Adherence to Food-Based Dietary Guidelines: A Systemic Review of High-Income and Low- and Middle-Income Countries

Research comparing the adherence to food-based dietary guidelines (FBDGs) across countries with different socio-economic status is lacking, which may be a concern for developing nutrition policies. The aim was to report on the adherence to FBDGs in high-income (HIC) and low-and-middle-income countries (LMIC). A systematic review with searches in six databases was performed up to June 2020. English language articles were included if they investigated a population of healthy children and adults (7–65 years), using an observational or experimental design evaluating adherence to national FBDGs. Findings indicate that almost 40% of populations in both HIC and LMIC do not adhere to their national FBDGs. Fruit and vegetables (FV) were most adhered to and the prevalence of adhering FV guidelines was between 7% to 67.3%. HIC have higher consumption of discretionary foods, while results were mixed for LMIC. Grains and dairy were consumed below recommendations in both HIC and LMIC. Consumption of animal proteins (>30%), particularly red meat, exceeded the recommendations. Individuals from HIC and LMIC may be falling short of at least one dietary recommendation from their country’s guidelines. Future health policies, behavioral-change strategies, and dietary guidelines may consider these results in their development.


Introduction
The obesity epidemic is becoming the greatest public health concern worldwide. Globally, current data suggest that 1307 million adults are overweight and 671 million are obese, and the number of cases in low-and middle-income countries (LMIC) are rapidly reaching those observed in high-income countries (HIC) [1]. Morbidity and mortality related to obesity have been shown to follow a socio-economic gradient, with higher rates of chronic non-communicable diseases (NCD) among those from lower socio-economic positions [2]. Diet, along with other lifestyle behaviors, is an important risk factor for many NCD, and a large number of dietary components have been shown to be socio-economically patterned [3][4][5]. The sharing of cultural ideas is enhanced with the rapid development of the flow of goods, services, and capital, and the broadening of social networks via advanced communication technologies and enhanced transportation systems. Thus, globalization has accelerated diet and lifestyle changes-as seen with the westernization of diet quality [6].
Although first demonstrated in HIC, changes in diet quality have also been found in LMIC [7,8]. For instance, results from multi-center cross-sectional study, assessing the diet quality of individuals (n = 9218) for eight Latin American countries, showed that better scores for healthy eating were found in higher socio-economic populations, while scores

Eligibility Criteria
This review was limited to studies published in English. All studies were assessed according to the following inclusion and exclusion criteria summarized according to the PICO (Participants, Intervention, Comparison, and Outcome) framework: Participants: Studies were eligible if they included free-living children, adolescents, and adults until 60-65 years. The cut-off of 60 y was used in studies in LMIC and 65 y was used for studies in HIC, which was based on the countries' definition for older adults. National FBDGs generally focus on these populations [11] as individuals outside of this age range typically have special energy and nutrient needs [21,22] Studies that included participants with a pre-existing disease, an organic cause for obesity and other chronic NCD, or who were taking medication that could affect diet were excluded.
Intervention/Exposure: Studies were included if they used FBDGs to evaluate dietary intake in their own country. Guidelines developed by non-government institutions were excluded. Studies were included if they assess FBDGs through dietary assessment methods, such as food records, 24 h recalls (24hDR) and food frequency questionnaires (FFQ). Studies assessing diet quality and/or adherence to guideline using indexes (e.g., adherence to Dietary Guidelines for Americans using Healthy Eating Index, Alternate Healthy Eating Index, and Dietary Diversity Score) were excluded because they may have assessed additional items outside the FBDGs. Adherence to recommendations in national FBDGS was assessed based on individual level meeting or not meeting the national FBDGs food groups recommendations.
Comparison: Different study designs, i.e., cross-sectional, cohort, and interventions (randomized and non-randomized trials) were included in this review. If intervention design was used, no exclusion criteria were placed on duration, length of follow-up or date. Outcome: The key outcome of this review was to assess the adherence of participants' dietary intake to their respective national FBDGs. Studies were excluded if they focused on an outcome other than adherence to a national FBDG, i.e., obesity or other chronic NCD.
A secondary outcome of this review was to assess the difference in adherence to each national FDBG according to their socio-economic classification for each country. Countries were dichotomized into two types of economy: HIC and LMIC; upper-middle, lowermiddle, and low-income countries were identified as LMIC [23,24].

Data Extraction
Data were independently extracted from eligible studies by two reviewers (A.L. and S.H.) and cross-checked for accuracy by a third reviewer (J.H.). The extracted data included sample characteristics (age, sex, race/ethnicity, educational level), country, guideline used (name) and adherence to the guideline (reported as mean (± standard deviation/error, SD/SE) or frequency (%)).
Studies included in this review are summarized in Figure 1. A total of 12,557 eligible papers were identified: 2851 from CIANHL/EBSCO, 411 from Lilacs/SciElo, 1307 from ProQuest, 2413 from PubMed, 4508 from Scopus, and 1052 from Web of Science. After excluding duplicates and reading titles, 2802 studies were assessed for eligibility. Finally, 616 full-text articles met the inclusion criteria and 49 were considered for the qualitative synthesis.

Adherence to National FBDGs in HIC and LMIC by Food Groups
The percentages of meeting the guidelines in the HIC ranged from 14.0% [75] in Greenland to 43.0% [65] in the USA, and in the LMIC from 40.0% [56] with Malaysians and 45.0% [52] with Egyptians meeting at least one recommendation from their countryspecific FBDG. The food groups that were most frequently reported as having been met for the dietary guidelines were fruits and vegetables (n = 19, 38.8%) [27,30,35,37,38,40,41,43,44,48,52,53,56,64,70,72,74,75]. The adherence to fruit group recommendations in HIC varies from 14% in a population of school-age children from Greenland [75] to 67.3% in the overall Spanish population [30], and in LMIC from 13.4% in Malaysian children and adolescents [56] to 49.4% in the overall Brazilian population [53]. The adherence to vegetable group guidelines in HIC varies from 11.4% in the US population [59] to 43.7% in the Spanish population [30], and in LMIC varies from 9.5% in Malaysian adolescents [56] to 74.1% in the Brazilian population [53]. Some HIC and LMIC reported the combined fruit and vegetable guideline adherence: 18% in Swiss population [40], 7-16% in Mexican population [41] and 33.4% in Egyptian adolescents [52]. The adherence reported in HIC was less than 1.5 and 3 servings/day for fruit and vegetables (FV), respectively [48,70]. More specifically, a Canadian study with 33,850 individuals over 2 years reported the adherence to recommendations for dark green vegetables and orange fruits of 12% and 8%, respectively [44].
Other food groups from national FBDGs reported in the studies included: grains (whole vs. refined), proteins (eggs, meats, and fish), and dairy and alternatives. Reported adherence to guidelines for grains and cereals were 3.8% for women in the USA [64], 79.2% for Greenlandic adolescents consuming potatoes [75] and 50.5% of Germans consuming whole grains [74]. Schwartz and Vernarelli [35] found that individuals that used the MyPlate ® guideline to inform eating patterns had higher intakes of whole grains (1.1 to 0.8 servings/day) and refined grains (6.0 to 6.6 servings/day) as compared to those who did not use the guideline. In LMIC, less than 40% population of the met recommendations for adequate grain intake in their national FBDGs, [53,56,58], especially among younger participants. In HIC, adherence to milk and alternatives guidelines ranged from 8.4% in Switzerland [43] to 51.9% in US women [64]; and in LMIC from 5.5% with Malaysian youth [56] to 12.5% in an overall sample of Brazilian adolescents and adults [53]. One Australian [70] and one USA [48] study reported an average of 1.0 serving/day of dairy group. Finally, HIC studies that evaluated the adherence to guidelines for meat and alternatives showed an average intake 30.0% higher than the recommendations [30,37,43,64,74,75], with one study showing 95% of the population meeting the recommendations [74]. Alternatively, the adherence to guidelines for meat and alternatives in LMIC showed mixed results. In Brazil [53] and in China [58] adherence was greater than 65.0%, but in Egypt [52], Malaysia [56], and Mexico [41] it was less than 23.0%. Koo et al. [56] only reported the intake for fish and seafood, and the majority of the studies both from HIC and LMIC reported a high prevalence for (red) meat consumption.

Risk of Bias of the Included Studies
From all the included studies, selection bias (2.2 ± SD 0.5) was the most reported bias, while study design (0.1 ± SD 1.2) the less reported bias. Figure 2 shows the risk of bias of each component rating for the included studies.

Discussion
This review synthesized the evidence from observational and intervention studies reporting the adherence to national FBDGs in individuals from both HIC and LMIC. The 48 studies included in this review were conducted across 15 HIC and 4 LMIC, thus representing a broad perspective on this study objective. This review found that a large proportion of individuals in both HIC and LMIC are not meeting national dietary guidelines. Meat and alternatives, and discretionary foods were consumed above the recommended amounts, and vegetable intake was below the recommendations. A global review of 90 FBDGs found that most of the dietary guidelines demonstrated that the food groups that were most adhered to were the starchy staples (e.g., rice and potatoes) and the fruit and vegetables, while other groups were less adhered to across the countries [16].
Evidence from this review suggests that the global population may not be meeting the minimum dietary recommendations for FV and whole-grains, with more pronounced deficits in those from HIC. All national FBDGs have common themes that FV and whole grains should be incorporated in a healthy diet for the prevention of obesity, other chronic NCDs, and some nutrient deficiencies [11,12,78]. These food sources usually provide a low amount of fat, and are key sources of vitamins, minerals, and dietary fiber [77]. In some countries, particularly LMIC, where diets of nutritionally vulnerable groups (i.e., children, adolescent girls, women of child-bearing age, and older adults) continue to be inadequate, the co-occurrence of deficiencies from more than one micro-nutrient is common [78,79]. Thus, population-level interventions to improve dietary intake are urgently needed to reverse these global trends.
Findings from this review also demonstrated that discretionary foods and other high sugar and fat sources may exceed the recommendations in national FBDGs. However, it may be noted that most of the included studies assessing fat consumption only reported total fat intake or the combination of fats and oils, with the exception of one study that evaluated food groups and omega-3 fatty acid consumption after a nutrition education program [32]. Thus, most studies did not differentiate between the types of fat consumed   After participating in an educational program, participants did not improve their intake for the groups: grains, vegetables, and meat and alternatives. They improved their intake for fruit, dairy, and oil.  Less than 43% of the participants followed at least one recommendation from the guideline. Fruits, vegetables, and milk were consumed less than the recommendations. More than 60% of the participants exceeded the recommendations for fats, sugar and sodium. After intervention participants increased their intake for milk, vegetables, fruits, and grains, and decrease intake of meats.

Discussion
This review synthesized the evidence from observational and intervention studies reporting the adherence to national FBDGs in individuals from both HIC and LMIC. The 48 studies included in this review were conducted across 15 HIC and 4 LMIC, thus representing a broad perspective on this study objective. This review found that a large proportion of individuals in both HIC and LMIC are not meeting national dietary guidelines. Meat and alternatives, and discretionary foods were consumed above the recommended amounts, and vegetable intake was below the recommendations. A global review of 90 FB-DGs found that most of the dietary guidelines demonstrated that the food groups that were most adhered to were the starchy staples (e.g., rice and potatoes) and the fruit and vegetables, while other groups were less adhered to across the countries [16].
Evidence from this review suggests that the global population may not be meeting the minimum dietary recommendations for FV and whole-grains, with more pronounced deficits in those from HIC. All national FBDGs have common themes that FV and whole grains should be incorporated in a healthy diet for the prevention of obesity, other chronic NCDs, and some nutrient deficiencies [11,12,76]. These food sources usually provide a low amount of fat, and are key sources of vitamins, minerals, and dietary fiber [77]. In some countries, particularly LMIC, where diets of nutritionally vulnerable groups (i.e., children, adolescent girls, women of child-bearing age, and older adults) continue to be inadequate, the co-occurrence of deficiencies from more than one micro-nutrient is common [76,78]. Thus, population-level interventions to improve dietary intake are urgently needed to reverse these global trends.
Findings from this review also demonstrated that discretionary foods and other high sugar and fat sources may exceed the recommendations in national FBDGs. However, it may be noted that most of the included studies assessing fat consumption only reported total fat intake or the combination of fats and oils, with the exception of one study that evaluated food groups and omega-3 fatty acid consumption after a nutrition education program [32]. Thus, most studies did not differentiate between the types of fat consumed in the diet, such as saturated fats, trans fats, and unsaturated fats (e.g., omega 3/omega 6 fatty acids). Given the importance of the associations between different types of fatty acid consumption in different age groups and prevention of several negative health outcomes, such as chronic NCD in adults and older adults [79] and cognitive development in children [80], this information might be valuable. For example, it would be beneficial to distinguish between the negative impacts of saturated/trans fats and positive effects of unsaturated fatty acids in the diet. Furthermore, the finding that the consumption of discretionary foods, such as SSB and sweet snacks, exceed recommendations (i.e., 1 serving = 600kj (143kcal)) [81,82] is crucial for researchers and practitioners given that consumption of these foods may contribute to unhealthy weight gain [83,84]. Unhealthy weight gain is associated with several risk factors for poor health outcomes, public health policies, behavioral-change strategies, and periodical updates on the national FBDGs are needed to tackle this problem.
The studies included in this review suggest that dietary intake patterns differ across age groups. For instance, children and adolescents usually consume fewer FV and other fiber sources [85,86] than adults. Notably, evidence from systematic reviews [87,88] showed that children and adolescents had lower adherence to national FBDGs, and that families play an important role in influencing their eating behaviors [89,90]. Thus, family aspects of eating behavior might be included as key messages on public health initiatives and other resources for the population.
The strengths of this systematic review include the examination of a topic that filled a gap in the existing literature. This systematic review aimed to identify adherence to national FBDGs and verify possible differences between HIC and LMIC. The studies included in this review were conducted in five different continents (North America, South America, Asia, Europe, and Oceania), providing a global perspective on the topic.
This study was not without limitations. Only studies including children older than 2 years, adolescents, and adults were included, as nutrition recommendations for younger children and older adults may differ. Only studies that directly compared intake to their national FBDGs were included; no other measurement tools that were based on the recommendations (e.g., Healthy Eating Indexes) or direct comparisons to international energy and nutrient recommendations (e.g., Institute of Medicine and World Health Organization) were included. This may have caused confirmation bias, when interpreting the studies [91]. Some studies were not performed in representative samples of the correspondent population, compromising their representativeness. Also, the methodology has its own limitations, as, some of the articles included evaluated the dietary intake with one 24hDR with limited items, which is not representative of the habitual diet. Nevertheless, this method is accepted for studying the intake in a large sample of population and estimating the mean nutrients intake [92]. Reported dietary intake may provide biased results (under or over-reporting by participants) due to social desirability. Furthermore, only English publications were included. Additionally, a greater proportion of the studies that evaluate adherence to the FBDGs were from HIC, especially from the USA, limiting the generalizability of the results to other countries. Finally, the selected studies included a variety of population sub-groups (e.g., sex, age, and weight status) that made it difficult to make conclusions across studies.

Conclusions
National dietary guidelines can be a useful tool to promote a healthy diet for different age groups. A diet based on these guidelines should provide adequate energy and nutrient intake and support a healthy weight status and positive health outcomes. The findings from this review demonstrate that individuals in both HIC and LMIC may be falling short in at least one recommendation from national guidelines. Overall, these results suggest that a substantial proportion of the population are not consuming enough FV and whole grains. Excess intake of discretionary foods was also observed, especially among younger populations from both HIC and LMIC, and the overall population in HIC. Thus, socio-demographic factors (e.g., age, sex, and income) may influence adherence to the guidelines. These findings can help inform the development of future health policies, behavioral-change strategies, and food-based dietary guidelines.