Nutritional adequacy and dietary disparities in an adult Caribbean population of African descent with a high burden of diabetes and cardiovascular disease

Abstract The Caribbean island of Barbados has a high burden of diabetes and cardiovascular disease. Dietary habits were last described in 2005. A representative population‐based sample (n = 363, aged 25–64 years) provided two nonconsecutive 24‐hr dietary recalls in this cross‐sectional study. Mean daily nutrient intakes were compared with the Dietary Guidelines for Americans. Subgroup differences by age, sex, and educational level were examined using logistic regression. High sugar intakes exist for both sexes with 24% (95% CIs 18.9, 30.0) consuming less than the recommended <10% of energy from added sugars (men 22%; 15.0, 31.6; women 26%; 18.9, 33.7). Sugar‐sweetened beverages provide 43% (42.2%, 44.4%) of total sugar intake. Inadequate dietary fiber intakes (men 21 g, 18.2, 22.8; women 18 g, 16.7, 18.9) exist across all age groups. Inadequate micronutrient intake was found in women for calcium, folate, thiamine, zinc, and iron. Older persons (aged 45–64 years) were more likely to report adequacy of dietary fiber (OR = 2.7, 1.5, 4.8) and iron (OR = 3.0, 1.7, 5.3) than younger persons (aged 25–44). Older persons (aged 45–64 years) were less likely to have an adequate supply of riboflavin (OR = 0.4, 0.2, 0.6) than younger persons. Men were more likely to have adequate intakes of iron (OR = 13.0, 6.1, 28.2), folate (OR = 2.4, 1.3, 4.6), and thiamine (OR = 3.0, 1.5, 5.0) than women. Education was not associated with nutrient intake. The Barbadian diet is characterized by high sugar intakes and inadequate dietary fiber; a nutrient profile associated with an increased risk of obesity, type 2 diabetes, and related noncommunicable diseases.

attack and stroke, type 2 diabetes and cancers are the leading causes of premature death in Barbados (IHME, 2013).The island's annual mortality rate from CVD is estimated at 275.6 per 100,000 with the annual years of life lost being approximated at 4,759 per 100,000 ("Cardiovascular Diseases in Barbados," n.d.).The prevalence of type 2 diabetes for Barbados is estimated at 18.7% (95% CIs 16.2%; 21.4%) (Howitt et al., 2015), which is double the global figure. Barbados and the wider Caribbean have higher rates of premature NCD mortality than in other parts of the Americas (Ordunez, Prieto-Lara, Pinheiro Gawryszewski, Hennis, & Cooper, 2015).
A suboptimal diet is a known major contributor to the development of obesity, type 2 diabetes, and CVD (Ezzati, Pearson-Stuttard, Bennett, & Mathers, 2018) and also to prevalent micronutrient deficiencies. In Barbados, for example, it has been estimated that one-in-five women of reproductive age were anemic (2018Global Nutrition Report, 2018. Poor dietary habits are responsible for more deaths globally than any other risk factor (GBD 2017Diet Collaborators, 2019. Nutrition programmes are increasingly being integrated into public health policies with the focus now shifting away from the excessive intakes of unhealthy nutrients (Bruins, Van Dael, & Eggersdorfer, 2019) to the optimization of adequate nutrient-rich diets (WHO Global Action Plan for the Prevention andControl of NCDs 2013-2020, n.d.). The Global Burden of Disease (GBD) Study estimates that a suboptimal diet contributes approximately 15% of total disability-adjusted life years (DALYs) among adults worldwide (GBD 2017Diet Collaborators, 2019. In 2017, more than half of diet-related deaths and two-thirds of diet-related DALYs were attributable to high intakes of sodium, low intakes of whole grains, and low intakes of fruits. For the Caribbean, the GBD study estimated suboptimal intakes of fruits, vegetables, whole grains, nuts, and seeds, with inadequate levels of fiber and calcium (GBD 2017Diet Collaborators, 2019. The estimates from the GBD study are in fact based on highly limited data from the region and rely on modeling and extrapolation. Dietary habits for Barbadians were last described in 2005 where it was noted that Barbados like many other countries of the world was undergoing a "nutrition transition" (Sharma et al., 2008). However, the data from this small study (n = 49) were based on a cancer case-control study and were not designed to be representative of the general adult population.
In 2012-2013, we conducted the population-based Barbados Salt Study in adults aged 25-64 years, which included 24-hr dietary recalls and urine collection. We have previously published the data on urine sodium and potassium excretion from this study (Harris et al., 2018). Here, we describe data on nutritional adequacy from this population, the most detailed account to date and including for the first time for this population analyses by social group. We aimed, therefore, to assess and describe dietary adequacy for Barbadians and to investigate the influence of age, sex, and educational level on nutrient intake. We anticipate that this work will add to an area with the current paucity of dietary data for the Caribbean region. These findings will be of importance for the formulation of effective public health policy and culturally appropriate nutritional interventions, targeting those groups at greatest risk, for the Barbadian population.

| Study population
The study population was a subsample of the Barbados Health of the Nation (HotN)Study, a cross-sectional survey conducted in 2011-2013, which recruited a nationally representative sample of adults aged ≥25 years (n = 1,234). Details of the sampling, recruitment, and data collection methods have been published elsewhere (Harris et al., 2018;Howitt et al., 2015). A sample of 441 adults aged 25-64 years was selected, stratified by sex and age group (25-44 and 45-64 years), with the aim of recruiting at least 100 persons in each group. The methods used to collect biochemical (cholesterol and blood glucose), anthropometric, behavioral, and medical history data have been previously described (Howitt et al., 2015). Information on diet was collected during two face-toface interviews carried out in 2012-2013. Data on diabetes, hypertension, and cholesterol were collected in the main HotN study (Howitt et al., 2015). Diabetes was defined by self-reported doctor-diagnosed diabetes or fasting plasma glucose ≥7 mmol/L while hypertension was defined by self-reported current use of antihypertensive medication, or systolic blood pressure ≥140 mm Hg or a diastolic blood pressure ≥90 mm Hg. Pregnant and lactating women were excluded because of their unique nutritional requirements. All procedures followed were in accordance with the Helsinki Declaration of 1975 as revised in 1983, with ethical approval granted by the Institutional Review Board of the Ministry of Health, Barbados and the University of the West Indies. All participants provided written informed consent.

| Procedures
Two nonconsecutive, interviewer-administered, 24-hr dietary recalls were collected from each participant by trained data collectors, using the United States Department of Agriculture (USDA) multipass method. The 24-hr dietary recall is considered to be the gold standard method for assessing mean intake at the population level (GBD 2017Diet Collaborators, 2019. Portion size consumed in one sitting was captured using three-dimensional Nasco food models (Nasco Company), standard measuring cups, and household utensils. The timing, food source, frequency of consumption, cooking method, seasoning use, and recipes were documented.

| Dietary data analysis
Individual food items were coded, and recorded portions converted into grams by a Registered Dietitian (author RH). These data were double entered, into the nutrition software, Nutribase Pro (version 9, Cybersoft Inc.). The underlying databases for this program are the USDA and Canadian food composition tables. The Association of Official Agricultural Chemists (AOAC) method was used to estimate the fiber levels in foods included in these food composition tables (Nutribase 9 User's Manual, 2010). Standardized traditional Barbadian recipes (Sharma et al., 2007) were added to the underlying food composition database, making this software more culturally appropriate.

| Food grouping
Foods were placed into food categories guided by previous work done in Barbados (Sharma et al., 2008). All food categories were reviewed by a Registered Dietitian (author RH). Food groups were combined as follows: "vegetables" consisted of all fresh, frozen, and canned vegetables, but not including legumes and potatoes; "fish" comprised fatty and lean fish, crustaceans and molluscs; "red meat" included all red meat including processed canned products, organ, and minced meat; and "poultry" included all chicken and turkey meats and their products. "Fruit" consisted of whole fruit including fresh, frozen, and canned fruits. "Sugar-sweetened beverages (SSBs)" consisted of sweetened juices, homemade juices with added sugar, juice drinks, fruit flavored drinks made from sugar crystals and flavored milks, sodas or carbonated drinks, sports and energy drinks, and any sweetened hot drinks (e.g., hot chocolate and coffee). "Beverages with no added sugar" included 100% pure fruit juice, smoothies, diet sodas, and unsweetened tea and coffee. "Ground provisions" included sweet potato, yams, cassava, eddoes, green bananas, breadfruit, plantains, English potato, and French fries; "dairy" included milk, yogurt, cheese, ice-cream, butter, and cream. Cereals were split into two broad categories, "hot porridges" and "other cereals"; breads were also divided into two groups, "multigrain and bran" and "white"; "nuts and legumes" included all tree nuts, peanuts, nut butters, and legumes; "rice" and "pasta" included both refined and unrefined products as well as all composite dishes, such as macaroni pie and rice and peas.
"Cakes and sweetbreads" included all baked products and cookies; "chocolate drinks" included all chocolate flavored water-based beverages, "candy/sweets, chocolate" included all sweets, candy, mints, and chocolates; "sugar" included all types of sugar, honey, syrups, jams, and jellies.
The 2015-2020 Dietary Guidelines for Americans defines added sugars as "syrups and other caloric sweeteners used as a sweetener in other food products" (USDA, Dietary Guidelines for Americans, 2015-2020).This does not include naturally occurring free sugars such as in fruit (fructose) and milk (lactose). For this work reported total sugar (g/day) is comprised of, monosaccharides (glucose, fructose, and galactose) and disaccharides (sucrose, maltose, and lactose) originating from free sugars including added sugar sources.
The percentage contribution made by each food category either in terms of calories for energy or by weight in grams for fat, protein, total sugar, and fiber were calculated. These percentages were then ranked and the "top 10" contributing food categories for each macronutrient and dietary fiber determined.

| Socioeconomic indicators and age
Data on education were collected by self-report as the highest level of education completed. The data were available for almost all study participants and education was therefore chosen as the measure of socioeconomic status. Due to the small sample size (n = 364), education was divided into less than tertiary education and tertiary education. Tertiary education was defined as postsecondary education including college, vocational, and university training. Age was grouped into older (45-64 years) and younger (25-44 years) age groups.

| Statistical analyses
All analyses were weighted to account for the sampling design, nonresponse and to match the age-sex distribution of the Barbadian population according to the 2010 Barbados Population and Housing Census (Barbados Statistical Service, 2010). We excluded dietary recalls from all statistical analyses where the energy intake was extreme being >4,780 kcal/day for men and >4,302 kcal/day for women and <500 kcal/day (Bradbury, Tong, & Key, 2017). All continuous variables approximated a normal distribution therefore means or proportions (as appropriate) are presented, with 95% confidence intervals (CI) for both macronutrient and specific micronutrients. To obtain the mean individual daily intake for specific nutrients, the sum total for each nutrient recorded for the two recall days was divided by 2 (Sharma et al., 2008). We performed logistic regression to investigate the association by each of age, sex, and educational level on nutrient adequacy, while controlling for the other two as potential confounder. Statistical significance is reported as p < .05. We provide the odds ratios (ORs) and their associated 95% confidence intervals for dietary variables only for which statistical significance was found at p < .05 (i.e., where the ORs do not cross 1). Data analysis was performed using STATA v12 (StataCorp. LP, College Station, TX, USA). All estimates and analyses were adjusted using the "SVY" module of STATA v12 to account for the complex survey sampling design.

| RE SULTS
The study population demographic and health characteristics are shown in Table 1. The overall prevalence of obesity (BMI ≥ 30) was 36.0%, being almost twice as high in women (46.4%) than men (25.3%). The prevalence of hypertension (defined by a systolic blood pressure >140 mm Hg or a diastolic blood pressure ≥90 mm Hg) and diabetes was 34.1% and 13.1%, respectively, being similar for both sexes. The mean total-to-HDL cholesterol ratio was higher in men (6.5) than women (5.4) ( Table 1).
A total of 722 days of dietary data were analyzed. More weekdays (75%) than weekend days (25%) were captured, and approximately 15% of persons reported to be on a "special diet". The nutrient intakes by age and sex are shown in Table S1a. The estimated energy intakes were higher in men (2,333 kcal/day; 2,172.4, 2,494.0) than women (1,840 kcal/day; 1,734.5, 1,946.4). The mean percentage energy from carbohydrate (53%), protein (17%), and fat (28%-30%) were similar for both sexes and across all age groups. The mean intake of total sugars was 127 g/day (117.7, 136.3) for men and 99 g/day (90.9, 106.9) for women. The intake of the essential fatty acids, linoleic (men 4 g; 3.4,

| D ISCUSS I ON
The current study highlights several dietary factors with discordance between the recommendations and reported intakes in a representative national sample of adults in Barbados. According to this study, the nutrient profile for Barbadians is characterized by high sugar intakes, with half of total sugar intake coming from added sugars and half of added sugars coming from SSBs. Moreover, the intakes of dietary fiber were inadequate. For this population, age and sex appear to be stronger determinants of dietary intake than education, with women generally being at greater risk of dietary inadequacy than men. We focus our discussion around those results which we consider to be of public health importance.
The reduction of the current level of consumption for SSBs (10% of total estimated daily calories) in Barbados should be an urgent public health priority. SSBs are energy-dense beverages of poor nutritional value which have been associated with an increased risk of obesity and type 2 diabetes (Fagherazzi et al., 2013) and a greater prevalence of dyslipidaemia (Malik, Popkin, Bray, Després, & Hu, 2010). Each serving/d of habitual intake of SSBs has been associated with an 18% greater incidence of type 2 diabetes (Imamura et al., 2016)   that it will provide baseline metrics which can be used in the development of culturally appropriate public health messages and policy for Barbadians. The influence of cultural factors such as the personal beliefs of Barbadians around SSBs is, however, an area yet to be explored.
The US average intake for dietary fiber is estimated at 15.7-17.0 g/day and 13.1 g/day for non-Hispanic Blacks (Grooms, Ommerborn, Pham, Djoussé, & Clark, 2013), similar to estimated levels for Barbadians. An increase of dietary fiber in the diet is of paramount importance as it has been shown to be inversely associated with coronary heart disease (Wu et al., 2015) and stroke (Wang et al., 2014). Results from a recent meta-analysis suggest that individuals who consume higher amounts of dietary fiber, especially cereal fiber, may benefit from a reduction in the incidence of developing type 2 diabetes (InterAct Consortium, 2015). The total-to-HDL cholesterol ratio (Atherogenic or Castelli Index) is above the ideal ratio of 3.5 recommended by the American Heart Association for both sexes. The predictive value of this ratio is known to be greater than its isolated parameters (Millán et al., 2009). This ratio is a powerful predictor of ischemic heart disease risk and has also been associated with insulin resistance syndrome (Lemieux et al., 2001). Fat intakes despite being at an acceptable level for Barbadians were mainly from sources which are high in saturated (poultry, red meat, and dairy products) and trans-fat (cakes and sweetbreads). The promotion of healthier cardio-protective monounsaturated fats, omega-3, and omega-6 fatty acids from foods such as oily fish, plant-based oils (such as flaxseed, soybean, and canola oils) eggs and nuts should be encouraged, in an effort to reduce LDL-cholesterol while preserving HDL cholesterol (Mensink, Zock, Kester, & Katan, 2003). Processed meats make a notable contribution (6%) toward total fat intake in Barbados. Red and processed meat consumption has been associated with the development of type 2 diabetes (Pan et al., 2013). Several additives found in processed meats such as nitrites and their by-products (e.g., peroxynitrite) experimentally promote atherosclerosis and vascular dysfunction (Förstermann, 2008), reduce insulin secretion (Förstermann, 2008), and impair glucose tolerance (Portha, Giroix, Cros, & Picon, 1980). Alternative protein sources such as poultry, fish, and legumes should be encouraged.
Folate, vitamin B6, and vitamin B12 are involved in homocysteine metabolism, which are thought to reduce CVD risk by lowering homocysteine levels (Clarke et al., 2010). Clouden for assistance with data collection.

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflicts of interest.

E TH I C A L A PPROVA L
Ethical Review: All procedures followed were in accordance with the Helsinki Declaration of 1975 as revised in 1983, with ethical approval granted by the Institutional Review Board of the Ministry of Health, Barbados and the University of the West Indies.

I N FO R M E D CO N S E NT
Written informed consent was obtained from all study participants.