Assessment of nationally representative dietary studies in the Gulf Cooperation Council: a scoping review

Introduction

Obesity is an epidemic in the countries of the Gulf Cooperation Council (GCC) (that is, Saudi Arabia, Bahrain, Kuwait, Oman, Qatar, and United Arab Emirates). Approximately one out of every three adults is obese (Body Mass Index ≥30), and the obesity prevalence has been rising in every member country. For example, between 2011 and 2016, the obesity prevalence rose in Saudi Arabia (KSA) from 32.1 to 35.4%, in Bahrain from 27.1% to 29.8%, in Kuwait from 35.1% to 37.9%, in Oman from 23.7% to 27%, in Qatar from 31.8% to 35.1%, and in the United Arab Emirates (UAE) from 28.3% to 31.7% (Global health observatory, 2017). Apart from obesity, the GCC countries are also leading countries in the world in diabetes and cardiovascular disease prevalence (M Alqarni, 2016; Alqurashi, Aljabri & Bokhari, 2011; Aljefree & Ahmed, 2015).

There is mounting evidence of a potential causal link between specific dietary factors (such as, fruit, vegetable, processed meat, and trans-fat intake) and the above mentioned chronic conditions (Micha et al., 2017; World Cancer Research Fund, 2018; World Cancer Research Fund, 2010). A recent systematic review of dietary data from 195 countries found that 22% of all adult deaths worldwide are due to unhealthy diet; more than half of diet-related deaths are attributable to a high sodium intake, low intake of whole grains, and low fruit intake (GBD 2017 Diet Collaborators, 2019).

Several factors likely contribute to obesity in GCC countries. With increased wealth from oil reserves, these countries have seen rapid economic growth. The urbanization of the landscape has seen a rise in international fast food chains, making it easier and quicker to consume processed foods (Alnohair, 2014; Al-Mahroos & Al-Roomi, 1999). This has resulted in a change of diet from traditional, locally produced goods such as wheat, vegetables and dates to fast foods high in fat, sugar and salt content (Al-Othaimeen, Al-Nozha & Osman, 2007). Whist all GCC countries have attempted to develop a national plan that addresses nutrition and physical activity, most have not followed up, which makes it difficult to evaluate the impact of such programs (Samara, Andersen & Aro, 2019). Changes in lifestyle such as increased use of cars, electrical home appliances, television and gaming devices have resulted in a more sedentary lifestyle (Alnohair, 2014; Musaiger, 2004). The extremely hot climate found in these countries also likely deters outdoor activities with people opting to use cars, even for short journeys (Alnohair, 2014; Al-Kandari, 2006). A combination of all these is likely to play a role in the current epidemic.

Given the high prevalence of chronic conditions in the GCC, one would expect that these countries engage extensively in diet and nutrition research. However, dietary studies have been limited; only approximately 1% of global dietary research has come from Arab countries (Sweileh et al., 2014). Their h-indices [measurement of performance by combining productivity (number of papers) and impact (number of citations)] are much lower than neighbouring non-Arab countries (Hirsch, 2005).

One would similarly expect that assessment tools used in dietary studies from GCC countries would differ from those in European or North American studies as Middle Eastern diets vary a great deal from their Western counterparts. For example, date palm fruit is highly consumed in Gulf regions, with daily consumption ranging from 68–164 g daily (Al-Mssalle, 2018; Ismail et al., 2006; Aleid, Al-Khayri & Al-Bahrany, 2015), whereas only 140 g of this fruit is consumed annually in Europe (Ordines, 2000). Differences such as these should be accommodated for when designing dietary assessment tools.

The usual assessment tools used in dietary research are 24-hour dietary recall (open-ended, food consumed the previous day, conducted by trained interviewer), diet records (open-ended, participants trained to record own diet), and food frequency questionnaires (FFQs) (closed-ended, typically a food list and frequency of consumption in a given period). All have strength and limitations (Shim, Oh & Kim, 2014), but due to low cost, low respondent burden and ease of use compared to other methods, FFQs are thought to be the best choice for measuring habitual diet in large populations. The usefulness and reliability of FFQs have been demonstrated with strong correlations with diet records (Rimm et al., 1992; Willett et al., 1985), dietary recalls (Katsouyanni, 1997; Brunner et al., 2001; Marques-Vidal et al., 2011), and objective biomarkers of diet (Katsouyanni, 1997; Brunner et al., 2001). As an FFQ is a self-reported subjective tool, FFQs should be tested for validity alongside a reference tool.

The authors’ aimed to conduct both a quantitative and qualitative review of all dietary studies conducted within each GCC country. To be as nationally representative as possible and to provide a current and more reflective picture of diet in the GCC, only studies carried out in multiple regions (must be a minimum of two regions) were included. Dietary research that used FFQs in individual GCC countries (Bahrain, Kuwait, Oman, Qatar, KSA, UAE) over the past ten years (2009–2019) were assessed. The characteristics of the studies were described and their quality was assessed using a widely accepted scoring tool (Herzog et al., 2013; Stang, 2010). The objectives were to (1) identify multi-regional GCC dietary studies that used FFQs, (2) assess the quality of the studies, and (3) offer recommendations for future dietary assessments.

Method

Exclusion of studies

Studies were excluded if they (1) examined data from only one specific region/city/population group and therefore were not necessarily nationally representative, (2) were multi-national studies that did not give Gulf-nation-specific results, (3) were not conducted in a GCC country, (4) were intervention studies where the diet had purposefully been changed, (5) were review or meta-analysis papers, (6) used an assessment tool other than a food frequency questionnaire, or (7) had no findings related to diet or did not report those findings. Therefore, the final analysis was limited to seven dietary studies (Fig. 1).

Results

Discussion

With such a high prevalence of diseases to which diet is a major contributor, it is surprising that there are so few multi-regional studies that investigated diet in the GCC in the past ten years. Five out of the seven studies included in this review did not use validated FFQs.

Dietary summaries show intake of fruit and vegetables being far below the recommended three servings of vegetables and two servings of fruit per day (Ministry of Health Saudi Arabia, 2012). In Saudi Arabia, only 5.2% of individuals met the recommendation for fruit intake and 7.5% for vegetable intake. In contrast, consumption of sugary beverages was oversubscribed, with an average of 36% of adolescents (14–19 years old) reporting daily consumption (Al-Hazzaa et al., 2011) and 27% of 15–60 year olds (Moradi-Lakeh et al., 2017), exceeding local and global recommendations of sugary-drink consumption (Eaton et al., 2012; Al Qaseer, Batarseh & Asa’ad, 2007; AlBuhairan et al., 2015). This low fruit and vegetable intake, combined with high sugary-beverage consumption, suggests a poor-quality diet across the GCC.

The varying methods of measuring diet made it difficult to compare consumption. For example, 57% of the studies assessed diet using frequency questions (how often), whilst 43% measured frequency and quantity (portions or serving size). At times, the response categories were too broad for in-depth analysis. For example, “Do you regularly consume meals? Yes/No” (Musaiger et al., 2011) does not specify which meals, how many meals, or the content of the meals. Similarly, “How often do you drink a glass of milk?” (Alsheridah & Akhtar, 2018) does not quantify the size of the glass or the amount of milk consumed.

Adaptability was one of the main issues relating to study quality according to the quality assessment scoring scale. Studies need to make it explicit how they have categorized foods, for example, whether they have classified potatoes as starch, tuber, snack, fast food, etc. Four studies attempted to include local foods, with a maximum of two or three items added (and mentioned in the article) (Donnelly, Fung & Al-Thani, 2018; Moradi-Lakeh et al., 2017; AlBuhairan et al., 2015; Al Baho & Badr, 2011). It is concerning that the other three studies did not mention any native foods at all. In Tabacchi’s review (Tabacchi et al., 2014), it is suggested that an FFQ with less than 70 food items reduces the quality of nutritional information that can be deduced. None of the studies included in this review had 70 items; the most was 20, the average being 11 items. Nutritional status and dietary patterns differ over time and from region to region; without the incorporation of local foods and without categorizing them under more common food groups, it is entirely possible to mask important epidemiological links between diet and disease.

An overall poor validity of FFQs was found in this review. Only one study used a validated Arabic FFQ and scored three points out of a possible four points on the quality assessment scale. Validation in large-scale studies is especially important as FFQs are prone to measurement errors and come with inherent self-bias. FFQs rely on an individual’s memory and his/her own perception of food sizes, thus under-reporting remains a common problem (Subar et al., 2003; Beaton et al., 1979; Freudenheim & Marshall, 1988; Kipnis et al., 2003). Researchers have made extensive efforts in the last two decades to mitigate some of the errors with self-reporting data (Freedman et al., 2014; Freedman et al., 2010; Kipnis et al., 2002), but diet and eating patterns are complex, and FFQs are still thought to have clear value and insight that solely objective measures cannot provide (Kirkpatrick et al., 2017; Subar et al., 2015). One of the ways to minimize errors is to use a validated FFQ. FFQs are not one-size-fits-all, and it is integral that questionnaires be adapted/modified to suit the population with which they are being used. This includes first developing a good FFQ to standard procedure (Willett, 2012), FFQs being in the native language, which for GCC is predominantly Arabic, and including as many local foods as possible.

Within obesity research, two areas are deficient: understanding the role of dietary habits in the obesity epidemic and sufficient intervention studies on weight loss via dietary change. Research on dietary habits in the obesity epidemic may be lacking due to a shortage of skilled researchers and research centers (Samara, Andersen & Aro, 2019). Obtaining accurate dietary data requires specialized nutritionists/dieticians and controlled research settings, but this is a problem across many Gulf states, where it is difficult to have sufficient numbers of well-trained staff to serve large populations and areas like Saudi Arabia (Table 1). Investments should be made in specialized university health education and research courses and training in hospital departments; this will take time and resources but is a necessary step to produce expert personnel that can adequately face the challenges of regional obesity research (Samara, Andersen & Aro, 2019).

Research in this field may also be looking at risk factors found in Western countries and not necessarily exploring factors that are unique to the socio-cultural environment of the GCC. For example, women have been shown to be less active than men across Gulf countries (Al-Nozha et al., 2007; Mabry et al., 2009) and more sedentary than their British counterparts (Al-Hazzaa et al., 2013; Al-Nakeeb et al., 2012), but reasons for this behavior was poorly understood. Only by exploring the socioeconomic, environmental and cultural contexts further was it understood that the greatest barrier to physical activity for women was a lack of facilities rather than assumed low levels of knowledge, dress codes (Samara et al., 2015) or high obese-body acceptance (Wills et al., 2006). Samara et al. suggested that future health strategies should focus on providing culturally sensitive exercise facilities for women (Samara et al., 2015). A similar approach needs to be taken for nutrition and diet, where interventions, based on survey results, acknowledge and work with, not against, local culture and social norms (O’Dea, 2008). Such intervention studies need to have tangible goals, clear action plans and sufficient follow-up to evaluate long-term effectiveness (Lawton et al., 2006).

Limitations of this review are that the search was carried out on four main databases; this may have missed studies published in other journals not found within these databases, and those that are currently underway or not yet published. However, additional cross-checking was performed with reference lists to ensure the maximum number of studies were screened. The small number of studies limited the generalizability of findings. To the authors’ knowledge, there are no other reviews similar to the current study. There are studies that have looked at other methods for country-specific dietary assessment (Kirkpatrick et al., 2017) and the Newcastle-Ottawa Assessment Scale, which was adapted for this study, has been used to assess study quality but not in the same context as the current study (Herzog et al., 2013; Modesti et al., 2016). Finally, although studies have looked at dietary research from other parts of the world, no study has quantified the number of dietary studies coming specifically from the GCC and assessed their quality. Our review is unique in these ways, so the results of this present study cannot be easily compared to other studies.

A particular strength is the quality assessment aspect of this review. Adapting a scoring system allowed for objective assessment of studies. It highlighted that most of the included studies were either satisfactory (n = 4) or excellent (n = 3), whilst making it clear that the greatest weaknesses were in the number of food items and the validity and adaptability of FFQs, which researchers should take into consideration when designing future studies. Another strength is that the review focused on large-scale, multi-regional studies, which are more representative of the respective GCC nations’ populations.

Conclusions

This is the first review to collect, quantify and critique the quality of data from dietary studies conducted in GCC countries by using an objective scoring system approach. Study quality varied, and major weaknesses of FFQ validity and adaptability have been highlighted.

Findings consistently showed that the majority of GCC populations are not meeting the recommended fruit and vegetable intake, and sugary-beverage consumption is on the rise, implying a poor diet. However, interpretations are made with caution due to the low study sample included (n = 7). In these GCC countries, where obesity levels are steadily rising, more dietary investigations are necessary. The use of validated FFQs in conjunction with other instruments like biomarkers, 24-hour recalls and/or food records is likely to provide more accurate dietary estimations.

In conclusion, it is essential that researchers develop well-designed, validated FFQs that are adapted for the GCC to standardise dietary assessments across studies.

Supplemental Information