Impact of a cash-for-work programme on food consumption and nutrition among women and children facing food insecurity in rural Bangladesh

854 Research Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994 Introduction Cash-for-work programmes are usually associated with disasters and emergencies and have been implemented in Afghanistan, Bangladesh, Haiti, Indonesia, Pakistan and Uganda.1 Many or- ganisations employed cash interventions after the Asian tsunami of 2004 and cash-for-work programmes were widespread in Aceh, Indonesia, and in Sri Lanka.2 However, when implemented incorrectly, these programmes can disrupt the local economy, artificially inflate wages and result in unsustainable shifts in the labour force.3 In north-western Bangladesh there is an annual period of food insecurity (the monga) that generally occurs between mid-September and mid-November and is primarily caused by unemployment and a lack of income before the large aman rice harvest.4 People who live on chars, which are large flat islands in the main river channels, are particularly affected by the monga as flooding occurs during the preceding months almost every year. When flooding is severe, there can be household damage, a loss of assets, disruption of agricultural activities and obstacles to rearing livestock. Chars dwellers may be confronted with river bank erosion, which can take away their homestead and any land they possess. The scarcity of work also has severe repercussions for the income of poor chars households and many are obliged to reduce their food intake during this period.5 The Chars Livelihood Programme (http://www.clp-ban- gladesh.org), funded by the United Kingdom Department for International Development, aims to lift over 55 000 extremely poor households out of poverty by providing income-generating assets. Households are provided with a raised earthen plinth on which their homes are reconstructed and homestead gardens can be established. In part, these plinths are created by a labour- intensive earthmoving process that involves members of poor chars households.6 In August and September 2007, widespread severe flood- ing across northern Bangladesh intensified and prolonged the impact of the monga for chars dwellers.6 In some households participating in the Chars Livelihood Programme, the cash-for- work intervention was implemented between September and December 2007 to coincide with the monga. The programme provided approximately 2.6 million person–days of paid work during which both men and women received 36 taka (about US$ 0.50) for each cubic metre of soil moved in the construction of the earthen plinths. No previous research has been conducted into the impact of cash-for-work programmes on nutritional status. In Bangladesh there was concern among some nongovernmental organisations that the physical labour required by the local cash-for-work pro- gramme would cause women to lose weight and to neglect their children because the mothers needed to work quite long hours. If this occurred, the programme would be counterproductive and would probably be detrimental to the health of the people it was intended to help. The present panel study was designed to determine whether women and children aged less than 5 years from landless households who were living on chars and who participated in a monga season cash-for-work programme Une traduction en français de ce résumé figure à la fin de l’article. Al final del artículo se facilita una traducción al español. .ةلاقلما هذهل لماكلا صنلا ةياهن في ةصلاخلا هذهل ةيبرعلا ةمجترلا Objective To determine whether a cash-for-work programme during the annual food insecurity period in Bangladesh improved nutritional status in poor rural women and children. Methods The panel study involved a random sample of 895 households from over 50 000 enrolled in a cash-for-work programme between September and December 2007 and 921 similar control households. The height, weight and mid-upper arm circumference of one woman and child aged less than 5 years from each household were measured at baseline and at the end of the study (mean time: 10 weeks). Women reported 7-day household food expenditure and consumption on both occasions. Changes in parameters were compared between the two groups. Findings At baseline, no significant difference existed between the groups. By the study end, the difference in mean mid-upper arm circumference between women in the intervention and control groups had widened by 2.29 mm and the difference in mean weight, by 0.88 kg. Among children, the difference in means between the two groups had also widened in favour of the intervention group for: height (0.08 cm; P < 0.05), weight (0.22 kg; P < 0.001), mid-upper arm circumference (1.41 mm; P < 0.001) and z-scores for height- for-age (0.02; P < 0.001), weight-for-age (0.17; P < 0.001), weight-for-height (0.23; P < 0.001) and mid-upper arm circumference (0.12; P < 0.001). Intervention households spent more on food and consumed more protein-rich food at the end of the study. Conclusion The cash-for-work programme led to greater household food expenditure and consumption and women’s and children’s nutritional status improved. Impact of a cash-for-work programme on food consumption and nutrition among women and children facing food insecurity in rural Bangladesh CGN Mascie-Taylor,a MK Marks,b R Gotoa & R Islamb a Department of Biological Anthropology, University of Cambridge, Pembroke Street, Cambridge, CB2 3RA, England. b Chars Livelihoods Programme, UK Department for International Development, Bogra, Bangladesh. Correspondence to CGN Mascie-Taylor (e-mail: nmt1@cam.ac.uk). (Submitted: 15 June 2009 – Revised version received: 1 July 2010 – Accepted: 10 August 2010 – Published online: 29 September 2010 ) Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994 855 CGN Mascie-Taylor et al. Nutritional impact of a cash-for-work programme in Bangladesh Research were able to maintain a better nutritional status than similar women and children living in the same geographical area who belonged to households not taking part in the programme. Methods Study design Over 100 000 households were enrolled in the Chars Livelihood Programme and the subsidiary cash-for-work programme was open to all members of the com- munity. The wage rate was specified in advance. The panel study involved 1009 households containing a child aged less than 5 years and an adult female (usually the mother) which were randomly select- ed from more than 50 000 taking part in the cash-for-work programme and 1051 similarly selected control households that were not taking part in the programme. The two groups were comparable in household size, the adults’ occupations and the age of the adult female family member. Several households were not in- cluded in the final analysis because the ob- served z-scores for some anthropometric measurements were outside of the accept- able ranges, which were −6 to +6 for the height-for-age z-score, −6 to +5 for the weight-for-age z-score, and −5 to +5 for the weight-for-height and mid-upper arm circumference z-scores.8 The final analysis included 1816 households (895 interven- tion and 921 control households). Tests of statistical power, carried out using SPSS Sample Power version 2 (SPSS Inc., Chicago, United States of America), showed that these numbers were sufficient to demonstrate significant differences at P < 0.01 and to provide a power of 90%, which was chosen to reduce errors. Of the 1816 households, 1800 (99%) had access to safe drinking water but had no cultivable or homestead land. Each inter- vention household contained either an adult female family member who actively worked in the cash-for-work programme (i.e. an “active” woman) or an adult female family member who did not work for the programme (i.e. a “non-active” woman) though a male household member did. There were 426 active women and 469 non-active women. In addition, all 895 intervention households contained one or more children aged less than 5 years. If there was more than one child aged less than 5 years in the household, one was selected randomly for assessment in the study. Data collection The study was carried out for the Chars Livelihood Programme by trained staff working for Helen Keller International, Dhaka, Bangladesh. Both intervention and control households were recruited be- tween mid-September and early October 2007 from the Gaibandha and Kurigram districts of north-western Bangladesh. A structured interview was completed by the same adult female at baseline and at the end of the study in all households. The study teams used standard interviewing techniques, including reverse chronologi- cal description of consumption. The ques- tionnaire used contained items on the number of days on which specific foods were consumed in the 7 days before the interview, the household’s dietary habits, morbidity, socioeconomic factors and the household’s contingency plans and coping strategies for the monga. Information on food consumption and the cost of food was collected verbally. Food costs were verified independently by research staff who visited local markets to determine the costs of the different food types men- tioned in the questionnaire. The women’s height, weight and mid- upper arm circumference were measured at both baseline and at the end of the study, which was typically in mid-Decem- ber 2007 since the mean duration on the panel was 10 weeks (range: 8–12). In ad- dition, the height, weight and mid-upper arm circumference of each woman’s child less than 5 years or of another child aged less than 5 years from the same household were also measured. All measurements were made using standard anthropomet- ric techniques.7 Although no particular provision was made for caring for children aged less than 5 years, all workers in the cash-for-work programme were assigned to a specific site in their respective villages and were given a 2-hour break during the day. Height-for-age, weight-for-age, weight-for-height and mid-upper arm circumference z-scores were determined for the children using international standards,8 and a child was classified as stunted, underweight, wasted or of low mid-upper arm circumference, respec- tively, if the z-score was < −2.0. Each woman’s body mass index (BMI) was calculated and women were categorized according to their BMI as having either grade III chronic energy deficiency (BMI: < 16.0), grade II chronic energy deficien- cy (BMI: 16.0–16.9) or grade I chronic energy deficiency (BMI: 17.0–18.49) or a normal nutritional status (BMI: ≥ 18.5).9 Data were verified by supervisors from Helen Keller International who revisited a random selection of 5% of households. Statistical analysis As this was a panel study, the analysis investigated within-woman and within- child changes between baseline and the end of the study. Sequential multiple regression analysis was used to test for changes in the mean differences in an- thropometric and nutritional variables between intervention and control groups; corrections were made for the linear and quadratic effects of age in women and for the linear and quadratic effects of age and sex in children. For categorical data, χ2 tests were used to examine differences between the groups and McNemar’s test was applied to paired samples. Results Baseline characteristics Households The baseline demographic characteristics of households participating in the study are presented in Table 1. There was no sig- nificant difference between intervention and control households in the mean age of the women or child in the household or in the proportion of children who were male, which was just over 50%. Illiteracy levels were very high in both genders, and higher in females than males. The significantly higher illiteracy rate found in intervention households was therefore partially attributable to the fact that sig- nificantly more intervention households had a female head. The main wage earner in most households was a labourer, who was usually employed on a daily basis. Women There was no significant difference in mean baseline height, weight, BMI or mid-upper arm circumference between women in intervention and control households. Overall, their mean height was 150.5 cm, their mean weight was 42.2 kg, their mean mid-upper arm circumfer- ence was 234.1 mm and their mean BMI was 18.6. Overall, 51.6% (937/1816) of all women had a BMI < 18.5. The percentages of women with different grades of chronic energy deficiency were very similar in the two groups: 30.4% and 34.6% of those in intervention and control households, respectively, had grade I chronic energy deficiency; 13.6% Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994856 CGN Mascie-Taylor et al.Nutritional impact of a cash-for-work programme in Bangladesh Research and 11.8%, respectively, had grade II chronic energy deficiency; and 7.2% and 5.7% respectively, had grade III chronic energy deficiency. In intervention house- holds, there was no significant difference between non-active women and active women in mean baseline height, weight, BMI or mid-upper arm circumference or in the percentage with different grades of chronic energy deficiency. Children No significant difference in mean base- line height, weight or mid-upper arm circumference was observed between children from intervention and control households. Overall, their mean height was 82.8 cm, their mean weight was 10.1 kg and their mean mid-upper arm circumference was 141.7 mm. Their mean height-for-age z-score was −1.98, their mean weight-for-age z-score was −1.94, their mean weight-for-height z-score was −1.17 and their mean mid-upper arm circumference z-score was −0.98. Overall, 51.3% (932/1816) of the children were stunted, 47.6% (864/1816) were under- weight, 18.6% (379/1816) were wasted and 9.4% (171/1816) had a mid-upper arm circumference z-score < −2.0. Household food There was no significant difference be- tween intervention and control house- holds in the mean quantity of pulses, eggs, fish, green leafy vegetables, meat, cereal, rice or fruit reported as being consumed in the 7 days before the baseline survey was carried out, nor was there any sig- nificant difference in mean expenditure on these food items. Changes during the study Women and children in the interven- tion group showed significantly greater improvements in all anthropometric and nutritional measures than the control group. Changes in anthropometric mea- surements between baseline and the end of the study are illustrated in Fig. 1. Women On average, women in the intervention group increased in weight and mid-upper arm circumference during the study, whereas those in the control group ex- perienced decreases: the mean increase in weight in women in the intervention group relative to those in the control group was 880 g and the mean relative increase in mid-upper arm circumfer- ence was 2.29 mm. Significantly fewer women in the intervention group had a BMI < 18.5 at the end of the study: the percentages were 48.4% and 56.6% in intervention and control groups, respectively (χ2 = 11.7; P < 0.001). Cor- respondingly, the percentages of women with grade I, grade II or grade III chronic energy deficiency were all lower in the in- tervention than in the control group (data not shown). In the intervention group, BMI improved in 9.7% of women while it worsened in 0.9%; the corresponding figures in the control group were 2.4% and 9.3% for improved and worsened BMI, respectively. Overall, 3.4% of women in the intervention group improved from having chronic energy deficiency to a Table 1. Baseline demographic characteristics of households participating in a cash- for-work programme and control households, Bangladesh, 2007 Characteristic Households P In programme (n = 895) Control (n = 921) Age of adult female family member in years, mean 25.0 25.0 0.966 Age of child in months, mean 30.2 29.4 0.295 Households with a male child, % 50.9 50.4 0.632 Households with a female head, % 8.5 5.1 0.004 Illiteracy of main earner, % 85.6 77.2 < 0.001 Usual occupation of main earner Day labourer, % 82.1 86.8 0.005 Other, % 17.9 13.2 Fig. 1. Change between baseline and the end of a panel study in anthropometric measurementsa of women and children aged less than 5 years from households in a cash-for-work programme and control households, Bangladesh, 2007 Weight, kg Control households Households in cash-for- work programmes Deteriorationa 0.32 0.00 Women Body mass index Mid-upper arm circumference, mm Height, kg Weight, kg Children < 5 years Mid-upper arm circumference, mm Height-for- age z-score Weight-for- age z-score Weight-for- height z-score Mid-upper arm circumference z-score –0.51 Improvementa 0.17 –0.22 1.14 –1.15 0.79 0.45 1.49 –0.33 0.07 –0.09 0.34 0.04 –0.08 0.72 0.24 0.10 –0.29 0.12 a Anthropometric measurements were corrected for the effects of age in women and for the effects of age and sex in children. Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994 857 CGN Mascie-Taylor et al. Nutritional impact of a cash-for-work programme in Bangladesh Research normal BMI, compared with no women in the control group. Conversely, 4.3% of women in the control group changed from having a normal BMI to having chronic energy deficiency, compared with 0.7% in the intervention group. There was no significant difference in weight gain between non-active women and active women in intervention households: the mean gains in these two subgroups were 0.95 kg and 0.80 kg, respectively. Nor was there a significant difference in the increase in mid-upper arm circumference (2.74 mm and 1.74 mm in the two sub- groups, respectively) or in the increase in BMI (0.42 and 0.36, respectively). Children During the study, children from inter- vention households gained, on average, 0.7 mm in height, 210 g in weight and 1.39 mm in mid-upper arm circumfer- ence more than those from control households, after adjustments for age and sex (Fig. 1). Among children from inter- vention households, but not those from control households, there were also signif- icant reductions in the percentages who were underweight or wasted or who had a low mid-upper arm circumference. For example, 7.3% of the children from inter- vention households improved from being underweight to having a normal weight, compared with only 3.3% of those from control households. Conversely, only 2.1% of children from intervention households became underweight, com- pared with 7.7% of those from control households (χ2 = 43.7; P < 0.001). In these two groups, respectively, 9.1% and 6.9% improved from being wasted to having a normal weight-for-height, while1.3% and 5.4% became wasted (overall χ2 = 25.6; P < 0.001). There was no significant difference between chil- dren from intervention households with active or non-active women. Household food Overall, 99.7% of intervention house- holds reported spending some of the money earned in the cash-for-work programme on food and nearly three- quarters stated that a family member had paid for medical treatment. Intervention households spent significantly more on food than control households (Table 2), with marked percentage increases in the consumption of eggs, meat, fish, milk, pulses and green leafy vegetables (Fig. 2). At the end of the study, the reported consumption of eggs, meat, fish, pulses, green leafy vegetables, milk and fruit was significantly greater and cereal consump- tion was lower among children from intervention households than among those from control households (Table 3). The differences in egg, meat and fish consumption were particularly marked. For example, 74.6% of children from intervention households had eaten fish on three days or more in the 7 days before the end of the study, compared with 2.7% of those from control households. As shown in Fig. 2, food consump- tion reported by intervention households increased between baseline and the end of the study for seven of the eight food types examined (i.e. all except cereals), while consumption by children from Table 2. Reported food expenditurea by households participating in a cash-for-work programme and control households in the 7 days before the end of a panel study, Bangladesh, 2007 Food typeb Households P In programme (n = 895) Control (n = 921) Mean (US$) SD (US$) Mean (US$) SD (US$) Cereals 4.11 2.60 1.60 1.10 < 0.001 Pulses 0.12 0.16 0.03 0.08 < 0.001 Green leafy vegetables 0.36 0.23 0.15 0.15 < 0.001 Eggs 0.05 0.10 0.01 0.04 < 0.001 Fish 0.50 0.35 0.12 0.17 < 0.001 Meat 0.06 0.27 0.01 0.09 < 0.001 Oil 0.23 0.16 0.14 0.10 < 0.001 SD, standard deviation. a Expenditure was recorded in Bangladeshi taka and converted to United States dollars. b Details of expenditure on milk and fruit were not available. Fig. 2. Change between baseline and the end of a panel study in food consumptiona by households in a cash-for-work programme and control households, Bangladesh, 2007 30.0 Control households Households in cash-for- work programmes Eggs 37.7 Ch an ge in c on su m pt io n (% ) 40.0 20.0 10.0 0 –10.0 –20.0 –30.0 –40.0 –5.0 Meat 12.1 –3.5 Fish 14.2 –22.1 Milk 13.7 –6.6 Fruit 6.3 –3.6 Cereals –9.4 Pulses 19.3 –20.3 Green leafy vegetables 36.5 –35.4 –24.8 a The change in consumption is the numerical difference between the percentage of households that had at least one day of consumption of the designated food in the 7 days before the interview at the study end and the percentage that had at least one day of consumption in the 7 days before the baseline interview. Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994858 CGN Mascie-Taylor et al.Nutritional impact of a cash-for-work programme in Bangladesh Research control households decreased for all eight food types. Moreover, these changes were all significant (McNemar’s paired test: P < 0.001 for all food types). Discussion The cash-for-work programme reported here may be viewed in the wider context of the longer-term conditional cash transfer programmes that have been increasingly used to transfer money to poor house- holds on the condition that they comply with certain requirements: for example, attending health-care sessions, using food and nutritional supplements or enrolling children in school. The few conditional cash transfer programmes that have fo- cused on nutritional status have produced conflicting results. One Brazilian study found a small negative impact,10 while a Mexican study showed that children in programmes involving both cash transfer and multi-micronutrient supplementa- tion grew about 1 cm more than those who received neither intervention.11 However, it was not possible to differen- tiate the effect of cash transfer from that of multi-micronutrient supplementation. Recent analyses of Mexican data have shown that doubling the quantity of cash transferred was associated with, for ex- ample, a better height-for-age ratio and a lower prevalence of childhood stunting.12 The results of the Bangladeshi study reported here indicate that implementing a cash-for-work programme at a crucial time of the year led to significant gains in the nutritional status of both children aged less than 5 years and women. These improvements also occurred in women who took part in the physically demand- ing cash-for-work programme. Although this was a short-term intervention, it could have longer-term benefits: selling household assets may become unneces- sary and individuals may not have to agree in advance to work for less money in the future. In addition, the earthen plinths constructed in the programme, which have a life expectancy of 25–30 years, improve household security. Although the improvements ob- served in child nutritional status are encouraging, with the main gain being less wasting due to acute malnutrition in intervention households (Fig. 1), many individuals in both intervention and control households continued to suffer from malnutrition. In a recent nationwide survey,13 39.2% of children aged less than 5 years in Bangladesh were Table 3. Reported number of days of food consumption by children from households participating in a cash-for-work programme and from control households in the 7 days before the end of a panel study, Bangladesh, 2007 Food type and no. of days consumed Households P a In programme (n = 895) Control (n = 921) % % Eggs < 0.001 0 45.5 89.3 1 28.2 7.7 2 18.4 2.2 3 4.5 0.5 4 or more 3.5 0.3 Meat < 0.001 0 80.8 96.7 1 15.0 3.1 2 4.2 0.1 Fish < 0.001 0 7.4 44.8 1 3.9 35.3 2 14.1 17.2 3 21.9 1.6 4 23.9 0.9 5 13.7 0.2 6 7.4 0 7 7.7 0 Cereals < 0.001 0 91.8 76.7 1 2.0 2.4 2 1.6 5.4 3 1.6 3.8 4 or more 3.0 11.7 Pulses < 0.001 0 25.7 62.3 1 23.2 23.2 2 27.3 11.3 3 12.5 1.8 4 or more 11.3 1.3 Green leafy vegetables < 0.001 0 4.9 7.7 1 2.7 7.8 2 13.2 22.6 3 12.7 24.2 4 19.1 19.8 5 11.6 9.6 6 10.8 2.7 7 24.9 5.6 Milk < 0.001 0 70.4 92.0 1 12.1 4.9 2 5.9 1.7 3 2.7 0.5 4 or more 8.9 0.9 Fruit < 0.001 0 87.3 97.2 1 5.6 1.4 2 3.2 0.7 3 or more 3.9 0.8 a The P-values were obtained by comparing the number of days each food type was consumed by children from households participating in a cash-for-work programme and by children from control households using the chi-squared test. Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994 859 CGN Mascie-Taylor et al. Nutritional impact of a cash-for-work programme in Bangladesh Research found to be stunted, while 45.7% were underweight and 11.9% were wasted. At the end of the study, the prevalence of underweight and wasting in intervention households resembled the national preva- lences: 43.7% and 11.4%, respectively. However, the corresponding figures in control households were worse, at 50.8% and 16.5%, respectively. Notably, the prevalence of stunting at the end of the study was much higher than the national average in both groups: it was 60.6% in children from intervention households and 64.5% in those from control house- holds. Moreover, it had increased during the study: by 10.7% in the intervention group and by 11.9% in the control group. The higher rates of chronic malnutrition may have been partly due to the seasonal changes in nutritional status that have been documented in Bangladesh13 or to increased levels of infection and morbid- ity occurring at the time of the year when the study took place.14 Although no specific nutritional education was provided, households tak- ing part in the cash-for-work programme spent more, in particular, on protein-rich foods such as fish, meat, eggs and milk, as well as on fruit, which is rich in folate, potassium and vitamin C. Detailed in- formation on the actual quantity of food consumed by women and their children was not available, as that would have ne- cessitated collecting data about portion sizes or weighing the food consumed.15 However, the level of food consumption reported is plausible. The resulting more diverse diet should lead to increased intake of vitamins important for growth, development and general health. In conclusion, our findings indicate that the cash-for-work programme led to a greater quantity and variety of food, particularly animal protein, being con- sumed during the annual period of food insecurity in Bangladesh and resulted in a significant improvement in the short term nutritional status of women and children living on chars. Even women who actively took part in the physically demanding work associated with the pro- gramme experienced an improvement in nutritional status. Over the longer term, the significant reduction of both acute and chronic malnutrition among chars dwellers will depend on lifting them out of poverty. ■ Funding: This study was funded by the United Kingdom Department for In- ternational Development as part of the research and impact monitoring activities of the Chars Livelihood Programme. Competing interests: None declared. صخلم قطانلما في ئياذغلا نملأا مادعنا نوهجاوي نيذلا لافطلأاو ءاسنلا ينب ةيذغتلا لىعو ءاذغلا كلاهتسا لىع لمعلا لباقم دقنلا جمانرب يرثأت شيدلاغنب نم ةيفيرلا ةيونسلا ةترفلا ءانثأ لمعلا لباقم دقنلا جمانرب ناك اذإ ام ديدحت ضرغلا ءاسنلل يوذغتلا عضولا نم نسحي شيدلاغنب في ئياذغلا نملأا مادعنلا .ةيفيرلا قطانلما في لافطلأاو تلائاعلا نم ددعل ةيئاوشع تانيع لىع ةعومجلما ةسارد تلمتشا ةقيرطلا لمعلا لباقم دقنلا جمانرب في ةكترشم ةلئاع 50000 نم ثركأ ينب نم 895 غلب 921 عم ،2007 ماع نم برمسيد/لولأا نوناك لىإ برمتبس/لوليأ نم ةترفلا في يولعلا فصنلا طيحمو نزولاو لوطلا سيق دقلو .دهاوشلا نم ةلثمام ةلئاع لك نم ماوعأ ةسمخ نع هرمع لقي دحاو لفطو ةدحاو ةديسل عارذلل تقولا طسوتم( ةساردلا ةياهن فيو يدعاقلا طخلا ىوتسم لىع كلذو ةلئاع ءاذغلا لىع قفنُي ام نع نغلبي ينتلاحلا في ءاسنلا تناكو .)عيباسأ 10 / تاتباثتلما ةنراقم تتم دقو .عوبسلأا نم مايأ ةعبسلا لاوط ءاذغلا كلاهتساو .ينتعومجلما ينب ينب ايئاصحإ هب دتعي فلاتخا كانه نكي لم ،يدعاقلا ىوتسلما لىع تادوجولما يولعلا فصنلا طيحم طسوتم ينب قرفلا دادزا ،ةساردلا ةياهنبو .ينتعومجلما 2.29 غلبو ،دهاوشلا ةعومجمو ةلخادلما ةعومجم في ءاسنلا ينب عارذلا نم ،لافطلأا في امأ .امارغوليك 0.88 نزولا طسوتم في فلاتخلاا ناكو ،ترم ليم ةعومجم حلاصل ًاضيأ دادزا دق ينتعومجلما طسوتم ينب فلاتخلاا ناكف نم لقأ لماتحلاا ةوق ؛ترم يتنس 0.08 غلب( لوطلل ةبسنلابف .ةلخادلما ،)0.001 نم لقأ لماتحلاا ةوق ؛مارغوليك 0.22 غلب( نزولل ةبسنلابو ،)0.05 لقأ لماتحلاا ةوق ،ترم ليم 1.41( ناك عارذلا نم يولعلا فصنلا طيحمو نم لقأ لماتحلاا ةوق ؛0.02( رمعلا-لباقم-لوطلل Z زرح غلبو ،)0.001 نم ،)0.001 نع لقأ لماتحلاا ةوق ،0.17( ناك رمعلا-لباقم-نزولاو ،)0.001 طيحمو ،)0.001 نم لقأ لماتحلاا ةوق ؛0.23( ناك لوطلا-لباقم-نزولاو فيو .)0.001 نم لقأ لماتحلاا ةوق ؛0.12( ناك عارذلا نم يولعلا فصنلا لىع ثركأ قفنت يتلا يه ةلخادملل ةعباتلا تلائاعلا ةنيع تناك ةساردلا ةياهن .تانيتوبرلاب ىنغأ ًءاذغ كلهتستو ءاذغلا لىع تلائاعلا تاقفن ةدايز لىإ لمعلا لباقم دقنلا جمانرب ىدأ جاتنتسلاا .لافطلأاو ءاسنلل يوذغتلا عضولا نسحتو ،هكلاهتساو ءاذغلا Resumé Impact d’un programme de rémunération en espèces du travail sur la nutrition et la consommation de nourriture chez les femmes et les enfants soumis à l’insécurité alimentaire dans les zones rurales du Bangladesh Objectif Déterminer si un programme de rémunération en espèces du travail lors de la période d’insécurité alimentaire annuelle au Bangladesh a amélioré l’état nutritionnel des femmes et des enfants pauvres vivant dans les zones rurales. Méthodes Le panel d’étude a concerné un échantillon aléatoire de 895 ménages sur plus de 50 000 ménages engagés dans un programme de rémunération en espèces du travail entre septembre et décembre 2007, mais aussi 921 ménages de contrôle semblables. La taille, le poids et la circonférence du bras à mi-hauteur de la femme et d’un enfant de moins de 5 ans de chaque ménage ont été mesurés au début et à la fin de l’enquête (durée moyenne: 10 semaines). Dans les deux groupes, les femmes ont indiqué des dépenses et une consommation alimentaire pour 7 jours pour le ménage. Les modifications des paramètres ont été comparées entre les deux groupes. Résultats Au départ, les deux groupes ne présentaient aucune différence significative. À la fin de l’étude, la différence dans la circonférence du bras à mi-hauteur entre les femmes du groupe d’intervention et le groupe de contrôle avait augmenté de 2,29 mm et la différence dans le poids moyen de 0,88 kg. Chez les enfants, la différence dans les moyennes entre les deux groupes avaient également augmenté en faveur du Bull World Health Organ 2010;88:854–860 | doi:10.2471/BLT.10.080994860 CGN Mascie-Taylor et al.Nutritional impact of a cash-for-work programme in Bangladesh Research groupe d’intervention comme suit: taille (0,08 cm; P < 0,05), poids (0,22 kg; P < 0,001), circonférence du bras à mi-hauteur (1,41 mm; P < 0,001) et scores z pour rapport taille-âge (0,02; P < 0,001), rapport poids-âge (0,17; P < 0,001), rapport poids-taille (0,23; P < 0,001) et circonférence du bras à mi-hauteur (0,12; P < 0,001). Les ménages du groupe d’intervention avaient dépensé plus d’argent pour acheter de la nourriture et avaient consommé davantage d’aliments riches en protéines à la fin de l’enquête. Conclusion Le programme de rémunération en espèces du travail a entraîné une augmentation des dépenses et de la consommation alimentaires des ménages. De plus, l’état nutritionnel des femmes et des enfants s’est amélioré. Resumen Influencia del programa Dinero por Trabajo en el consumo de alimentos y en la nutrición entre las mujeres y los niños que sufren una situación de inseguridad alimentaria en las zonas rurales de Bangladesh Objetivo Determinar si el programa Dinero por Trabajo mejoró el estado nutricional de las mujeres y los niños más desfavorecidos de las zonas rurales de Bangladesh durante un periodo de un año de inseguridad alimentaria. Métodos El estudio de cohortes comprendió una muestra aleatoria de 895 hogares, de entre los más de 50 000 inscritos en el programa Dinero por Trabajo, entre septiembre y diciembre de 2007, y 921 hogares de características similares como grupo de control. En cada hogar se midieron la altura, el peso y la circunferencia braquial de una mujer y de un niño menor de 5 años al inicio y al final del estudio (duración media: 10 semanas). En ambas ocasiones, las mujeres informaron sobre los gastos y el consumo alimentarios semanales de sus hogares. Se compararon los cambios paramétricos existentes en los dos grupos. Resultados Al inicio del estudio no se observó ninguna diferencia significativa entre los grupos. Al final del estudio, la diferencia en la circunferencia media braquial entre las mujeres del grupo de intervención y el de control había aumentado 2,29 mm y la diferencia en el peso medio era de 0,88 kg. Entre los niños, la diferencia de las medias entre los dos grupos también había aumentado a favor del grupo de intervención: altura (0,08 cm, p < 0,05), peso (0,22 kg, p < 0,001), circunferencia braquial (1,41 mm, p < 0,001) y las puntuaciones z de la altura/edad (0,02; p < 0,001), peso/edad (0,17, p < 0,001), peso/altura (0,23, p < 0,001) y circunferencia braquial (0,12, p < 0,001). Al final del estudio, los hogares del grupo de intervención gastaban más en alimentos y consumían más alimentos ricos en proteínas. Conclusión El programa Dinero por Trabajo tuvo como resultado un mayor gasto en y consumo de alimentos en los hogares y una mejora del estado nutricional de mujeres y niños. References 1. CARE starts “cash-for-work” programme for school cleaning in Haiti. Geneva: Reliefweb (CARE); 2008. Available from: http://www.reliefweb.int/rw/rwb.nsf/ db900sid/EDIS-7K6MQB?OpenDocument) [accessed 3 September 2010]. 2. Doocy S, Johnson D, Robinson C. Cash grants in humanitarian assistance: a nongovernmental organization experience in Aceh, Indonesia, following the 2004 Indian Ocean Tsunami. Disaster Med Public Health Prep 2008;2:95– 103. doi:10.1097/DMP.0b013e318170b5b4 PMID:18525372 3. Cuny F, Hill R. Famine, conflict and response: a basic guide. West Hartford: Kumarian Press; 1999. 4. Zug S. Monga – seasonal food insecurity in Bangladesh: bringing the information together. J Soc Stud 2006:111. 5. Conroy K, Marks M. The use of coping strategies by extreme poor households on the Jamuna Chars during monga. Bogra: Chars Livelihood Programme; 2008. Available from: http://www.clp-bangladesh.org/index2. php?option=com_docman&task=doc_view&gid=80&Itemid=99 [accessed 20 September 2010]. 6. Conroy K, Islam R, Marks M. The impact of the 2007 CLP infrastructure and employment programme. Bogra: Chars Livelihood Programme; 2008. Available from: http://www.clp-bangladesh.org/index2.php?option=com_ docman&task=doc_view&gid=37&Itemid=99 [accessed 20 September 2010]. 7. Lohmann TG, Roche AF, Martorell R. Anthropometric standardization reference manual. Champaign: Human Kinetics Books; 1989. 8. World Health Organization child growth standards: length/height-for age, weight-for-age, weight-for-length, weight-for-height and body mass index- for-age: methods and development. Geneva: World Health Organization; 2006. Available from: http://www.who.int/childgrowth/standards/Technical_ report.pdf [accessed 20 September 2010]. 9. Shetty PS, James WPT. Body mass index: a measure of chronic energy deficiency in adults. Rome: Food and Agriculture Organization; 1994. 10. Morris SS, Olinto P, Flores R, Nilson EA, Figueiró AC. Conditional cash transfers are associated with a small reduction in the rate of weight gain of preschool children in northeast Brazil. J Nutr 2004;134:2336–41. PMID:15333725 11. Berhman JR, Hoddinott J. Program evaluation with unobserved heterogeneity and selective implementation: the Mexican PROGRESA impact on child nutrition. (PIER Working Paper 02-006). Philadelphia: University of Pennsylvania; 2002. 12. Fernald LCH, Gertler PJ, Neufeld LM. Role of cash in conditional cash transfer programmes for child health, growth, and development: an analysis of Mexico’s Oportunidades. Lancet 2008;371:828–37. doi:10.1016/S0140- 6736(08)60382-7 PMID:18328930 13. HKI. Bangladesh in facts and figures: 2005 annual report of the Nutritional Surveillance Project, Dhaka. Dhaka: Helen Keller International Bangladesh; 2006. 14. HKI. Household and community level determinants of malnutrition in Bangladesh. Nutritional Surveillance project Bulletin No. 17. New York & Cambridge: Helen Keller International &Institute of Public Health; 2006. 15. Rousham EK, Mascie-Taylor CGN. Seasonality and child morbidity in rural Bangladesh. Am J Hum Biol 1995;7:369–79. doi:10.1002/ ajhb.1310070313