1. Introduction
The incidence rate and prevalence of type 2 diabetes have steadily increased in Asian populations. The International Diabetes Federation (IDF) estimated that 163 million people (35.2% of global diabetic population) in the Western Pacific region had prevalent type 2 diabetes in 2019 [
1], contributing the most to type 2 diabetes in the world. In Korea, the prevalence of type 2 diabetes has increased from 6.9% in 1998 to 10.8% in 2017 [
2,
3]. Additionally, type 2 diabetes contributed to 17.1% of the total deaths in Korea in 2018 [
4].
Coffee consumption has been suggested to lower several chronic diseases, including type 2 diabetes [
5], metabolic syndrome [
6], coronary heart disease [
7], liver disorders [
8], and several types of cancers [
9]. The bioactive compounds in coffee, such as caffeine and chlorogenic acids, have been investigated as potential compounds that lower the risk of type 2 diabetes. Caffeine has been shown to stimulate the metabolic rate [
10,
11], and its thermogenic effect has been hypothesized to decrease the risk of metabolic disease development. Antioxidants, including chlorogenic acids, commonly found in coffee, have also been highlighted as a preventing factor for type 2 diabetes by inhibiting the generation of free radicals and removing hyperglycemia-induced oxidative stress [
12,
13,
14].
A heritability study on caffeine [
15] and genome-wide association studies (GWASs) have suggested that coffee consumption behavior may be linked to genetic polymorphisms. The first genome-wide meta-analysis of coffee consumption was conducted in a European population and identified two independent loci, rs4410790 nears
AHR and rs2470893 between
CYP1A1 and
CYP1A2, which are caffeine metabolism-related genes [
16]. Additional European/Caucasian GWASs also discovered single nucleotide polymorphisms (SNPs) located in
AHR,
CYP1A1 and
CYP1A2 as well as
ABCG2,
POR,
BDNF,
SLC6A4,
GCKR and
MLXIPL [
17], near
NRCAM or
ULK3 [
18]. A Japanese GWAS, the first GWAS of coffee consumption in Asia, identified 24 SNPs on chromosome 12, showing rs2074356 in
HECTD4 as the strongest significant SNP [
19]. Another Japanese GWAS found two loci located in
CUX2 (rs7910258) and
AHR (rs10251701) [
20]. The few GWASs in Asia warrant further investigation of the coffee-related genetic polymorphisms in Asian populations because there has been an increase in coffee consumption and type 2 diabetes in Asia. Because coffee consumption has the potential to prevent type 2 diabetes [
21], it is important to investigate whether coffee consumption is linked to a lower risk of type 2 diabetes and whether this association is modified by genetic variations common in Asian populations.
The objective of this study was to identify genetic polymorphisms associated with habitual coffee consumption and examine whether the association between coffee consumption and the risk of prediabetes and type 2 diabetes combined was modified by these genetic variants in Korean population.
4. Discussion
Our first GWAS of coffee consumption in the Korean population identified five SNPs (rs2074356 in HECTD4, rs11066015 in ACAD10, rs12229654 in MYL2, rs11065828 and rs79105258 in CUX2) related to habitual coffee consumption. Compared with non-coffee consumers, the risk of prediabetes and type 2 diabetes being combined was inversely associated with habitual coffee consumption, either black coffee or sugared coffee. Individuals with black coffee consumption had a lower risk of prediabetes and type 2 diabetes combined compared with non-black-coffee consumers among those with multiple minor alleles for these five SNPs.
The significant SNPs discovered to be related to habitual coffee consumption in our GWAS were all introns. Although the introns were noncoding regions of genes, they may affect the transcription rate and translation efficiency, further regulating gene expression [
33]. Recent GWASs have identified several loci on the
AHR,
CYP1A1, and
CYP1A2 genes associated with coffee consumption [
16,
17]. However, most of the GWASs on coffee consumption were conducted in European populations, only two GWASs, to our knowledge, reported SNPs related to coffee consumption in Asian populations. A GWAS in the Japan Multi-Institutional Collaborative Cohort (J-MICC) study, the first GWAS on coffee consumption in Asia, found that rs2074356 located in 12q24 was most strongly associated with habitual coffee consumption (
p = 2.2 × 10
−6) [
19]. Similarly, in our GWAS, a total of 18 SNPs were associated with coffee consumption at
p < 1 × 10
−5, all of which were found to be related to habitual coffee consumption in the J-MICC study, and the strongest significant variant was rs2074356 as well. Another Japanese coffee GWAS identified two independent loci (rs79105258 in 12q24 and rs10252701 in 7p21) that were associated with coffee consumption [
20]. rs79105258 was also selected as a coffee-related variant in our study. In addition to the association with habitual coffee consumption, these SNPs were associated with type 2 diabetes [
34], blood glucose levels [
35], blood pressure levels [
36], and obesity [
37]. rs2074356 in
HECTD4 was associated with prevalent type 2 diabetes and blood glucose level in a Korean population [
34,
35]. Three SNPs (rs12229654, rs11066015 and rs2074356) were also identified to be linked with both systolic and diastolic blood pressure in a Japanese GWAS [
36].
Previous epidemiologic studies have shown that coffee consumption was inversely associated with the risk of type 2 diabetes. A meta-analysis of 28 cohort and nested case-control studies reported that participants who consumed 5 cups/day of coffee had a 30% lower risk of type 2 diabetes compared with almost non-consumers, and the associations were similar between men and women [
5]. Although we found a stronger inverse association among men than among women, further investigation is needed to explore a larger amount of coffee consumption, e.g., 3 or more cups/day, and whether the inverse association holds for women.
The lower risk of type 2 diabetes linked to coffee consumption could be linked to several biological mechanisms. As a main polyphenolic compound in coffee, chlorogenic acids have been shown as inhibitors of hepatic glucose-6-phosphatase, the rate-limiting enzyme of glucose hydrolysis [
38]. Reduced hepatic glucose-6-phosphatase may affect the glucose output and thus decrease the blood glucose concentration. In addition, chlorogenic acids act as antioxidants to lower oxidative stress shown in both in vitro and in vivo studies [
39]. Additionally, caffeine and magnesium, both of which are commonly found in coffee, have been suggested to have roles in type 2 diabetes prevention by improving insulin resistance. Previous studies suggested that caffeine could improve insulin resistance by stimulating insulin secretion from pancreatic β cells [
40]. In addition to insulin secretion, caffeine increases thermogenesis, lipolysis, and β-oxidation [
41]. Magnesium supplementation has reduced the development of type 2 diabetes and improved glucose disposal in experimental studies [
42,
43], and cohort studies have reported a significant inverse association between magnesium intake and type 2 diabetes risk [
44].
In this study, we observed that both black coffee and sugared coffee decreased the risk of prediabetes and type 2 diabetes combined for individuals who consumed more than 2 cups of coffee per day. But participants who consumed 1 to <2 cups/day of sugared coffee had a 45% higher risk of prediabetes and type 2 diabetes combined among men alone, but not among women. Although sugar-sweetened beverage intake including carbonated beverages has been positively associated with the risk of type 2 diabetes [
45,
46], the results of coffee with sugar remained equivocal. In a French cohort study, compared to non-coffee consumers, the incident type 2 diabetes decreased by 40% and 31% among participants consuming more than 1.1 cups of coffee per lunch with or without sugar, respectively [
47]. In a small clinical trial, where eight lean, young and healthy adults drank six types of beverages or water 1 h before a potato-based meal, postprandial hyperglycemia, an early abnormality of type 2 diabetes [
48], was significantly reduced when they drank sweetened coffee before their meals [
49]. Further studies are needed to examine whether the benefit of coffee remains even after adding a small amount of sugar.
To our knowledge, this study is the first GWAS of habitual coffee consumption in a Korean population. The strengths of this study include good ascertainment of prediabetes and type 2 diabetes, adjustment for potential confounding factors, and a 12-year follow-up. The incidence of prediabetes and type 2 diabetes were identified based on the circulating levels of FPG, 2-h plasma glucose, and HbA1c, which could minimize the misclassification. Adjustment for potential confounding factors, including smoking status and alcohol consumption, may enable us to remove the effect of the confounding factors. However, we cannot rule out the possibility that residual confounding factors remained. There are several limitations to our study. First, the rate of revisits to the clinic for the blood draw decreased to 60% in the last sixth follow-up. Therefore, our study may not be representative of the full cohort of KARE study. However, the internal validity may not be impaired as we obtained relatively accurate information on the incidence of prediabetes and type 2 diabetes during the 12-year follow-up. Second, we could not distinguish caffeinated and decaffeinated coffee or boiled and filtered coffee. However, previous studies have shown that the associations between coffee and type 2 diabetes were similar by the amount of caffeine [
5] or preparation methods [
50]. Third, we were not able to examine high coffee consumption (e.g., 3 or more cups/day) because only a few participants consumed more than 3 cups/day. Fourth, we did not consider medicinal caffeine intake when we performed a GWAS. However, beverages may mainly contribute to daily caffeine intake in Korea.