A glycaemic index compendium of non-western foods

Current international tables published on the glycaemic index (GI) of foods represent valuable resources for researchers and clinicians. However, the vast majority of published GI values are of Western origin, notably European, Australian and North American. Since these tables focus on Western foods with minimal inclusion of other foods from non-Western countries, their application is of limited global use. The objective of this review is to provide the GI values for a variety of foods that are consumed in non-Western countries. Our review extends and expands on the current GI tables in an attempt to widen its application in many other regions of the world.


Introduction
In many non-Western countries, cereal-based carbohydrates provide~60% of total energy intake 1 compared with 42% for Caucasians 2 . The consumption of such highcarbohydrate diets yield high glucose and insulin response, thus contributing to insulin resistance. Nonetheless, the quality of carbohydrate consumed is as important as the quantity. Poor quality carbohydrates are quickly digested and absorbed, thereby giving rise to high blood glucose and insulin 'spikes'. Observational studies have shown that the consumption of low glycaemic index (GI) foods is associated with a lower risk of type 2 diabetes mellitus (T2DM) 3 , significantly less insulin resistance and a lower prevalence of the metabolic syndrome 4 . However, the vast majority of these published GI values are of European, Australian and North America origin.

Glycaemic index
The GI is defined as a numerical figure used to represent the ability of a carbohydrate food to raise blood glucose levels. It is expressed as a percentage of the incremental area under the glycaemic response curve (AUC) elicited by a portion of food containing 50 g available carbohydrate in comparison with the AUC elicited by a standard reference food of 50 g glucose or white bread in the same participant 5 . The principle is that the slower the rate of carbohydrate absorption into the bloodstream, the lower the rise of blood glucose level and the lower the GI value. A GI value of ≥70 is considered high, a GI value 56-69 inclusive is medium and a GI value ≤55 is low, where glucose = 100.
Following the approach of these authors, we are for the first time providing a compendium of GI values of non-Western foods 6 . Since many of these GI values were published in uncommon journals or located in various sources, it is not surprising that many previous authors may have found it a challenge to access and retrieve such information.
With a global pandemic of T2DM escalating, especially in emerging countries 7 , it is now recognised that the GI food-based intervention is an important tool in the management and prevention of T2DM 8 . Ironically, in regions of the world where there is a pandemic of T2DM, there is a shortage of a taxonomy of GI data of non-Western foods (e.g. Middle East, South Asia, Indian sub-continent) in contrast to the current international GI tables 6,9 . In 1997, the FAO/WHO Expert Consultation suggested that the concept of GI might provide a useful means of helping to select the most appropriate carbohydratecontaining foods for the maintenance of health and the treatment of several diseases 10 . A meta-analysis by Brand-Miller et al 11 . demonstrated that choosing low GI foods in place of conventional or high GI foods exhibited a small but clinically important effect on medium-term glycaemic control in patients with diabetes. Low GI foods have been shown to reduce insulin demand and lipid concentrations, improve blood glucose control and reduce body weight, thus preventing diabetes-related cardiovascular events [12][13][14][15] .
A typical non-Western diet, such as in South Asia, is high in carbohydrates with cereals such as polished rice, white flour, finger millet, semolina and wheat providing the bulk of the energy 16 . Furthermore, it has been shown that a unique metabolic feature of South Asians, for an identical carbohydrate load, elicits postprandial glucose peaks that are 2-3 times larger than Caucasians [17][18][19] . Hence, a compilation of the GI of non-Western foods is necessary for proper selection and modifications that may be of particular benefit to not only these groups of people but to a wider audience.
The objective of this review is, therefore, to consolidate the GI values for a variety of foods that are consumed in non-Western countries. This is in order to capture and encapsulate all the data available on GI that have not been reported in the general literature. Given that the largest preponderance of type 2 diabetes is in Asia, the Middle East, South America and parts of Africa, it is imperative that the database on GI is expanded in order for it to have global utility. With this in mind, papers were critically evaluated based on a strict criterion. The emphasis of this review has inevitably been to record and document the GI of various foods.

Research design and method
We conducted a comprehensive literature search for relevant, original articles published from January 2000 through May 2020. Since 2000 marked the exponential growth in GI testing in non-Western countries, we have decided to take this as the year of data analysis. Briefly, the following string of search terms was used in PubMed and Google Scholar, with no language or other restrictions: (glycaemic index) AND (foods) AND ('country'). The electronic search was supplemented by manual searches through the reference sections of selected publications, as well as with linked articles that were found to have cited these particular publications. Non-Western countries included in this search were as follows: Singapore, Malaysia, Indonesia, Brunei, Cambodia, Thailand, Japan, Korea, China, Taiwan, Hong Kong, Nepal, India, Myanmar, Vietnam, Sri Lanka, Philippines, United Arab Emirates, Yemen, Oman, Saudi, Qatar, Kuwait, Lebanon, Egypt, Pakistan and Bangladesh. The compiled GI studies in our article have utilised the recommended GI testing method 10,20 and fulfilled the minimum requirements for the following inclusion criteria for GI testing studies: minimum 10 participants (healthy/T2DM), instruments such as Yellow Spring Instruments (YSIs) and the use of handheld glucometers such as HemoCue® and other similar devices widely used in clinical studies for GI testing, amount of available carbohydrate and reference food (glucose/white bread/white rice). In the table, most of the foods are based on 50 g available carbohydrate. However, for foods with low to moderate carbohydrate density, it is justified by Brouns et al. 20 to use a lower amount of carbohydrate to prevent consumption of an excessively large amount of food. Finally, the GI of non-Western foods were grouped according to the countries. The food list was arranged according to the country of origin so as to allow individuals who are keener on knowing the GI variability of foods from their own country to retrieve this information conveniently. Table 1 lists 940 food items, citing 159 separate studies, representing reliable data derived from healthy subjects or individuals with type 2 diabetes. Figure 1 shows a flow diagram indicating a number of studies screened, excluded and included in this article. Non-Western countries included in this compilation were as follows: Singapore, Malaysia, Thailand, Indonesia, Philippines, Japan, Korea, China, Taiwan, Hong Kong, India, Sri Lanka, Emirates, Oman, Saudi and Lebanon. These countries were chosen based on published studies on GI from these locations, with validated methodologies used and the papers followed the inclusion criteria as described in our paper. The GI of non-Western foods was firstly arranged according to the country it was derived from. This was followed by the food item, the GI, serving size (if any), available carbohydrate portion, subject type and number, and lastly the reference food with time period of GI testing. An important feature of our paper is to encourage readers to interpret the data in a way that will enable them to select healthy foods from the GI range of foods available. Therefore, using the data generated from our GI tables, the illustrative example presented below are simple methods that may be adapted to reduce the GI values of carbohydrate-rich staples (Fig. 2).

Conclusion
We believe that the inclusion of the additional GI values of foods from non-Western countries will enhance the use                         and application of GI both in research and clinical practice. Many of the staples consumed in these regions are high in GI, notably rice, flatbread, noodles, buns, paus, pastries and so on. The use of these GI tables will also enable consumers to make informed choices on how best to select low GI foods. The GI data compiled in this article consists of both single and mixed meals. This is a major advance to many GI tables that have focused on single foods. Mixed meals in this region are complex in relation to ingredients used and taste. Given its complexity, our table that includes the GI of mixed meals is a major advantage. It is hoped that this compendium will further stimulate additional data collection and enhance the utility of GI tables for a worldwide audience.

WAYS TO REDUCE GI OF CARBOHYDRATE-RICH STAPLES
Use coarse grains instead of refined flour for food preparations Add legumes and high fibre vegetables to highly polished grains to lower GI Add fat such as olive oil, ground nut oil, butter to white bread or rice Choose parboiled rice or basmati rice instead of processed white rice Include protein sources such as meat, fish, eggs when consuming carbohydrates as proteins lowers GI and promotes insulin secretion Incorporate condiments such as cinnamon, vinegar, ginger into meals as they have shown to lower blood glucose levels Fig. 2 Recommendations of ways to reduce the GI of carbohydrate-rich staples. Figure shows how combination of food ingredients and foods may be used to reduce the glycaemic response of rice-based staples.