The authors have declared that no competing interests exist.
Conceived and designed the experiments: AJS CSGE CW JCKW MCB MCT. Analyzed the data: CSGE. Wrote the first draft of the manuscript: CSGE. Contributed to the writing of the manuscript: AJS CD CM CSGE CW JCKW MCB MCT NSU PS.
Surveying women and children from refugee camps in Algeria, Carlos Grijalva-Eternod and colleagues find high rates of obesity among women as well as many undernourished children, and that almost a quarter of households are affected by both undernutrition and obesity.
Households from vulnerable groups experiencing epidemiological transitions are known to be affected concomitantly by under-nutrition and obesity. Yet, it is unknown to what extent this double burden affects refugee populations dependent on food assistance. We assessed the double burden of malnutrition among Western Sahara refugees living in a protracted emergency.
We implemented a stratified nutrition survey in October–November 2010 in the four Western Sahara refugee camps in Algeria. We sampled 2,005 households, collecting anthropometric measurements (weight, height, and waist circumference) in 1,608 children (6–59 mo) and 1,781 women (15–49 y). We estimated the prevalence of global acute malnutrition (GAM), stunting, underweight, and overweight in children; and stunting, underweight, overweight, and central obesity in women. To assess the burden of malnutrition within households, households were first classified according to the presence of each type of malnutrition. Households were then classified as undernourished, overweight, or affected by the double burden if they presented members with under-nutrition, overweight, or both, respectively.
The prevalence of GAM in children was 9.1%, 29.1% were stunted, 18.6% were underweight, and 2.4% were overweight; among the women, 14.8% were stunted, 53.7% were overweight or obese, and 71.4% had central obesity. Central obesity (47.2%) and overweight (38.8%) in women affected a higher proportion of households than did GAM (7.0%), stunting (19.5%), or underweight (13.3%) in children. Overall, households classified as overweight (31.5%) were most common, followed by undernourished (25.8%), and then double burden–affected (24.7%).
The double burden of obesity and under-nutrition is highly prevalent in households among Western Sahara refugees. The results highlight the need to focus more attention on non-communicable diseases in this population and balance obesity prevention and management with interventions to tackle under-nutrition.
Good nutrition is essential for human health and survival. Insufficient food intake causes under-nutrition, which increases susceptibility to infections; intake of too much or inappropriate food, in particular in interaction with sedentary behaviour, can lead to obesity, which increases the risk of non-communicable diseases such as diabetes. During the past 30 years, the prevalence (the proportion of a population affected by a condition) of obesity has greatly increased, initially among adults in industrialized countries, but more recently among children and in less-affluent populations. Now, worldwide, overweight people outnumber under-nourished people. Furthermore, some populations are affected by both under-nutrition and obesity, forms of malnutrition that occur when the diet is suboptimal for health. So, for example, a child can be both stunted (short for his or her age, an indicator of long-term under-nutrition) and overweight (too heavy for his or her age). The emergence of this double burden of malnutrition has been attributed to the nutrition transition—the rapid move because of migration or urbanization to a lifestyle characterized by low levels of physical activity and high consumption of refined, energy-dense foods—without complete elimination of under-nutrition.
Refugees are one group of people in whom under-nutrition and obesity sometimes coexist. Worldwide, in 2010, 15.4 million refugees were dependent on host governments and international humanitarian agencies for their food security and well-being. It is essential that these governments and organizations provide appropriate food assistance programs to refugees—policies that are appropriate during acute emergencies may not be appropriate in protracted emergencies and may contribute to the emergence of the double burden of malnutrition among refugees. Unfortunately, the extent to which the double burden of malnutrition affects refugees in protracted emergencies is unknown. In this cross-sectional study (an investigation that looks at the characteristics of a population at a single time), the researchers assessed the double burden of malnutrition among people from Western Sahara who have been living in four refugee camps near Tindouf city, Algeria, since 1975.
The researchers used data from a 2010 survey that measured the height and weight of children and the height, weight, and waist circumference of women living in 2,005 households in the Algerian refugee camps. For the children, they estimated the prevalence of global acute malnutrition (which includes thin, “wasted” children, as indicated by a low weight for height based on the World Health Organization growth standards, and those with nutritional oedema), stunting, and underweight and overweight (low and high weight for age and gender, respectively). For the women, they estimated the prevalence of stunting, underweight (body mass index less than 18.5 kg/m2), overweight (body mass index greater than 25 kg/m2), and central obesity (a waist circumference of more than 80 cm). Among the children, 9.1% had global acute malnutrition, 29.1% were stunted, 8.6% were underweight, and 2.4% were overweight. Among the women, 14.8% were stunted, 53.7% were overweight, and 71.4% had central obesity. Notably, central obesity and overweight in women affected more households than global acute malnutrition, stunting, and underweight in children. Finally, based on whether a household included members with under-nutrition or overweight, alone or in combination, the researchers classified a third of households as overweight, a quarter as undernourished, and a quarter as affected by the double burden of malnutrition.
These findings indicate that there is a high prevalence of the double burden of malnutrition among households in Western Saharan refugee camps in Algeria. Although this study provides no information on men and does not investigate whether the obesity seen in these camps leads to an increased risk of diabetes and other non-communicable diseases, these findings have several important implications for the provision of food assistance and care for protracted humanitarian emergencies. For example, they highlight the need to promote long-term food security and to improve nutrition adequacy and food diversity in protracted emergencies. In addition, they suggest that current food assistance programs that are suitable for acute emergencies may not be suitable for extended emergencies. They also highlight the need to focus more attention on non-communicable diseases in refugee camps and to develop innovative ways to provide obesity prevention and management in these settings. However, as the researchers stress, careful policy and advocacy work is essential to ensure that efforts to deal with the threat of obesity among refugees do not jeopardize support for life-saving food assistance programs for refugees.
Please access these websites via the online version of this summary at
Wikipedia provides background information about the
The World Health Organization provides information on all aspects of
The
The
Oxfam also provides detailed information about its work in the
An article published by the Food and Agriculture Organization of the United Nations explains the
In 2010, the United Nations High Commissioner for Refugees (UNHCR) reported a worldwide estimate of 43.7 million displaced persons
During the last 30 years obesity has increased worldwide
The presence of obesity among refugees, generally assumed to be rare, has nevertheless been described in populations fleeing conflict zones where economic development has already impacted dietary intakes and activity patterns. For example, female Bosnian refugees granted asylum in Sweden had significantly higher values of body mass index (BMI) and waist circumference (WC) than age-matched Swedish women
Refugee populations from less-developed countries that migrate to camps are known to present with nutritional deficiencies, such as iron deficiency anaemia
UNHCR routinely monitors health and nutrition indicators of children aged <5 y and women of childbearing age (15–49 y) in the Western Sahara refugee setting through nutrition surveys. Results are primarily used to better evaluate needs and allocate resources; they are also used to indirectly estimate the impact of programme implementation by assessing secular changes in health and nutrition indicators.
The nutrition survey used in this study, besides estimating nutrition indicators, also aimed to provide pre-intervention baseline data for a future impact evaluation of a UNHCR-led micronutrient supplementation programme intended to reduce micronutrient deficiencies.
When considering whether ethical approval would be required for the impact evaluation, we were guided by a recent expert meeting report
The nutrition survey was approved by UNHCR and the refugee health authorities. Informed verbal consent was obtained from all adults and caregivers.
Since 1975, people from Western Sahara (also known as Sahrawi) have lived as refugees in camps near Tindouf city, in southwest Algeria, an area with a harsh desert environment. Their situation is considered a protracted emergency, as there is a stalemate to negotiations, with no sign of imminent resolution. Although accurate estimates are not available, the host country estimates that there are ∼165,000 people living in four camps (Awserd, Dakhla, Laayoune, and Smara), mostly dependent on food assistance from international organisations.
A two-stage household cluster survey with four strata (one per camp) was conducted in October–November 2010 to collect nutrition indicators from children (<5 y) and women of childbearing age (15–49 y). The survey design followed UNHCR survey guidelines
Sample size was calculated using previous prevalence data for global acute malnutrition (GAM), stunting, and anaemia in children, and anaemia and obesity prevalence in women, using ENA for SMART software (beta version, November 2008;
To detect a reduction in stunting prevalence from 39.1% to 29.6%, with 80% power and 10% significance level, in children aged 6–35 mo, a sample of 592 children, aged 6–59 mo, was calculated to be required from each stratum. The calculation assumed a normal distribution for HAZ, with a standard deviation of one, a design effect value of 1.5, a desired precision of 10%, and the assumption that 60.6% of children aged 6–59 mo would be 6–35 mo of age.
For women, a target sample size of 574 non-pregnant women, from each stratum, was calculated, based on an expected prevalence of BMI≥25 of 47%, a design effect of 1.5, and a desired precision of 5%.
Households were defined as a group that shared meals and slept under the same roof. An estimated 508 households per camp were needed to reach the required sample size, assuming 1.2 children aged 6–59 mo per household, and a 3% non-response rate. For women, 328 households were estimated required, assuming 1.8 women of childbearing age per household. All eligible children and women present in each household were sampled.
To determine the appropriate number and size of the survey clusters, a pilot field data collection exercise was conducted following training of health and nutrition workers. It was decided to set the cluster size equal to the number of households that a survey team could complete in one day. This resulted in a design using 30 clusters of 17 households in each stratum.
The survey was carried out by ten survey teams (four members each) and four supervisors. A two-stage sampling method was used
Children's date of birth was recorded from health cards. If cards were absent, caregivers were asked to recall the date. Women were asked their age in years, and whether they were currently pregnant and/or lactating. Those reporting either status were excluded from anthropometric measurements.
Weight was measured to the nearest 0.1 kg using a digital scale (Seca 876, Seca). Children were weighed without clothes, and those unable to stand were weighed with the caregiver, using the mother/child function of the scale. Women were asked to remove all accessories, shoes, and excess clothing before being weighed.
Height (or recumbent length for children aged <2 y or measuring <87 cm) was measured to the nearest 0.1 cm using a portable stadiometer (ShorrBoard, Shorr Productions). BMI was obtained by dividing weight (in kilograms) by height (in metres) squared.
WC is considered by some a better predictor of metabolic risk for the individual than BMI
Presence of bilateral pitting oedema in children was recorded if an imprint remained in both feet after pressing for 3 s.
For children, all anthropometric measurements were computed into
For women, the BACON algorithm, a tool for outlier detection of related variables (i.e., weight, height, and WC), was used to flag and exclude outliers from analyses
Only children aged 6–59 mo and non-pregnant, non-lactating women were included in the analyses and reported in this study.
In children, acute malnutrition (based on WHZ and/or the presence of oedema), stunting (based on HAZ), and underweight (based on WAZ) were defined and classified as global/total (<−2), moderate (<−2 but ≥−3), and severe (<−3)
For women, stunting was classified as total (<−2), moderate (<−2 but ≥−3), and severe (<−3); underweight was defined as BMI<18.5 kg/m2, overweight as BMI≥25 but <30 kg/m2 and obesity as BMI≥30 kg/m2. Metabolic risk by central obesity was defined as increased (WC≥80 but <88 cm) or substantially increased (WC≥88 cm)
Although both women and children were classified as underweight based on international standards
All sampled households were initially classified as to whether or not they contained cases of (1) under-nutrition in children (GAM: WHZ<−2 and/or oedema; stunting: HAZ<−2; underweight: WAZ<−2) or in women (stunting: HAZ<−2, underweight: BMI < 18.5 kg/m2); or (2) overweight in children (BAZ>2) or women (overweight: BMI≥25 kg/m2; central obesity: WC≥80 cm).
Next, to quantify the proportion of households with the double burden of malnutrition, households were selected and classified using a modified method previously described
We extended the standard definition of double burden by using WC, considered by some a better predictor of metabolic risk than BMI
All statistical analyses were carried using Stata (Stata IC release 11, StataCorp). To check for sample bias, initial comparisons were carried out between strata for mean age in both target groups, and for the proportion of males in children.
Means and proportions were calculated using Stata's
Camp | Households | Children <5 y | Women aged 15–49 y | |||||
Sampled | Consented | Refused | Empty | Total | 6–59 mo | Total | Non-Pregnant, Non-Lactating | |
Awserd | 510 | 506 | 2 | 2 | 408 | 366 | 635 | 381 |
Dakhla | 510 | 489 | 14 | 7 | 444 | 418 | 719 | 460 |
Laayoune | 511 | 505 | 6 | 0 | 392 | 349 | 626 | 416 |
Smara | 513 | 505 | 3 | 5 | 514 | 475 | 768 | 524 |
Total | 2,044 | 2,005 | 25 | 14 | 1,758 | 1,608 | 2,748 | 1,781 |
For children <5 y, date of birth was missing in four (0.2%); 146 (8.3%) were aged <6 mo, and 91.5% were aged 6–59 mo. For women of childbearing age, age was missing in 15 (0.5%); 26 (0.9%) had no pregnancy or lactating status recorded, 318 (11.6%) were classified as pregnant, 612 (22.1%) as lactating, and 64.8% as non-pregnant, non-lactating.
Summaries of age distributions for both target groups are presented in
Characteristic | Camp | Total ( |
|||
Awserd ( |
Dakhla ( |
Laayoune ( |
Smara ( |
||
|
|||||
Male sex (percent [95% CI]) | 50.0 (45.0–55.0) | 52.2 (46.4–57.9) | 51.9 (47.7–56.1) | 51.6 (46.8–56.4) | 51.4 (48.9–53.9) |
Age (months [95% CI]) | 28.8 (27.3–30.4) | 31.3 (29.7–32.9) | 29.9 (28.1–31.6) | 31.0 (29.4–32.7) | 30.3 (29.4–31.1) |
6–17 mo (percent) | 31.1 | 27.5 | 29.2 | 25.7 | 28.2 |
18–29 mo (percent) | 24.0 | 23.7 | 24.1 | 25.3 | 24.3 |
30–41 mo (percent) | 21.0 | 17.2 | 22.1 | 21.3 | 20.3 |
42–53 mo (percent) | 15.0 | 21.5 | 17.8 | 17.1 | 17.9 |
54–59 mo (percent) | 8.7 | 10.0 | 6.9 | 10.7 | 9.3 |
Age (years [95% CI]) | 30.3 (29.5–31.2) | 30.8 (29.6–32.1) | 30.5 (29.8–31.4) | 30.1 (29.3–31.0) | 30.4 (30.0–30.9) |
15–24 y (percent) | 36.2 | 36.5 | 35.6 | 38.7 | 36.9 |
25–34 y (percent) | 29.9 | 27.0 | 28.8 | 24.4 | 27.3 |
35–44 y (percent) | 20.5 | 19.3 | 19.7 | 24.2 | 21.1 |
44–49 y (percent) | 13.4 | 17.2 | 15.9 | 12.6 | 14.7 |
Non-pregnant, non-lactating women.
95% CI, 95% confidence interval.
Of the total sample of children selected for analysis (
Overall, in children, the main form of under-nutrition observed in the camps was stunting, followed by underweight and, lastly, GAM (
Indicator/Variable |
|
Mean ± SD | Prevalence (95% CI) |
|
|||
|
|||
Mean WHZ | 1,541 | −0.38±1.24 | |
Moderate (WHZ<−2 but ≥−3) | 6.3 (5.1–7.5) | ||
Severe (WHZ<−3 and/or oedema) | 2.8 (2.0–3.7) | ||
Global (WHZ<−2 and or oedema) | 9.1 (7.6–10.7) | ||
|
|||
Mean HAZ | 1,488 | −1.29±1.24 | |
Moderate (HAZ<−2 but ≥−3) | 20.9 (18.7–23.2) | ||
Severe (HAZ<−3) | 8.2 (6.7–9.6) | ||
Total (HAZ<−2) | 29.1 (26.4–31.9) | ||
|
|||
Mean WAZ | 1,584 | −1.00±1.18 | |
Moderate (WAZ<−2 but ≥−3) | 13.5 (11.8–15.2) | ||
Severe (WAZ<−3) | 5.1 (3.8–6.4) | ||
Total (WAZ<−2) | 18.6 (16.5–20.8) | ||
|
|||
Mean BAZ | 1,527 | −0.25±1.23 | |
Overweight (BAZ>2 but ≤3) | 2.4 (1.6–3.3) | ||
Obesity (BAZ>3) | 0.8 (0.4–1.3) | ||
|
|||
Mean HAZ | 1,700 | −1.07±0.93 | |
Moderate (HAZ<−2 but ≥−3) | 13.0 (11.2–14.7) | ||
Severe (HAZ<−3) | 1.9 (1.1–2.7) | ||
Total (HAZ<−2) | 14.8 (12.9–16.8) | ||
|
|||
Mean BMI (kilograms/metre2) | 1,699 | 26.1±5.2 | |
Underweight (BMI<18.5) | 5.1 (4.0–6.2) | ||
Overweight (BMI≥25 but <30) | 31.8 (29.6–34.0) | ||
Obesity (BMI≥30) | 21.9 (19.6–24.2) | ||
Overweight or obesity (BMI≥25) | 53.7 (51.0–56.4) | ||
|
|||
Mean WC (centimetres) | 1,689 | 87.4±12.4 | |
Increased (WC≥80 but <88) | 23.5 (21.3–25.7) | ||
Substantially increased (WC≥88) | 47.9 (45.2–50.6) | ||
Central obesity (WC≥80) | 71.4 (68.7–74.1) |
Non-pregnant, non-lactating women.
95% CI, 95% confidence interval; SD, standard deviation.
Of the total sample of women selected for analysis (
For women, stunting was the main form of under-nutrition observed (
Scatterplot of the relationship between BMI (in kilograms/metre2; linear regression coefficient 0.22, constant 19.3) and WC (in centimetres; linear regression coefficient 0.60, constant 69.0) with age among Western Sahara refugee women.
Proportions of households with a member affected by malnutrition in women and children, Western Sahara refugee camps.
Regarding overweight and central obesity in women, central obesity affected a greater proportion of households than overweight. The proportion of households with children classified as overweight was low. Overall, a total of 53.5% of households had at least one case of either overweight or central obesity, in any target group.
Our results showed a negative association between the presence of cases with under-nutrition and the presence of cases with overweight in households (
Proportion of households classified as normal, double burden, overweight, and undernourished in Western Sahara refugee camps. Overweight and the double burden in each stacked bar is based on two different indicators used to classify either obesity (BMI) or central obesity (WC).
Similar results were observed if overweight classification in women was based on central obesity (WC≥80 cm), although the proportions of households classified as overweight or double burden were slightly greater than those observed using BMI (
To our knowledge, this is the first detailed study of the prevalence and the coexistence of under-nutrition and overweight in a protracted refugee setting where the population has not experienced economic development and is dependent on food assistance for survival. Our results demonstrate that both stunting (in children and women) and obesity (in women) are highly prevalent among Sahrawi refugees, with central obesity being even more prevalent and appearing at a younger age in women than obesity. Second, more households were affected by overweight and central obesity than by under-nutrition, although the latter affected over one-third of households. Third, an important proportion of refugee households, one in four, are affected by the double burden of malnutrition. The results raise crucial and challenging issues for the design of refugee assistance programmes, and the future provision of care for obesity-associated co-morbidities among Sahrawi refugees and other similar populations.
At the proximate level, how could a population that was previously nomadic, possibly experiencing chronic energy insufficiency, have developed the observed high levels of overweight and obesity while living in refugee camps in the absence of economic development? Various factors previously suggested to be associated with obesity among Sahrawi women living in Western Sahara urban centres
One factor is that the Sahrawi were traditionally nomadic and culturally associate larger bodies with wealth and beauty, thus fattening practices involving periods of ritual overfeeding, and the use of appetite enhancers and traditional medication (suppositories composed of a mix of dates, seeds, and medicinal plants that are believed to increase peripheral fat accumulation), were common among Sahrawi
Another factor is an excessive sugar consumption habit among the Sahrawi
These factors help to partially explain the high prevalence of overweight in this population; however, they are complemented by other factors affecting refugees living in the camps, which at a more ultimate level help explain the high prevalence of both under-nutrition and overweight in this population. Importantly, some factors that are associated with under-nutrition in early life appear to increase susceptibility to overweight in later life (see the thrifty phenotype hypothesis
One crucial aspect is that Sahrawi refugees are dependent on food assistance to cover most of their nutritional needs and thus lack agency over their food system. A typical food assistance basket for this population will often be rich in starchy foods (refined grain cereals, pulses, and blended foods) and sugar. The refugee food assistance package typically contains low quantities, if any, of fresh or dried vegetables and fruit, therefore providing a low-diversity diet
Historically, a high prevalence of obesity, as observed among Sahrawi refugee women, is commonly described among groups that have suffered from severe cultural and economic disruptions with prolonged food insecurity, followed by a rapid transition to more refined foods
There are various strengths and innovations in this study. Data collection followed a robust and detailed nutrition survey design using internationally recognised protocols for household sampling. This allowed for assessment of nutritional status at both individual and household levels. In addition, to our knowledge, this is the first time obesity and the double burden of malnutrition were assessed using a measure of central obesity, as well as BMI.
However, there are some limitations. First, assessing obesity via BMI and WC presents methodological challenges, as both indicators are affected by variations in lean mass
In addition to methodological challenges, women wore clothes during weight measurement, which could have increased the overweight prevalence. However, this is unlikely to change the findings, as the prevalence of BMI≥25 kg/m2 was only 2.6 percentage points lower after deducting 1 kg for women's clothing during a sensitivity analysis.
Further, we lack additional data that would allow further interpretation of our results. For example, no anthropometric data were obtained for men or other household members, and therefore we could not evaluate their nutritional status. While females are often more at risk of obesity, as demonstrated by the slightly greater secular changes in females' BMI that have occurred worldwide in the last 30 years
We also did not collect clinical or biochemical markers of metabolic syndrome, and therefore could not ascertain the associated burden of disease in this population. However, data from urbanised Sahrawi women, with comparable levels of central obesity (75%), showed that 16.3% had metabolic syndrome and 28.6% were hypertensive
As no socioeconomic data were collected, we could not assess the role of known social determinants of obesity and under-nutrition, such as poverty or economic and gender inequalities
Lastly, as observed in
Overall, our study highlights an evolving need to focus more effort on NCDs in protracted refugee settings, particularly on obesity, associated co-morbidities, and the double burden of malnutrition described here. The high prevalence of obesity in this Sahrawi refugee population should not be assumed to imply that the population receives excessive or even adequate nutrition. Both under-nutrition and overweight may be considered as alternative forms of malnutrition, where the diet is suboptimal for health
First, the emergence of obesity and the double burden of malnutrition has serious implications for how international organisations should plan and provide assistance, especially for those exposed to conflict or displacement of protracted duration. For example, food assistance policies need to be revised and adapted, as those currently designed to meet population minimum needs during an acute emergency will need to consider their potential contribution to the later development of NCDs. Additionally, efforts are needed to promote long-term food security and higher nutrition adequacy in protracted emergencies. The actions needed range from improved food security assessments, with special focus on diversity within food groups, to provision of cash or vouchers, to community involvement in sustainable livelihood programmes such as gardening and small-scale business. UNHCR will need to work with the World Food Programme and other organisations on this issue. The Sahrawi refugees have been residing in camps since 1975. Generations of adults from birth have received food assistance as their main source of food. Their children are now the second or third generation exposed to a consistently low-quality diet. The intergenerational impact of this exposure is of serious concern in this and similar protracted emergencies
Second, efforts are needed to evaluate and monitor the health impact of obesity and the double burden in refugee situations. Obesity and NCDs should be routinely included in nutrition and health assessment exercises in protracted refugee settings, and should be incorporated into the UNHCR Health Information System database.
Third, the development of appropriate and effective behaviour change interventions to prevent and tackle obesity in these contexts will need innovative approaches. These will require health personnel and community participation in the identification of needs and implementation of solutions. Additionally, a detailed economic assessment is needed to correctly evaluate the resources needed for prevention and treatment.
Lastly, careful policy and advocacy work will be required to convey the complexity of the situation, and to ensure that continued support for life-saving food assistance programmes and the tackling of under-nutrition and nutritional deficiencies is not jeopardised as the threat of obesity to refugee health receives the attention it deserves.
French translation of the abstract by David Beran and Aurore Virayie.
(DOC)
Spanish translation of the abstract by Carlos S. Grijalva-Eternod and Alejandra J. Cantoral-Preciado.
(DOC)
Arabic translation of the abstract by Elham Aljaaly and AlBandary AlJameel.
(DOC)
We are extremely grateful to all the Western Sahara refugee families who took part in this nutrition survey, the Western Sahara refugee authorities who facilitated the implementation of the survey, and the surveyors who collected the data.
body-mass-index-for-age
body mass index
global acute malnutrition
height-for-age
non-communicable disease
United Nations High Commissioner for Refugees
weight-for-age
waist circumference
weight-for-height