, more to the north of the island than the south.

62Martinique is generally under-equipped with regard to sporting facilities compared to mainland France, also with a disparity between the north and the south of the island: for example, there are no public swimming pools in the north or south-west of the island, and few tennis courts in the north and south, as opposed to the centre.

63In the overseas territories the type of activity, the age of the infrastructure as well as anti-cyclonic and anti-seismic standards to be maintained impose additional restrictions to improving the territory’s provision of sporting infrastructure as well as to maintaining what already exists. Nevertheless, in La Réunion despite the lack of sports facilities the optimisation of existing facilities remains possible for the foreseeable future, for example by using time slots available for different user groups (companies, retirees, etc.) or by providing night lighting.

2.2.4. Relationship to nutritional status

64As noted in the study of the connections between nutritional status and diet, the same desirability- and reverse causality biases are found in the surveys which estimate individual physical activity, which are all cross-sectional and the results of which are ultimately far from conclusive. In addition, the only significant relationships found in these surveys only concerned sporting activities and not physical activity in general, and related mainly to women. In Martinique the Escal survey (Quenel et al., 2008; Merle et al., 2008) showed that the practice of a sport other than walking at least once a week was more common among women without excess weight than among those who were overweight or obese (respectively 41 and 28 %; p < 0.05), but among girls aged 3-15 the opposite was found: the practice of a sporting activity was reported more often by obese girls (p < 0.05). For adults in La Réunion, according to the Reconsal study (2002), the practice of a sporting activity was more common among subjects without an “at-risk” waist size.

65When it comes to estimating the level of sedentary behaviour within the same studies the relationship to the nutritional status is clearer, and is in line with expectations of greater corpulence being associated with increased sedentary behaviour. However, this still only concerns women in the Escal survey in Martinique: in adulthood the percentage of women watching television for four hours per day or more actually increased with corpulence (not overweight: 13 %; overweight: 16 %; obese: 22 %; p < 0.05) and the same relationship was found for girls aged 3-15 (p < 0.05). The Reconsal study in La Réunion similarly showed that 37 % of obese people, both men and women, watched television for more than three hours a day, compared to 30 % of those with a normal BMI and 20 % of thin people (p < 0.05). The other studies which estimated the physical activity and/or sedentary behaviour in adults and/or children (particularly Kannari in Martinique, Aphyguad in Guadeloupe and Nutriel in French Guiana) showed no significant relationship to corpulence, provided that the comparisons were adjusted in relation to the main socio-economic characteristics. However, an analysis of the Kannari survey (Colombet, to be published) revealed that the risk of hypertriglyceridemia was lowest when the subjects exhibited a high level of physical activity.

66Lastly, in the specific population of adults suffering from chronic diet-related illness the results are once again rather inconclusive. In Martinique, in the Escal survey, 26 % of people with high blood pressure performed a sporting activity compared to 40 % of people with normal blood pressure (p < 0.0001), but this relationship was not found in the Kannari survey or the Aphyguad survey in Guadeloupe. As for diabetes, conversely, the reporting of a sporting activity was more common among known diabetics in the Redia2 survey in La Réunion (2006-2009) (Lenclume and Favier, 2012).

2.3. Genetic factors

67Among the risk factors of developing obesity is the interaction of the consumption of high energy products, low energy expenditure and genetic factors (gene mutations controlling the appetite as well as the operation of a hundred genes linked to obesity). Our diet is indeed likely to cause temporary or permanent gene mutations, including possibly hereditary epigenetic mutations, based on what some research suggests. These nutri-epigenetic studies15 help determine profiles of more at risk individuals in the population.

68Only a few of these mutations relate to the DROM inhabitants however, whose origins are diverse and more or less mixed (for example, 80 % of the Guadeloupe population have African ancestors). One of the studies highlighted five mutations in four genes associated with obesity in children of Afro-Antilles origin, the frequency of genetic polymorphisms associated with obesity thus varying for this population compared to the inhabitants of predominantly European origin. In this regard, through the analysis of the total genome, the Genome Wide Association Study (GWAS) evaluated the increased risk of obesity associated with certain variants of the Fat mass and obesity-associated protein (FTO) gene, which could be counteracted by an increase in physical activity.

69Some genetic variations also affect the appetite for sweet tastes more or less markedly depending on the populations and age. Children of mainly Afro-American ancestry would be especially prone to a high consumption of sugar, which does not hold true for adults with the same ancestry (Mennella et al., 2005). This observation suggests that for sweet taste preferences the food experience could be a greater determinant than genetic heritage. Another study revealed that genetic factors account for 30 % of the perception of the sweet taste (Hwang et al., 2015), suggesting that genetic factors are also involved in the perception of certain molecules that influence our dietary choices – these perceptions are found to be very different among populations of European or African origin however.

70Ultimately the existing studies show the strong interaction between genetics and lifestyle in terms of dietary practices and physical activity, raising the prospect of using best practices as a driver to combat genetic predispositions to chronic diseases.

2.4. Food contamination and environmental pollution

71The impact of environmental pollution on the health of populations will not be dealt with here, as changes in the behaviour of the population with regard to food contamination or environmental pollution cannot be described due to a lack of studies on the subject. Our presentation therefore aims instead to summarise the specific situation facing each DROM and to discuss the population’s susceptibility to information about these issues (see II.2).

2.4.1. Chlordecone in the Antilles

72Chlordecone was used as an insecticide in the Antilles from the 1960s to 1993, especially in relation to banana cultivation. Its low rate of degradation into the environment explains the significant and persistent soil contamination resulting from its use. This pollutant is most commonly found in the blood of inhabitants of the Antilles (Anses, 2017), who are persistently exposed to low doses. Humans are contaminated mainly through their food as aquatic environments are affected with a bio-accumulation of chlordecone in crustaceans and other species. As for soil pollution, it also contaminates vegetables and tubers as well as eggs and meat.

73In 2013 it was estimated that 52 % of Guadeloupe inhabitants and 31 % of Martinique inhabitants obtained supplies through short distribution channels (Anses, 2017), with the people most exposed to chlordecone having obtained supplies in these informal ways.

74Knowledge of this soil contamination and its effects on health was gradually disseminated among the general population, and in Martinique and Guadeloupe awareness of the use of chlordecone and its environmental and health consequences have led to changes in the dietary habits of every four people in ten (Inpes, 2009) due to food safety fears. For households in Guadeloupe and Martinique, where the gardens were contaminated with chlordecone above the maximum residue limits, the Jafa programme (family gardens) was set up to reduce exposure to chlordecone through produce obtained from gardens or family farms situated on contaminated soil. Where land is contaminated with more than 100 µg/kg, Jafa advisors provide tailored monitoring, giving advice on behaviour aimed at reducing risk (limit root vegetables to twice a week for example, carefully wash and peel thoroughly before cooking, diversification with more green vegetables, etc.), as well as on which crop types and growing methods to prioritise (either on mounds or off the ground).

2.4.2. Mercury and lead in French Guiana

75Gold-mining activities are responsible for mercury being the main food safety concern in French Guiana. In the aquatic environment, mercury is converted into methylmercury, which is ingested by fish which in turn are consumed by humans. When ingested, mercury can cause impaired neurological or behavioural development in foetuses and children. The Amerindian populations of Upper Maroni and Oyapock are particularly exposed due to a diet rich in fish.

76Lead poisoning in French Guiana children is another public health problem of which there is a strong suspicion of it being caused by diet. The Guyaplomb study conducted between 2015 and 2017 showed that lead mostly affects boys, children in disadvantaged socio-economic environments and those living along the coastline or in villages on the Oyapock River (Camopi and Trois-Sauts).

2.4.3. Multiple risk factors in La Réunion

77Mercury contamination exists in both La Réunion and French Guiana, and pregnant women and young children may be exposed to an increased health risk related to the consumption of swordfish and other large marine predators due to high levels of methylmercury ingested by these creatures (Afssa, 2006; Anses, 2013).

78As in mainland France, current media coverage is focused on glyphosate, used in La Réunion as a herbicide in sugar cane fields. Contamination mainly comes from drinking water. According to the president of the OasisRéunion association, the results for those tested in mainland France as part of the “glyphosate in our urine” campaign were 100 % positive, even for those who only consume organic products.

79With regard to animal health, bovine leukosis is also a health risk in La Réunion, and may affect 70-80 % of the island’s livestock. Yet in La Réunion, unlike in mainland France, since the meat was not exported the contaminated cattle were not slaughtered and their produce still used given that the virus found in the meat and milk is destroyed by heat (milk by sterilisation and meat through cooking). The virus is therefore not implicated in human illnesses, but a certain mistrust of local products is starting to be created by the media.

2.4.4. Water-related health risks

80Drinking water carries health risks common to several DROM. In Mayotte there are dangers inherent in both water consumption and waste water treatment, in addition to which a large number of Mayotte inhabitants do not have easy access to drinking water in the first place. In 2013 the Insee estimated that only 72 % of primary residences had a water source in the home (Luce et al., 2017). Contamination of certain rivers could also be the source of a number of hepatitis A cases (Aboudou et al., 2019) and there is widespread inadequate sanitation leading to faecal-oral transmission with a high risk of an epidemic.

81Drinking water also raises a number of questions in La Réunion, with the incidence of acute gastroenteritis in La Réunion double compared to that in mainland France. Parasites have been detected in surface water resources, in addition to which such surface catchments – source of more than half of the water distributed by public networks – make the water vulnerable to pollution. More generally, the construction of public waterworks infrastructure for water destined for human consumption is lagging far behind requirements in La Réunion, hence the insufficient quality of sanitation (depending on the sectors or periods), a potential shortage during the dry season and an increase in nitrates and pesticides. As a result it remains an important priority to set up water treatment systems to address the health risks to the population.

82It should be noted that, while the subject of food contamination and health crises, as well as contamination mechanisms, are extensively studied, consumer concerns and how these issues affect the population’s behaviour are less so.

3. Prevention and use of healthcare

83Healthcare in the DROM should be assessed, and particularly in accordance with the profile of the DROM population – an ageing population in the Antilles, a population with a late demographic transition in French Guiana and in Mayotte, the populations of these two DROM still growing as in La Réunion – but also in relation to nutritional diseases (see II.6). Disparities in the provision and use of healthcare, as well as treatment plans for nutritional diseases are even more important to consider, as they can be factors in causing health inequality.

3.1. Health care provision and use

84French Guiana and the Antilles are characterised by a substantial shortage of GPs (half as many in French Guiana as in mainland France), but they have a relatively larger number of private practice nurses than mainland France (Baert et al., 2017; Girdary et al., 2017). For its part, Mayotte offsets the low number of GPs to some extent with four local hospitals and 13 clinics spread across the territory. In French Guiana, the relocated prevention and healthcare centres, branches of the Cayenne hospital, help to deal with the most urgent cases. In La Réunion, on the other hand, there is a high density of GPs and specialists, as is also the case for private practice nurses, the latter especially in the eastern part of the island.

85Intra-territorial disparities are still significant, sometimes with areas of high medical density (e.g. in Fort-de-France) and others with great shortages (Port-Louis) in the same DROM. In French Guiana, this disparity is reflected particularly between the coast, where nearly three-quarters of the population have access to healthcare exceeding their needs, and the rest of the territory, where one in ten French Guinean have to travel more than 20 minutes for medical care.

86Recourse to healthcare varies depending on the region. In Martinique, the rate of healthcare consultation is quite similar to Guadeloupe, where the number of medical procedures is much lower than in mainland France. In French Guiana, access to primary healthcare proves difficult (added to an overload of emergency cases). The same difficulty in accessing consultations may explain the relatively high employment of nurses in Guadeloupe, and may also be explained by a political desire to keep elderly people at home: 93 % of Guadeloupe inhabitants receiving the Personal Independence Allowance actually live at home (compared to only 61 % in mainland France). As regards the Guadeloupe practitioners, the age factor is also an important consideration for healthcare since more of them are older than the national average (Tantin-Machecler and Camus, 2014). Mayotte is characterised by a large foreign population (over 40 % of the inhabitants in 2007), which makes greater use of traditional medicine (through the fundi and mwalimu) than private medicine. The use of healthcare in La Réunion mainly involves GPs, with most inhabitants generally satisfied with their consultations, except when the waiting times for an appointment exceed one month. In general recourse to healthcare in relation to nutritional diseases is insufficient to meet the needs of the overseas regions (particularly in French Guiana and Mayotte).

87Non-use of healthcare in the DROM appears to be above the national average, except in La Réunion. In Guadeloupe more than one person in ten does not use healthcare, which can be explained by the population’s standard of living since almost one in five households was living below the poverty line in 2012. Other explanations include lack of time (especially for women and young people), the distance from the office or transportation difficulties. As in Guadeloupe, financial reasons are the main reason for the non-use of healthcare in French Guiana (more than one person in three) especially for dental care, unlike other DROM (54 % vs. rates varying between 28 and 36 %). The rate of non-use of healthcare in Martinique for financial reasons is a little lower, as it relates to a only a fifth of the inhabitants, while the rate of non-use of healthcare is lowest in La Réunion (Richard et al., 2014a). A third of the inhabitants of La Réunion self-medicate and this behaviour is reinforced by the conservation of unconsumed medicines. All generations in La Réunion use alternative remedies (such as herbal infusions) (Boulevart, 2016), mainly due to cultural reasons.

3.2. Treatment plans for nutritional diseases

88Treatment plans are important insofar as chronic diet-related diseases require preventive, therapeutic and follow-up measures in the long term. The various sources available highlight the fact that women follow treatment plans more closely and that young people, the most economically vulnerable and geographically remote as well as patients with multiple conditions do not follow them as closely.

89The Entred study done in 2007-2010 (Ndong et al., 2010) on the DROM (except Mayotte) reveals a lack of consultations and follow-up examinations while the Rédia 2 study showed the importance for health of an early diagnosis of particularly diabetes, as newly screened patients faced more uncontrolled risk factors for complications (Lenclume and Favier, 2012). In La Réunion the qualitative studies (Balcou-Debussche, 2006, 2016a, 2016b) and the preliminary data from the study on treatment plans for patients with diabetes based on information from the national inter-scheme information system for health insurance (Sniiram) and the information system medical orientation programme (PMSI) (Chan Wai Nam et al., 2018) clearly show breaks in the treatment plans, particularly among the most disadvantaged: discontinuation of follow-up and treatment after the 1^st phase of diagnosis and treatment by the treating doctor, early onset of complications with “re-discovery” of diabetes, the important role of the GP in determining the relevance of the treatment plan and the hospital as expert point of reference (Balcou-Debussche and Debussche, 2009).

90For cases of under-nutrition, an insufficient use of primary healthcare can be highlighted. The treatment plan consists of screening and care during hospitalisations for surgery or serious diseases, then outpatient follow-up. The question of earlier screening can be also raised, including for elderly people, as highlighted by the Gramoune Care study in La Réunion in 2016- 2017.

91The number and density of specialists in nutritional diseases in the DROM are slightly lower than in mainland France, with a slightly lower density for 100,000 inhabitants. However, in relation to real needs – with twice as many people with diabetes than in mainland France, faced with more complications – the number of specialists is actually very low, especially in French Guiana and Mayotte. The increased socio-economic vulnerability in the DROM is an aggravating factor. The density and distribution of specialists in the Antilles and La Réunion imply a greater overall provision of primary care, whereas the lower density of GPs is accentuated in some areas by intra-territorial disparities. The larger number of nurses goes some way towards rebalancing the situation, though requiring some adjustment. Out of all the DROM, French Guiana has the highest disparity in the follow-up rate for tests recommended for people with diabetes (Fosse-Edorh and Mandereau-Bruno, 2015), whereas follow-up figures appear higher in La Réunion than elsewhere in France.

92For obesity, care is mostly provided by GPs. In the Antilles and La Réunion the creation of three centres specialising in obesity in 2012 has improved the coordination of healthcare use, as have dedicated bariatric surgery centres. However, for the latter their level of activity is lower than in mainland France (Hazard et al., 2018). In Mayotte and French Guiana there is minimal treatment of obesity, with a low number of specialists, no dedicated centre and the obligation of going to another DROM or mainland France to seek treatment. During pregnancy however, the treatment rate of obesity in Mayotte and La Réunion is twice as high as in mainland France.

3.3. Prevention

93La Réunion and the Antilles offer a significant number of varied preventative actions operating through institutions and associations.

94There is a similar level of information on chronic diet-related diseases in the DROM and mainland France: between 73 and 83 % of the inhabitants of French Guiana, La Réunion, Martinique and Guadeloupe feel they are well-informed about diabetes, less so about high blood pressure (52-53 % in French Guiana and La Réunion, 63-64 % in the Antilles). More than eight out of ten Guadeloupe inhabitants report being well-informed thanks to the media, the medical profession (especially among women) and those around them (especially men) (Naud and Cornely, 2010). The population’s interest is evident in the reading of product ingredient labels – four in ten Guadeloupe inhabitants do so routinely– although for one third this information appears difficult to understand. The link between available information and nutritional knowledge is important for the population to be able to feed itself in full knowledge of the facts.

95The difficulty accessing information on chronic diet-related diseases is sometimes compounded by general difficulties involving lack of command and adequate comprehension of the language used – in the DROM command of the written and spoken language is not always at the same level as in mainland France. Health literacy is characterised by the ability to access and use information for one’s own health (Kickbusch et al., 2013). For this reason in Guadeloupe the Karudiabète survey looked for explanations for inadequate diabetes care in relation to relevance of communication vis-à-vis the audience targeted (Vail and Cornely, 2014). The difficulty in finding and evaluating good information on Type 2 diabetes was also evident in the Ermies study in La Réunion, highlighting the critical role of health professionals and services in helping patients manage their diabetes.

96In conclusion then, the provision and use of healthcare are characterised by a shortfall in the use of general and specialised medicine, with more recourse to home nurses due to a centralised provision which is sometimes very far away geographically. As a result, the significant regional and intra-territorial disparities accentuate the effect of socio-economic vulnerabilities, with the most significant discrepancies found in French Guiana and Mayotte. In spite of this however, the continued presence of local facilities (prevention and healthcare centres, health clinics) remains an asset in these territories, even though access and recourse to specialists in nutritional diseases are poor in relation to the needs of the DROM (and critical in French Guiana and Mayotte) – the use of primary healthcare may be increasingly important, but it is limited by the inadequacy of the services available in relation to the populations’ needs and chronic illnesses.