The relationship between food intake and non-communicable diseases (NCD), where excessive and unbalanced intakes of energy, SFA, trans-fatty acids, salt and sugar are associated with nutrition-related NCD, is well established(1). The Gulf countries have the highest reported prevalence of NCD in the world(Reference Ng, Zaghloul and Ali2). Kuwait is a Gulf country with a total land area of 17 818 km2 and a population of 3·3 million, of whom about 31 % are Kuwaiti citizens. About 80 % of adult Kuwaiti citizens(Reference Ng, Zaghloul and Ali2, Reference Al Rashdan and Nesef3) are reported to be either overweight or obese and many of them suffer from one or more nutrition-related NCD such as diabetes (15 %)(Reference Al Rashdan and Nesef3, Reference Abdella, Al Arouj and Al Nakhi4), metabolic syndrome (24 %)(Reference Al Rashdan and Nesef3), CHD, hypertension (26 %)(Reference Jackson, Al-Moussa and Al-Raqua5) and/or dyslipidaemia (33 %)(Reference Jackson, Al-Moussa and Al-Raqua5–7). Kuwait has the highest childhood overweight problem among the Gulf countries and obesity is reportedly on the rise(Reference Ng, Zaghloul and Ali2). El-Bayoumy et al.(Reference El-Bayoumy, Shady and Lotfy8) reported that 30·7 % and 14·6 % of children between the ages of 10 and 14 years were overweight and obese, respectively; while 43·3 % and 21·3 % of children between 14 and 19 years of age were overweight and obese, respectively(7).
Before the discovery of oil, Kuwait's food supply was limited. Available foods consisted mainly of rice, dates, seafood, camel milk, sheep and goat meat and their by-products(9). Meat was an indicator of wealth and the frequency of consumption of meat for the general population was weekly or monthly. Arabic coffee and milk represented main beverages while sweetened tea was served only on special occasions. Kuwait has experienced continued economic growth represented as increases in gross national income per capita (purchasing power parity)(10), gross domestic product(11) and population growth rate(12). Economic growth in Kuwait has been accompanied by an increase in food availability, mainly of imported foods, which comprise 85 % of all foods available in the market. Traditional foods have been replaced by foods typical of the Western diet through fast-food outlets and restaurants(9). Increased food availability and reduced food prices (due to government subsidy) have contributed to increased energy intakes(Reference Musaiger13). Increased daily consumption of energy-dense, nutrient-poor foods, increasing frequency of snacks and meals consumed away from home, shifts from drinking water and milk to drinking sugar-sweetened beverages and increased portion sizes are among the most commonly reported dietary changes associated with the nutrition transition, consequently increasing rates of nutrition-related NCD(Reference Popkin14–Reference Astrup, Dyerberg and Selleck20). Increased prevalences of obesity and nutrition-related NCD occur with increased intakes of energy, sugar, fat and protein and decreased intake of fibre(1, Reference Ng, Zaghloul and Ali2, Reference Popkin14–Reference Astrup, Dyerberg and Selleck20).
Except for a few studies reporting low fruit and vegetable intakes or increased consumption of fast foods and sugary snacks among Kuwaitis, mainly in children(Reference Zaghloul, Waslien and Al Somaie21–Reference Al-Ansari, Al-Jairan and Gillespie24), there is a paucity of national data on dietary intakes and assessment of compliance with dietary guidelines. Therefore the aims of the current study were to describe nutrient intakes and the prevalence of overweight and obesity in a nationally representative sample of Kuwaitis and to investigate the dietary determinants for nutrition transition by comparing dietary data with reference intake guidelines.
Experimental methods
Sample
This is a random representative national sample of Kuwaiti households. Kuwaiti households from the six governorates (Al Asema, Hawalli, Al Jahra, Al Farwania, Al Ahmadi and Mubarak Al Kabeer) were divided into eighty-two localities proportionate to Kuwaiti population density. Each locality was divided into clusters. Clusters of twenty households were selected using stratified sampling. Out of the total 5418 households contacted, 2862 households agreed to participate with a response rate of 53 %. At the household level, 1830 individuals (48 % males and 52 % females) were randomly selected from 545 separate Kuwaiti households from all six geographical strata, taking into consideration census gender distribution and age category. A screening form was completed per household and included demographic and socio-economic data and household composition. Data obtained were used to identify randomly selected individuals to survey from each age category and gender. Male household heads were recruited from odd-numbered households, while female household heads were selected from even-numbered households. In the case of having more than one participant within the same age group, both were selected if they were of different genders, or the male with odd serial number or the female with even serial number if both were of same gender. Interviews were conducted at seven primary health-care clinics of the Ministry of Health located at various districts of Kuwait during the period July 2008 to November 2009 with a response rate of 24 %.
For the current paper, socio-economic, anthropometric and dietary data were analysed on a subsample of 1704 participants between the ages of 3 and 86 years for whom 24 h recalls were completed. Socio-economic, health and dietary data were collected on 655 children aged 1–18 years and 1049 adults aged ≥19 years. Anthropometric measurements and blood indices were taken. The study was approved by the Ethics Committee of the Kuwait Ministry of Health. Two consent forms were obtained, one from the head of the household and the other from each participant, including children, selected at the household level. Parents signed on behalf of their children. Consent forms were written in Arabic, as were the questionnaires. A raking method was used to calculate non-response-adjusted weights, producing a final set of person weights to perform data analyses.
Anthropometric data
Weight was measured using a Tanita 310 body composition analyser for children aged ≥7 years and adults, while SECA scales were used for weight measurements of children <7 years of age. Measurements were taken to the closest 100 g. A SECA 416 infantometer was used to obtain the length measurement of infants, while a SECA 214 stadiometer was used for measuring the height of older children and adults to the nearest 1 cm.
BMI (kg/m2) was calculated by dividing weight in kilograms by the square of height in metres. Overweight and obesity were defined based on WHO standards. For adults, overweight was defined(25) as BMI ≥ 25·0 kg/m2 and obesity as BMI ≥ 30·0 kg/m2. For participants ≤5 years of age, overweight was defined as BMI Z-score ≥2 and obesity as BMI Z-score ≥3 using WHO standards(26). For participants from 6 to 19 years of age(Reference de Onis, Onyango and Borghi27), overweight was defined as BMI Z-score ≥1 and obesity as BMI Z-score ≥2.
Dietary data
A single 24 h recall was collected from participants over 2 years of age using the multiple-pass method(Reference Conway, lngwersen and Vinyard28, Reference Conway, Ingwersen and Moshfegh29) developed by the US Department of Agriculture and a food instruction booklet developed to standardize dietary data collection and reflect cultural and traditional eating behaviours(Reference Zaghloul, Ilyan and Al-Hamly30). In addition, food photographs and household measures (cups, spoons, etc.) were used to estimate portion sizes of foods. Food photographs were developed at the Kuwait Institute for Scientific Research where foods were cooked, served, weighed and photographed. Experienced dietitians were trained to collect the dietary data using the multiple-pass method and food instruction booklet. Mothers and/or other family members who were knowledgeable about the child's food intake were asked to provide types and quantities of food and beverages the child had consumed within the 24 h period preceding the interview. Adolescents responded for themselves. Given the lack of experience in collecting dietary 24 h recalls in Kuwait and the cultural context, we tested the quality of dietary data during the fieldwork for indicators of completeness and reasonability using a sample of 737 adults. For each adult participant, energy intake (EI) and BMR were calculated, the latter using the Schofield equations(Reference Schofield31), and the ratio EI:BMR was determined. A cut-off value for EI:BMR of <0·9 was used to classify participants as under-reporters(Reference Harrison, Galal and Ibrahim32, Reference Briefel, Sempos and McDowell33). Furthermore, the estimated energy requirement (EER) was calculated according to the US Dietary Reference Intakes (DRI)(34) from each participant's age, weight, height, gender and physical activity, and the ratio EI:EER computed. EI:EER provides an additional comparison capturing significant under-reporting or over-reporting. These analyses revealed that 89 % of males and 62 % of females were adequate reporters. In addition, mean EI:BMR was 1·55 for adult females and 1·26 for adult males, denoting a mean underestimation percentage within acceptable cut-off points. We concluded that under-reporting is not likely to be a problem with the current study and in this cultural context.
Dietary data processing
The ESHA Food Processor software version 10·3(35) was used for dietary data entry after adding 103 chemically analysed local Kuwaiti composite dishes to the software database(Reference Al-Amiri, Al-Otaibi and Al-Foudari36–41). Recipes were created for an additional forty-six local foods reported in the 24 h recall. The nutrient composition of twenty breads and bakery products was obtained from Kuwait Flour Mills Company's nutrient analysis laboratory. Many commonly eaten food products in the market were investigated for nutrient content and adjusted in the nutrient database to reflect food fortification and enrichment in the Kuwaiti market. In the case of missing nutrient data for unanalysed food items, the closest similar food in the ESHA Food Processor program was selected to impute the missing nutrients.
The 24 h recalls were coded and foods closest in description and nutrient content were selected from US dietary databases. After initial data entry, a second coder checked each recall to verify accuracy. All recall surveys were rechecked by survey research supervisors. Dietary recalls were analysed for macronutrients, nine vitamins and seven minerals.
Nutrient intakes were compared with the US DRI since Kuwait does not have its own dietary guidelines. Kuwaitis’ nutrition status was evaluated when appropriate against the Estimated Average Requirement (EAR), Adequate Intake (AI) and Acceptable Macronutrient Distribution Range (AMDR) developed by the Institute of Medicine's Food and Nutrition Board(34, 42–48) for all reported nutrients. The percentages of consumers meeting 100 % or more of the appropriate DRI were calculated.
Statistical analysis
The SPSS statistical software package version 16 was used to calculate weighted means and standard errors of the nutrient intakes using individual weights developed based on census data and taking into account the complex sampling design. The percentage of energy from macronutrients and the percentage of participants who over-consumed energy and nutrients by age category and gender were estimated. Student's t test was used to compare between male and female participants. The percentage of overweight and obese participants was estimated by age and gender. ANOVA was performed to test differences in mean intake by social factors at 95 % confidence interval with a significance level of P < 0·05.
Results
Description of the population
As shown in Table 1, more than half of all adult participants were female (55·3 %) and most (74·1 %) were married. Some 52·9 % of adult participants had a monthly income of 1000–2000 Kuwaiti Dinars (approximately $US 3500–7000), 7·4 % were either illiterate or functionally illiterate (can read and write but with no formal education), 44·1 % were salary employed in either the government or private sector, and 23·8 % of them worked at home (including housewives).
Table 1 Characteristics of the sample of Kuwaiti adults (n 1049), 2009–2010
KD, Kuwaiti Dinars.
Prevalence of overweight and obesity
Weighted data revealed that mean BMI increased with age among males and females (Table 2). Overweight and obesity were more prevalent among male children (≤18 years) compared with female children especially the 9–13 years age group. More than two-thirds of the adults were either overweight or obese, reaching a high of 78 % and 93 % in males and females aged ≥51 years, respectively.
Table 2 Prevalence of obesity, overweight and over-consumption of energy and macronutrients by age group and gender: nationally representative sample of Kuwaitis, 2009–2010
EER, estimated energy requirement; AMDR, Acceptable Macronutrient Distribution Range.
Nutrient intakes
Table 2 shows that a large percentage of Kuwaitis exhibited over-consumption of energy. However, that over-consumption decreased with age: 31·5–72·6 % of children and 15·5–31·4 % of adult participants exceeded the recommended energy requirements. The daily energy intake of children ranged from 5347 to 11 068 kJ (1278–2645 kcal), while adults consumed between 6078 and 10 524 kJ (1454–2515 kcal). Males consumed more energy than females. Although macronutrient (protein, carbohydrates, fat) consumption fell within the AMDR for most ages, almost a third of the sample exceeded the upper limit of the AMDR for fat (35 % of total energy) except for children aged 1–3 years and adults aged ≥51 years. On the other hand, ≥51-year-old adults exceeded the upper limit of the AMDR for carbohydrates. Marked under-consumption of n-3 and n-6 fatty acids occurred across all age and gender groups, showing no gender differences.
Table 3 shows that the ranges of mean intake of carbohydrates, protein and fat for females were 184–264 g, 48–71 g and 44–75 g, respectively; corresponding values for males were 175–355 g, 48–110 g and 45–96 g. The mean intake of cholesterol was less than the recommended 300 mg, except for males aged 19–50 years. Mean fibre intake was less than the recommended AI value for all age groups among both males and females. Mean intake of fibre increased with age, with about a third of males and females aged ≥51 years consuming 100 % or more of the recommended amount of fibre for their age.
Table 3 Weighted means, standard errors and percentage consumption of macronutrients in excess of the Dietary Reference Intakes† by age group and gender: nationally representative sample of Kuwaitis, 2009–2010
EAR, Estimated Average Requirement; ND, not defined.
†Percentage of participants who consumed ≥100 % of the EAR.
‡Percentage of participants who consumed ≥100 % of the Adequate Intake for fibre (g).
§Percentage of participants who consumed ≥300 mg/d.
Excluding intakes of vitamins and minerals obtained from supplements, Figs 1 and 2 show the general inadequacy of nutrient intakes among adults and children. Recommended intakes of vitamin E, fibre, folate and Ca were met by less than half of the adult population with a significantly low intake of most nutrients by women. Fewer than 10 % of adults met the EAR for vitamin E, and less than 50 % for fibre, Ca, folate and vitamin C. Similarly, low intakes were observed among children; however, significant gender differences were observed only for vitamin B12, Zn, Ca and P. More adult Kuwaitis met the recommended EAR values for Zn, P and Mg than did children.
Fig. 1 Percentage of Kuwaiti adults meeting the Dietary Reference Intakes (1 = Estimated Average Requirement, 2 = Adequate Intake; 3 = estimated energy requirement, 4 = Acceptable Macronutrient Distribution Range) by gender (——, males; – – –, females), 2009–2010. Significance levels for comparisons by gender: *P < 0·05, **P < 0·01, ***P < 0·001, ****P < 0·0001
Fig. 2 Percentage of Kuwaiti children meeting the Dietary Reference Intakes (1 = Estimated Average Requirement, 2 = Adequate Intake, 3 = estimated energy requirement, 4 = Acceptable Macronutrient Distribution Range) by gender (——, males; – – –, females), 2009–2010. Significance levels for comparisons by gender: *P < 0·05, **P < 0·01, ***P < 0·001
Table 4 shows mean intake of vitamins by age category and gender. Results revealed extremely low percentages of participants consuming 100 % or more of the EAR for vitamin D, vitamin E and folate (<2 % for vitamin D and <10 % for vitamin E). Prevalence of low serum levels of 25-hydroxyvitamin D (<25 nmol/l) was also found among the participants (17 % for males and 33 % for females), with the prevalence being significantly higher among females than males (data are not shown).
Table 4 Weighted means, standard errors and percentage of consumption of vitamins in excess of the Dietary Reference Intakes by age group and gender: nationally representative sample of Kuwaitis, 2009–2010
RE, retinol equivalents; EAR, Estimated Average Requirement.
†To convert to μg, divide IU by 40.
Conversely, mean intakes of vitamin A, vitamin C, thiamin, riboflavin, niacin, vitamin B6 and vitamin B12 were greater than the EAR values across all age categories in both males and females. Vitamins A, C and B12 were measured in serum and the majority of participants had all vitamin and trace element levels within the normal range (data not shown).
Table 5 shows the mean intake of minerals by age category and gender. Mean Fe intake for males was two to three times the recommended value. On the other hand, the mean Ca intake was markedly less than the recommended amount for both males and females across different age groups. The lowest mean Ca intakes were 543·8 mg and 518·7 mg for females aged 9–13 years and 14–18 years, respectively.
Table 5 Weighted means, standard error and percentage consumption of minerals in excess of the Dietary Reference Intakes by age group and gender: nationally representative sample of Kuwaitis, 2009–2010
EAR, Estimated Average Requirement; UL, Tolerable Upper Intake Level.
†The EAR for Ca is 800 mg for women aged 19–30 years and 1000 mg for women aged 31–50 years.
†The UL for Na was used to estimate the percentage of participants exceeding the safe Na intake level.
Females 14–18 years of age reported the lowest percentage of those consuming 100 % or more of the EAR for Ca, Mg and P, although 50 % of them consumed 100 % or more of their energy requirement. Mean intake of Zn was low only among females aged 14–18 years (6·8 mg). Mean intake of Na exceeded the Tolerable Upper Intake Level for all age groups. Salt added at the table was not quantified in the survey.
Tables 6–8 show mean macronutrient, vitamin and mineral intakes by education, occupation, marital status and family income among adults. Males consumed significantly more nutrients than females. Mean nutrient intakes were significantly highest among the self-employed compared with other occupations and lowest among illiterate and functionally illiterate participants compared with those of higher educational attainment. Surprisingly, mean nutrient intakes did not differ by family income except for riboflavin and niacin. Mean nutrient intakes differed significantly by marital status but the direction of differences was inconsistent.
Table 6 Unweighted mean macronutrient intakes of adults according to demographic characteristics: nationally representative sample of Kuwaitis, 2009–2010
KD, Kuwaiti Dinars.
P < 0·05.
Table 7 Unweighted mean vitamin intakes of adults according to demographic characteristics: nationally representative sample of Kuwaitis, 2009–2010
RE, retinol equivalents; KD, Kuwaiti Dinars.
P < 0·05.
Table 8 Unweighted mean mineral intakes of adults according to demographic characteristics: nationally representative sample of Kuwaitis, 2009–2010
KD, Kuwaiti Dinars.
P < 0·05.
Discussion
The present study is the first to our knowledge to describe nutrient intakes of a nationally representative sample of Kuwaitis based on actual food consumption and not on estimated food availability. The large percentage of the population which over-consumes energy and energy-yielding nutrients, coupled with the high prevalence of obesity and overweight, clearly show that the nutrition transition is extant in Kuwait(Reference Sibai, Nasreddine and Mokdad49). Almost half of the population consumed more than their energy, protein and carbohydrate requirements. The range of mean energy intakes was 5347–11 068 kJ (1278–2645 kcal) for males and 5595–8336 kJ (1337–1992 kcal) for females. As expected, these empirically derived findings are lower than those estimated in the FAOSTAT data set for total energy (13 010 kJ or 3108 kcal), carbohydrates (100 g), protein (92 g) and fat (114 g)(Reference Popkin14, Reference Musaiger50). This difference is due the fact that FAOSTAT is an estimation of energy intake based on aggregated data sources. We expect our data set to contain age and gender differences contributing to the observed discrepancies in addition to other uncalculated losses.
The prevalence of overweight and obesity was high across almost all age groups, with the highest prevalence being seen among women ≥20 years of age. This finding is consistent with previous reports(Reference Ng, Zaghloul and Ali2, 7, Reference Al-Kandari51). Our results showed a slightly higher prevalence of obesity among adults in the same age range compared with those reported by Al Rashdan and Nesef(Reference Al Rashdan and Nesef3), who found 36·4 % of adult males to be obese and 47·9 % of adult females. For children and adolescents, obesity prevalence ranged from 7 % to 37 % for males and from 10 % to 24 % for females. The comparison of childhood obesity rates with those found in previous studies is difficult because of the use of different cut-offs and standards. However, it is worth noting that Ng et al.(Reference Ng, Zaghloul and Ali2) reported increased obesity prevalence for all children in Kuwait compared with the other Gulf countries and estimated an increased rate of 1·7 % and 3·9 % among women and men between 30 and 60 years of age, respectively.
Our study also showed over-consumption of Na and cholesterol at an early age, which may help to explain the high prevalence and early onset of CVD among Kuwaitis(Reference Jackson, Al-Moussa and Al-Raqua5, Reference Jackson, Al-Moussa and Al-Raqua6). Additionally, the high Na intake is consistent with the reported increased consumption of fast foods and French fries among children and the high Na content shown in the nutrient composition data of dishes commonly consumed in the Arab Gulf countries(Reference Musaiger, Takruri and Hasan52).
Despite their increased energy intake, more than 90 % of Kuwaiti adults and children did not consume 100 % or more of the EAR for vitamin D, vitamin E, n-3 and n-6 fatty acids; more than two-thirds did not meet the EAR for Ca, Mg, folate and fibre; and half did not meet the EAR for vitamins A and C and Zn. This indicates low-nutrient-dense food choices. The low intakes of Ca, Zn, vitamin D and folate found in the present study are consistent with findings from college women in Kuwait(Reference Al-Shawi53) and from adolescents in Saudi Arabia(Reference Washi and Ageib54, Reference Sadat-Ali, AlElq and Al-Turki55) and Qatar(Reference Bener, Al-Ali and Hoffmann56). Moreover, increased prevalence of osteoporosis has been reported in Kuwait(Reference Dougherty and Al-Marzouk57) and Saudi Arabia(Reference Sadat-Ali, AlElq and Al-Turki55, Reference Naeem58). The US National Health and Nutrition Examination Survey 2001–2002 report identified vitamins A, E and C as potential problems for most gender/age groups based on comparisons with EAR values(Reference Moshfegh, Goldman and Cleveland59). Vitamin B6 was considered a potential problem for older adult females, Zn for older adult males and females and teenage females, and P for pre-teens and teenage females. These findings are comparable to the results of the current study.
Unlike a previous report showing low Fe intake among anaemic and non-anaemic girls aged 14–20 years in Kuwait (10·6 and 10·8 mg/d, respectively)(Reference Al Mousa, Prakash and Jackson60), the current study showed a mean Fe intake of 11·7 mg/d for the same gender and age group.
Similar to our findings, the British Columbia Nutrition Survey reported that income level did not affect nutrient intakes while educational attainment affected only the intakes of vitamins C and B12(Reference Forster-Coull, Milne and Barr61). On the contrary, the trend in nutrient intake among Mexican Americans between 1982 and 2006 showed increased energy and fat intakes and decreased total protein and carbohydrate intakes with increasing income and educational attainment(62). The lack of effect of income on food intake among Kuwaitis could be explained by low food prices and current subsidy policies. All Kuwaiti citizens are eligible to subsidized food. Subsidized food items include polished white rice, refined table sugar, lentils, tomato paste, powdered full-fat milk or long-life liquid full-fat or skimmed milk, infant formula, instant cereal, vegetable oil, frozen chicken and full-fat spread cheeses. The current Kuwaiti consumption patterns highlight the importance of revisiting subsidy policies to gear them towards subsidizing healthy foods. There is a need to substitute energy-dense foods (oil, sugar, ghee) with more nutrient-rich, Ca-rich fruits and vegetables.
Limitations
Estimation bias of the dietary measurements cannot be excluded because of the present study's reliance on the assessment of a single 24 h recall. Multiple 24 h recalls would be necessary to confirm individual intakes. Privacy of held food and cultural beliefs was a real challenge to dietary data collection and limited our ability to collect a second 24 h recall even for a subsample. However, the results of biochemical tests revealed that the majority of participants had all vitamin and trace element levels within the normal range. The detailed blood sample analysis and its association with dietary indices are beyond the scope of the current paper.
Conclusions
Kuwait is a country with a food-abundant environment. Food is available at low cost and Kuwaitis have low nutrition awareness. In this context Kuwaitis are experiencing a nutrition transition that is evidenced by increased prevalences of obesity, overweight and nutrition-related NCD and high consumption of foods that are energy dense, high in macronutrients and low in fibre and micronutrient density. The present study quantifies food intake in a representative sample of Kuwaiti individuals. As such, it represents a critical step in understanding how to modify dietary intake to reduce the prevalence of nutrition-related NCD in Kuwaitis. There is an urgent need to increase nutrition awareness of healthy food choices and to conduct interventions aimed at modifying subsidy policies in Kuwait and limiting Na intake.
Further analyses are needed to identify high energy and Na sources in the Kuwaiti diet and to estimate the differential contributions of Western v. traditional foods (such as soft drinks and fruit drinks v. milk and fruit juices) to energy and nutrient intakes and to determine whether Kuwaitis are meeting the recommended intakes of fruit and vegetables.
Acknowledgements
Sources of funding: The study was supported in part by grant #2003-1202-02 from the Kuwait Foundation for the Advancement of Science; the Kuwait Supreme Council for Planning and Development; and the UN Development Programme. Conflicts of interest: The authors declare they have no conflicts of interest. Authors’ contributions: S.Z. conceived of and designed the dietary assessment tools and questionnaires for the specific age groups, developed the dietary assessment database, verified the field survey dietary data collection, conducted the analysis and interpretation of data, and wrote the manuscript. S.N.A.-H. conceived of the study, obtained the funding grant, ensured the provision of materials, analysis tools and consultations needed for dietary and anthropometric measurements, and was responsible for data compilation and interpretation. N.A.-H. participated in the experimental design and ensured the provision of all facilities – including study sites, health clinics and team members from the Ministry of Health – needed for receiving household members, completing the survey questionnaires and taking measurements. S.A.-Z. and H.A. recruited and supervised the field survey teams required for recruitment of clusters of households from different localities and line listing of household individuals, and developed the database of recruited households for demographic data and line listing of household members. I.I. and H.A.-A. were responsible for entry and coding of all 24 h dietary data and dietary data processing and tabulation. A.A.-O. developed and maintained the study database by developing a web-based demographic template for recruited households and a template of age-specific questionnaires for daily upload of completed questionnaires, and assisted in data cleaning and data analysis. E.A.-S. and M.A.-S. were responsible for supervision and setting of criteria for anthropometric data collection and interpretation. R.T.J. reviewed and edited the manuscript. Acknowledgements: The authors express their gratitude to the management of the Kuwait Institute for Scientific Research and the Ministry of Health for support of their scientific pursuit. Thanks are also extended to the funding bodies for their contribution towards the project.
