Abstract
OBJECTIVE
To examine the association between dietary intake of choline and betaine and the risk of type 2 diabetes.
RESEARCH DESIGN AND METHODS
Among 13,440 Atherosclerosis Risk in Communities (ARIC) study participants, the prospective longitudinal association between dietary choline and betaine intake and the risk of type 2 diabetes was assessed using interval-censored Cox proportional hazards and logistic regression models adjusted for baseline potential confounding variables.
RESULTS
Among 13,440 participants (55% women, mean age 54 [SD 7.4] years), 1,396 developed incident type 2 diabetes during median follow-up of 9 years from 1987 to 1998. There was no statistically significant association between every 1-SD increase in dietary choline and risk of type 2 diabetes (hazard ratio [HR] 1.01 [95% CI 0.87, 1.16]) nor between dietary betaine intake and the risk of type 2 diabetes (HR 1.01 [0.94, 1.10]). Those in the highest quartile of dietary choline intake did not have a statistically significant higher risk of type 2 diabetes than those in the lowest choline quartile (HR 1.09 [0.84, 1.42]); similarly, dietary betaine intake was not associated with the risk of type 2 diabetes comparing the highest quartile to the lowest (HR 1.06 [0.87, 1.29]). Among women, there was a higher risk of type 2 diabetes, comparing the highest to lowest dietary choline quartile (HR 1.54 [1.06, 2.25]), while in men, the association was null (HR 0.82 [0.57, 1.17]). Nevertheless, there was a nonsignificant interaction between high choline intake and sex on the risk of type 2 diabetes (P = 0.07). The results from logistic regression were similar.
CONCLUSIONS
Overall and among male participants, dietary choline or betaine intakes were not associated with the risk of type 2 diabetes. Among female participants, there was a trend for a modestly higher risk of type 2 diabetes among those with the highest as compared with the lowest quartile of dietary choline intake. Our study should inform clinical trials on dietary choline and betaine supplementation in relationship with the risk of type 2 diabetes.
Introduction
Type 2 diabetes is a major health problem globally and in the U.S. (1,2). Dietary habit is a leading contributor to morbidity and mortality globally (3) and is a most prominent risk factor for type 2 diabetes in addition to obesity, fat distribution, physical inactivity, age, and family history of diabetes (4,5). There is strong evidence for the association between food groups and the risk of type 2 diabetes, but there remains a huge gap in evidence on the potential association between specific dietary nutrients such as choline and betaine and the risk of type 2 diabetes.
Given the involvement of choline (an essential nutrient) and betaine (main methyl donor) in the one-carbon cycle with possible epigenetic effects, there is a growing interest in the association between dietary choline and betaine intake, both found in a variety of foods, and the risk of diseases in humans. However, there are only a few previous human studies on dietary choline and betaine intake and the risk of type 2 diabetes with inconsistent results. A recent study among men found a lower risk of type 2 diabetes in those with the highest choline intake as compared with those with the lowest choline intake (6). Extant literature points to an association between dietary consumption of phosphatidylcholine (7), as well as plasma, serum, or urine metabolites of choline, and risk of type 2 diabetes (8). Trimethylamine-N-oxide (TMAO), a microbiota-dependent metabolite of dietary choline, is found to be directly associated with the risk of type 2 diabetes (9). At the same time, some studies suggest that dietary choline and betaine intakes are associated with better insulin sensitivity in humans (10). The enzymes in the one-carbon cycle that involves choline and betaine are differently expressed in males and females and result in higher choline and betaine metabolites in females than males (11). There are also differences in choline requirements (12).
Given that dietary habits are modifiable, examination of the association of dietary choline intake with the risk of type 2 diabetes may inform dietary modifications, which may help reduce the risk attributable to choline. In this study, we aimed to assess the prospective longitudinal association between dietary choline and betaine intake and the risk of type 2 diabetes among participants of the Atherosclerosis Risk in Communities (ARIC) cohort.
Research Design and Methods
Study Population
Data for the current study were obtained from the ARIC study cohort (13). The baseline ARIC cohort examination was conducted from 1987 to 1989 and included 15,792 men and women from four U.S. communities including Forsyth County, NC; Jackson, MS; suburbs of Minneapolis, MN; and Washington County, MD. Participants were followed up through annual telephone interviews and review of hospitalizations. Extensive examinations were conducted initially every 3 years, with the latest examination currently underway. Participants were aged 45–64 years at baseline. Participants were excluded from this analysis if they had baseline type 2 diabetes (n = 2,014); if they had missing values for dietary choline, dietary betaine, type 2 diabetes, or other relevant covariates (n = 127); or if they were of a race other than Black or White (n = 48). Participants (n = 163) with improbable total calorie intake (≤500 kcal/day or >5,000 kcal/day) were also excluded from the analysis (14). All participants (n = 13,440) with complete data on exposure variables, covariates, and outcome variables were included in the final analysis. Insulin resistance, an impaired response of peripheral tissues to insulin action, was defined based on the HOMA of insulin resistance, which was calculated using: (FPI × FPG)/22.5, in which FPI is fasting plasma insulin concentration (milliunits per liter) and FPG is fasting plasma glucose (millimoles per liter) (15). An HOMA of insulin resistance cutoff value of ≥4.70 (80th percentile) was used to define insulin resistance (16). Impaired glucose tolerance was defined based on fasting blood glucose. Participants with fasting blood glucose of 100–125 mg/dL were classified as meeting the definition of impaired fasting glucose (17). In a sensitivity analysis, we excluded participants with impaired glucose tolerance and insulin resistance conditions with an underlying abnormality of glucose metabolism and considered precursors of diabetes.
Primary Exposures of Interest
Baseline dietary choline and dietary betaine intake were measured using a 66-item version of the Willett semiquantitative food frequency questionnaire (FFQ) (18). The dietary choline and betaine intakes were estimated based on the daily serving sizes of the food items known to have choline and betaine by summing the amount of choline and betaine in each food item per serving size based on the U.S. Department of Agriculture food content database (19,20).
Primary Outcome
Incident type 2 diabetes was the primary outcome of interest in the study and was ascertained at visit 2 (1990–1992), visit 3 (1993–1995), and visit 4 (1996–1998) among those free of type 2 diabetes at baseline. Participants with fasting blood glucose level ≥126 mg/dL, taking diabetes medication, or reporting a physician diagnosis of diabetes were considered to have type 2 diabetes, while those with fasting blood glucose level <126 mg/dL who were not taking diabetes medications were considered free of type 2 diabetes.
Covariates
The covariates in the study collected at study baseline included sociodemographic variables (age, sex, race, and education); BMI; health behaviors (smoking and alcohol use); total cholesterol; and intake of total calories, folate, magnesium, vitamin B6, vitamin B12, methionine, and animal fat.
Statistical Analysis
Baseline participant characteristics across the choline quartiles were compared using ANOVA for normally distributed and Kruskal-Wallis tests for nonnormally distributed continuous variables and χ2 test for categorical variables. Interval-censored Cox proportional hazards regression models sequentially adjusted for baseline covariates were fit to determine the longitudinal association between dietary choline and betaine intake and the risk of type 2 diabetes considering cubic splines of baseline hazard. Covariates adjusted for in model 1 were age, race, sex, BMI, alcohol intake, and smoking, and the exposure variables (choline and betaine) were mutually adjusted. Model 2 further adjusted for baseline fasting blood glucose and total calorie intake. Model 3 further adjusted for intake of dietary vitamin B6, vitamin B12, folate, methionine, animal fat, and dietary magnesium. Nonnormally distributed variables were log-transformed. We also conducted logistic regression analysis adjusting for covariates listed from model 1 to model 3. The interactions between the exposure variables and race or sex were tested using the likelihood ratio test. P values ≤0.05 were considered statistically significant. P-trend values were computed using the median of choline and betaine in each quartile as continuous (21,22). A sensitivity analysis excluding participants with insulin resistance at baseline was also conducted. Multicollinearity was examined using variance inflation factor. A variance inflation factor >10 was considered high. Animal protein intake was removed from the models, as it was collinear with methionine intake. All statistical analyses were conducted using R Statistical Software (version 3.5.1) and SAS software, version 9.4 (SAS Institute Inc., Cary, NC).
Results
The mean baseline age of the participants was 54 (SD 7.4) years, and 55% were female. Compared with participants with the lowest quartile of choline intake, those with the highest dietary choline intake had a higher BMI and higher intakes of alcohol, total calories, folate, vitamin B6, vitamin B12, animal fat, animal protein, and dietary magnesium. The proportion of male participants and participants with less than a high school education was greater among those with high as compared with low dietary choline intake (Table 1). We observed that betaine intake was positively correlated with choline intake (ρ = 0.62).
Table 1.
Baseline ARIC cohort participants’ characteristics
| Choline quartiles | P value | ||||
|---|---|---|---|---|---|
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | ||
| N | 3,392 | 3,392 | 3,392 | 3,391 | |
| Choline, mg/day (median [IQR]) | 175.04 (35.00) | 258.71 (20.18) | 333.88 (24.37) | 487.04 (114.62) | <0.001 |
| Betaine, mg/day (median [IQR]) | 56.79 (21.62) | 74.51 (25.86) | 87.65 (27.84) | 114.84 (43.06) | <0.001 |
| Age (mean [SD]) | 53.79 (5.75) | 54.07 (5.72) | 54.02 (5.74) | 54.00 (5.75) | 0.175 |
| BMI, kg/m2 (median [IQR]) | 26.73 (4.94) | 27.17 (5.17) | 27.44 (5.14) | 27.63 (5.11) | <0.001 |
| FBG, mg/dL (median [IQR]) | 98.15 (9.33) | 98.68 (9.14) | 98.79 (9.44) | 98.97 (9.32) | 0.002 |
| Race | 0.082 | ||||
| Black | 814 (24.0) | 782 (23.1) | 803 (23.7) | 869 (25.6) | |
| White | 2,578 (76.0) | 2,610 (76.9) | 2,589 (76.3) | 2,522 (74.4) | |
| Sex | <0.001 | ||||
| Female | 2,138 (63.0) | 1,955 (57.6) | 1,836 (54.1) | 1,547 (45.6) | |
| Male | 1,254 (37.0) | 1,437 (42.4) | 1,556 (45.9) | 1,844 (54.4) | |
| Education, n (%) | <0.001 | ||||
| Grade school | 305 (9.0) | 266 (7.9) | 271 (8.0) | 299 (8.8) | |
| Less than high school | 471 (13.9) | 432 (12.8) | 404 (11.9) | 495 (14.6) | |
| High school graduate | 1,223 (36.1) | 1,107 (32.7) | 1,064 (31.4) | 1,047 (30.9) | |
| Vocational school | 271 (8.0) | 291 (8.6) | 311 (9.2) | 299 (8.8) | |
| College | 819 (24.2) | 921 (27.2) | 968 (28.6) | 886 (26.2) | |
| Graduate/professional school | 298 (8.8) | 370 (10.9) | 372 (11.0) | 360 (10.6) | |
| Alcohol, g/week (mean [SD]) | 39.88 (90.58) | 43.30 (93.93) | 45.14 (96.68) | 50.12 (111.36) | <0.001 |
| Magnesium, mg/day (median [IQR]) | 176.61 (56.96) | 227.98 (64.84) | 267.66 (69.79) | 343.34 (94.39) | <0.001 |
| Total calorie intake, kcal/day (mean [SD]) | 1,082.90 (323.77) | 1,411.50 (342.88) | 1,714.48 (394.38) | 2,335.27 (639.38) | <0.001 |
| Cholesterol, mg/dL (median [IQR]) | 215.78 (42.54) | 214.62 (40.99) | 213.46 (40.99) | 212.69 (40.22) | 0.011 |
| Folate intake, μg DFE/day (mean [SD]) | 157.17 (68.35) | 205.20 (77.42) | 240.71 (84.35) | 306.43 (111.41) | <0.001 |
| Vitamin B6, mg/day (mean [SD]) | 1.14 (0.41) | 1.52 (0.44) | 1.82 (0.49) | 2.35 (0.67) | <0.001 |
| Vitamin B12, μg/day (mean [SD]) | 4.70 (2.69) | 6.44 (3.16) | 8.11 (3.64) | 11.10 (4.81) | <0.001 |
| Methionine, g/day (mean [SD]) | 0.98 (0.26) | 1.43 (0.25) | 1.81 (0.29) | 2.53 (0.62) | <0.001 |
| Animal fat, g/day (mean [SD]) | 20.52 (8.36) | 29.44 (9.60) | 38.45 (10.82) | 55.82 (18.89) | <0.001 |
| Animal protein, g/day (mean [SD]) | 29.79 (9.24) | 44.58 (9.61) | 57.49 (10.97) | 81.56 (22.57) | <0.001 |
DFE, dietary folate equivalent; FBG, fasting blood glucose; IQR, interquartile range.
Among our 13,440 ARIC participants free of type 2 diabetes at baseline, during a median of 9 years (interquartile range 3 years), 1,396 developed incident type 2 diabetes during the follow-up. In the fully adjusted multivariable model (model 3), we observed no association of dietary choline with the risk of type 2 diabetes (hazard ratio [HR] 1.01 [95% CI 0.87, 1.16] per 1 SD of choline intake). Likewise, betaine intake was not associated with the risk of type 2 diabetes (HR 1.01 [95% CI 0.94, 1.10] per 1 SD of betaine intake) (Table 2). In a sensitivity analysis only among those without insulin resistance at baseline, there was no association between 1-SD increase in dietary choline or betaine intake and the risk of type 2 diabetes (data not shown). The logistic regression analysis result per 1 SD of choline and betaine intake was similar to that of the survival analysis (Supplementary Table 1).
Table 2.
HRs and 95% CIs for the association of dietary choline and betaine intake per 1-SD increase as a continuous variable with the risk of type 2 diabetes
| Overall | Female | Male | |
|---|---|---|---|
| Choline | |||
| Model 1 | 1.06 (0.99, 1.12) | 1.05 (0.96, 1.15) | 1.06 (0.97, 1.15) |
| Model 2 | 0.99 (0.91, 1.07) | 1.06 (0.97, 1.15) | 0.99 (0.91, 1.07) |
| Model 3 | 1.01 (0.87, 1.16) | 1.16 (0.93, 1.44) | 0.93 (0.76, 1.13) |
| Betaine | |||
| Model 1 | 0.97 (0.91, 1.04) | 1.03 (0.94, 1.14) | 0.92 (0.84, 1.00) |
| Model 2 | 1.02 (0.95, 1.10) | 1.09 (0.98, 1.22) | 0.94 (0.85, 1.05) |
| Model 3 | 1.01 (0.94, 1.10) | 1.05 (0.94, 1.18) | 0.98 (0.87, 1.09) |
Model 1 mutually adjusted for choline and betaine and further adjusted for age, sex (in overall models), race, log-BMI, education, smoking, and alcohol. Model 2 further adjusted for total calorie intake and fasting blood glucose. Model 3 further adjusted for magnesium, vitamin B6, vitamin B12, methionine, animal fat, animal protein, and folate intake.
Similarly, comparing those in the highest quartile of dietary choline intake with those in the lowest quartile, there was no significant association between dietary choline intake and the risk type 2 diabetes (HR 1.09 [95% CI 0.84, 1.42]) nor was dietary betaine intake significantly associated with the risk of type 2 diabetes, comparing the highest to the lowest quartile (HR 1.06 [95% CI 0.87, 1.29]) (Table 3). In a sensitivity analysis among study participants without insulin resistance at baseline, effect estimates were similar to those observed in the main analyses (data not shown).
Table 3.
HRs and 95% CIs for the association of dietary choline and betaine intake with the risk of type 2 diabetes
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | P value | |
|---|---|---|---|---|---|
| Choline quartiles | |||||
| Model 1 | 1.00 (Reference) | 1.03 (0.88, 1.20) | 1.14 (0.98, 1.33) | 1.20 (1.01, 1.43) | 0.022 |
| Model 2 | 1.00 (Reference) | 0.98 (0.84, 1.16) | 1.02 (0.85 1.23) | 1.09 (0.84, 1.42) | 0.685 |
| Model 3 | 1.00 (Reference) | 0.98 (0.84, 1.16) | 1.02 (0.85, 1.23) | 1.09 (0.84, 1.42) | 0.440 |
| Sensitivity analysis | |||||
| Model 3 | 1.00 (Reference) | 1.05 (0.83, 1.32) | 1.01 (0.77, 1.31) | 0.97 (0.67, 1.41) | 0.260 |
| Betaine quartiles | |||||
| Model 1 | 1.00 (Reference) | 1.03 (0.88, 1.19) | 1.09 (0.94, 1.28) | 0.99 (0.83, 1.18) | 0.923 |
| Model 2 | 1.00 (Reference) | 1.02 (0.87, 1.18) | 1.08 (0.93, 1.27) | 1.08 (0.89, 1.30) | 0.362 |
| Model 3 | 1.00 (Reference) | 1.00 (0.86, 1.17) | 1.06 (0.90, 1.25) | 1.06 (0.87, 1.29) | 0.510 |
| Sensitivity analysis* | |||||
| Model 3 | 1.00 (Reference) | 0.92 (0.74, 1.15) | 1.11 (0.88, 1.40) | 1.15 (0.87, 1.52) | 0.176 |
P value is P for trend computed by modeling the quartile median of each nutrient as the predictor variable. Model 1 mutually adjusted for choline and betaine and further adjusted for age, sex, race, log-BMI, education, smoking, and alcohol. Model 2 further adjusted for total calorie intake and baseline fasting blood glucose. Model 3 further adjusted for magnesium, vitamin B6, vitamin B12, methionine, animal fat, and folate intake.
In sensitivity analyses, we limited estimates to participants without insulin resistance at baseline.
The result from logistic regression was similar to the interval-censored Cox proportional hazards regression analysis for the highest choline quartile compared with the lowest (odds ratio [OR] 1.22 [0.89, 1.66]) and for betaine (OR 0.99 [0.77, 1.26]) (Supplementary Table 2).
In crude models, there were significant effect modifications by sex and race on the association between choline intake and the risk of type 2 diabetes. In fully adjusted models, there was no effect modification by race but there was effect modification by sex. In a sex-stratified fully adjusted analysis, while there was no significant association between 1-SD increase in dietary choline or betaine intake and the risk of type 2 diabetes (Table 2), in analysis comparing the highest quartile to the lowest, dietary choline intake was associated with a higher risk of type 2 diabetes among women (HR 1.54 [1.06, 2.25]; P-trend = 0.029) but not among men (HR 0.82 [0.57, 1.17]; P-trend = 0.360) (Table 4). The result of the logistic regression was also similar among female participants for the highest choline quartile versus the lowest (OR 1.87 [95% CI 1.17, 3.00]), while among male participants, there was no significant association (OR 0.89 [95% CI 0.58, 1.37]) (Supplementary Table 3). The P for interaction between choline and sex was 0.070. In interval-censored Cox proportional regression analysis, there was a direct association between betaine and risk of type 2 diabetes among females (HR 1.31 [0.99, 1.74]; P-trend = 0.039) but not among males (HR 0.82 [0.62, 1.08]; P-trend = 0.188) (Table 4). In logistic regression analysis, there was no statistically significant association between dietary betaine intake and the risk of type 2 diabetes in stratified analysis by sex (Supplementary Table 3). The interaction of choline and betaine with race was not statistically significant (P > 0.05). In both the overall and sex-stratified multivariate-adjusted models, the association between each nutrient without mutually adjusting for betaine or choline, respectively, and the risk of type 2 diabetes did not materially change. The HR for the highest choline quartile versus the lowest was 1.10 (0.85, 1.42) and for betaine, 1.06 (0.87, 1.29). In sex-stratified analysis, among females, the HR for the highest quartile was 1.57 (1.08, 2.29) for choline and 1.35 (1.02, 1.78) for betaine. Among males, the HR for choline was 0.81 (0.57, 1.16) and for betaine, 0.81 (0.62, 1.07) comparing the highest quartile to the lowest.
Table 4.
HRs and 95% CIs for the association of dietary choline and betaine intake with the risk of type 2 diabetes stratified by sex
| Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | P value | |
|---|---|---|---|---|---|
| Female | |||||
| Choline quartiles | |||||
| Model 1 | 1.00 (Reference) | 1.09 (0.89, 1.35) | 1.22 (0.98, 1.51) | 1.28 (0.99, 1.64) | 0.044 |
| Model 2 | 1.00 (Reference) | 1.10 (0.89, 1.36) | 1.15 (0.92, 1.45) | 1.37 (1.03, 1.83) | 0.032 |
| Model 3 | 1.00 (Reference) | 1.14 (0.91, 1.43) | 1.22 (0.93, 1.59) | 1.54 (1.06, 2.25) | 0.029 |
| Betaine quartiles | |||||
| Model 1 | 1.00 (Reference) | 1.13 (0.91, 1.39) | 1.34 (1.08, 1.65) | 1.24 (0.96, 1.60) | 0.065 |
| Model 2 | 1.00 (Reference) | 1.14 (0.92, 1.41) | 1.40 (1.13, 1.74) | 1.45 (1.11, 1.90) | 0.003 |
| Model 3 | 1.00 (Reference | 1.11 (0.90, 1.37) | 1.33 (1.06, 1.66) | 1.31 (0.99, 1.74) | 0.039 |
| Male | |||||
| Choline quartiles | |||||
| Model 1 | 1.00 (Reference) | 0.96 (0.77, 1.20) | 1.07 (0.86, 1.34) | 1.13 (0.89, 1.44) | 0.196 |
| Model 2 | 1.00 (Reference) | 0.85 (0.67, 1.07) | 0.88 (0.70, 1.12) | 0.80 (0.60, 1.08) | 0.219 |
| Model 3 | 1.00 (Reference) | 0.85 (0.67, 1.07) | 0.88 (0.68, 1.14) | 0.82 (0.57, 1.17) | 0.360 |
| Betaine quartiles | |||||
| Model 1 | 1.00 (Reference) | 0.92 (0.74, 1.14) | 0.86 (0.69, 1.08) | 0.77 (0.60, 0.98) | 0.032 |
| Model 2 | 1.00 (Reference) | 0.88 (0.71, 1.10) | 0.81 (0.65, 1.03) | 0.78 (0.60, 1.02) | 0.079 |
| Model 3 | 1.00 (Reference) | 0.89 (0.77, 1.02) | 0.81 (0.64, 1.03) | 0.82 (0.62, 1.08) | 0.188 |
P value is P for trend computed by modeling the quartile median of each nutrient as the predictor variable. Model 1 mutually adjusted for choline and betaine and further adjusted for age, race, log-BMI, education, smoking, and alcohol. Model 2 further adjusted for total calorie intake and baseline fasting blood glucose. Model 3 further adjusted for magnesium, vitamin B6, vitamin B12, methionine, animal fat, and folate intake.
Conclusions
In this large cohort study, we observed no statistically significant association between dietary choline intake and the risk of type 2 diabetes, nor was dietary betaine intake significantly associated with the risk of type 2 diabetes. The associations remained similar among study participants without insulin resistance at baseline. Among female (but not male) participants, there was a trend for a modestly higher risk of type 2 diabetes among those with the highest as compared with the lowest quartile of dietary choline intake.
Choline, an essential nutrient (23,24) present in high quantities in red meat, eggs, beef, low-fat milk, chicken without skin, liver, and mustard (20), is involved in neurotransmission, lipid transport, cell membrane phospholipid synthesis, the one-carbon cycle as a precursor of the methyl donor betaine, and gene methylation (25). Betaine, mainly found in bread, spinach, and cold breakfast cereals (20), is a main methyl donor (one-carbon metabolism) in a variety of body cells (6). The distributions of top 10 food sources of choline and betaine in the ARIC study are listed in Bidulescu et al. (20). Humans can produce choline endogenously in the liver, mostly as phosphatidylcholine, but the amount that the body naturally synthesizes is not sufficient to meet human needs (24). Therefore, humans must obtain some choline from the diet. The most common sources of choline in foods are the fat-soluble phospholipids phosphatidylcholine and sphingomyelin as well as the water-soluble compounds phosphocholine, glycerophosphocholine, and free choline (20,23–25). When these choline-containing compounds are ingested, pancreatic and mucosal enzymes liberate free choline (19). Free choline, phosphocholine, and glycerophosphocholine are absorbed in the small intestine, enter the portal circulation, and are stored in the liver, where they are subsequently phosphorylated and distributed throughout the body (19,24). The remaining fat-soluble phospholipids (phosphatidylcholine and sphingomyelin) are absorbed intact, incorporated into chylomicrons, and secreted into the lymphatic circulation, where they are distributed to tissues and other organs (19,24). Despite the hypothesis that choline might affect cardiovascular outcomes, several large observational studies have found no significant associations between choline intakes and cardiovascular disease risk (20,24,25). Other studies suggest that higher dietary choline might increase cardiovascular disease risk because some choline and other dietary compounds are converted to TMA by intestinal bacteria. TMA is then absorbed and converted by the liver into TMAO, a compound that has been associated with a higher risk of cardiovascular disease (25–27).
Studies investigating the direct association between choline and betaine intake and the risk of type 2 diabetes are scant and contradictory. An animal study showed that choline-deficient mice on a high-fat diet had reduced fasting insulin and improved glucose tolerance, as compared with low-fat diet mice (28). In rodents, low intakes of dietary choline and methionine were related to enhanced insulin sensitivity (29), and intraperitoneal administration of choline increased serum insulin levels, perhaps by augmenting acetylcholine release in the pancreas (30). In other animal models, glucose downregulated methylene tetrahydrofolate reductase, an enzyme in the one-carbon cycle that is involved in the conversion of homocysteine to methionine and upregulated cystathionin β-synthase, another enzyme in one-carbon metabolism that converts homocysteine to cystathionine. Insulin had the opposite effect (31). This indicates an interaction between glucose metabolism and the one-carbon cycle. High homocysteine level was associated with the risk of type 2 diabetes (32).
In a study that combined results from three large cohorts, those in the highest as compared with the lowest phosphatidylcholine intake had 30% higher risk of type 2 diabetes; however, the authors found that high dietary free choline and betaine were associated with reduced risk of type 2 diabetes (33). In a cross-sectional study, participants with type 2 diabetes had high plasma concentrations of choline and low concentrations of betaine (34). Another study found that dietary choline and betaine intake were inversely correlated with fasting blood glucose and insulin levels (10). In a case-control study, plasma TMAO (a choline metabolite) was associated with 2.6 times higher odds of type 2 diabetes (9). In a metabolomics study conducted among European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam participants, choline-containing metabolites were positively associated with the risk of type 2 diabetes (8). In a Mexican American study sample, betaine and betaine to choline ratio were inversely associated with BMI and indicators of metabolic stress (35). In a small betaine supplementation study conducted among Japanese men, investigators did not find a statistically significant difference in fasting blood glucose, HbA1c (%), and fasting insulin levels between the supplementation group and placebo group over a 12-week follow-up period (36). In another study, plasma betaine levels were observed to be low among insulin-resistant humans, and betaine supplementation improved glucose homeostasis in animal models (37). In a previous ARIC study by our research group, dietary choline intake but not combination of choline and betaine intake was associated with a nonsignificant higher risk of coronary heart disease (20).
There is an established difference in choline metabolism by sex. Women of childbearing age have higher plasma betaine and choline, and various enzymes in the one-carbon cycle are upregulated or downregulated in women compared with men (38). Specifically, increase in phosphatidylethanolamine N-methyltransferase (PEMT) from male value to female value increased both plasma choline and betaine, while increasing betaine-homocysteine methyltransferase from male value to female value decreased betaine from male level (38). The differences are suggested to be the effect of sex hormones. It appears that the effect of choline and betaine persists more in females. PEMT+/+ diet-fed mice had 60% increased body mass and displayed insulin resistance, while PEMT−/− mice did not. The PEMT−/− mice had increased energy expenditure and maintained normal insulin sensitivity (39). Whether choline or its metabolites have epigenetic impacts on the risk of type 2 diabetes has not yet been established. Choline is converted to betaine, which donates the methyl group in the one-carbon cycle and thus is involved in DNA methylation. Whether the tendency for the modest association between choline and the risk of type 2 diabetes (as observed in this study among women) can be attributed to methylation of DNA or epigenetic modification is a subject of future investigation.
The strength of this study includes a large sample size, long prospective follow-up, and consistent results from both interval-censored survival analyses and logistic regression analysis. The study also has some limitations relevant to the interpretation of the results. The dietary intake data were collected using self-report with an FFQ, and measurement errors cannot be ruled out for nutrient intake; however, the FFQs have been validated by dietary records (18), and participants were ranked according to choline and betaine intake reducing the impact of measurement errors. Another limitation is that TMAO levels were not available to adjust for.
Conclusion
In our study, dietary choline or betaine intake were not associated with the increased risk of type 2 diabetes, although differences by sex were present. Our study should inform clinical trials on dietary choline and betaine supplementation in relationship with risk of type 2 diabetes.
Article Information
Acknowledgments. The authors thank the staff and participants of the ARIC study for the important contributions. A complete list of principal ARIC study staff can be found in Am J Epidemiol 1989;129:687–702.
Funding. The ARIC study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022. Laboratory testing and biospecimen collection at ARIC visit 6 was supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (NIH) (grant R01DK089174). A.B. was supported in part by two NIH Center for Scientific Review institutional training grants (HL07055 and DK07686). S.H.Z. was supported by research grants from the U.S. Department of Agriculture (58-1235-5-126) and the NIH Center for Scientific Review (DK55865 and DK56350).
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. D.T.D. conceived the current study, conducted data analysis, interpreted the results, and drafted the manuscript. A.B. led the study on dietary choline and betaine data collection/compositions, provided the data sets, and contributed to the revisions on the successive versions of the manuscript. D.T.D., K.C.J., A.M.K.-N., K.M., and S.H.Z. critically reviewed and revised the manuscript. All authors approved the final version of the manuscript. A.B. and D.T.D. are the guarantors of this work and, as such, had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Footnotes
This article contains supplementary material online at https://doi.org/10.2337/figshare.12814169.
References
- 1.Ogurtsova K, da Rocha Fernandes JD, Huang Y, et al. . IDF Diabetes Atlas: global estimates for the prevalence of diabetes for 2015 and 2040. Diabetes Res Clin Pract 2017;128:40–50 [DOI] [PubMed] [Google Scholar]
- 2.Xu G, Liu B, Sun Y, et al. . Prevalence of diagnosed type 1 and type 2 diabetes among US adults in 2016 and 2017: population based study. BMJ 2018;362:k1497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Forouhi NG, Misra A, Mohan V, Taylor R, Yancy W. Dietary and nutritional approaches for prevention and management of type 2 diabetes. BMJ 2018;361:k2234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Barker J, Su F, Alwan NA. Risk factors for type 2 diabetes after gestational diabetes: a population-based cohort study. Lancet 2017;390:S21 [Google Scholar]
- 5.Stern MP, Williams K, Haffner SM. Identification of persons at high risk for type 2 diabetes mellitus: do we need the oral glucose tolerance test? Ann Intern Med 2002;136:575–581 [DOI] [PubMed] [Google Scholar]
- 6.Virtanen JK, Tuomainen TP, Voutilainen S. Dietary intake of choline and phosphatidylcholine and risk of type 2 diabetes in men: the Kuopio Ischaemic Heart Disease Risk Factor Study. Eur J Nutr. 20 March 2020 [Epub ahead of print]. DOI: 10.1007/s00394-020-02223-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Li Y, Wang DD, Chiuve SE, et al. . Dietary phosphatidylcholine intake and type 2 diabetes in men and women. Diabetes Care 2015;38:e13–e14 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Floegel A, Stefan N, Yu Z, et al. . Identification of serum metabolites associated with risk of type 2 diabetes using a targeted metabolomic approach. Diabetes 2013;62:639–648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shan Z, Sun T, Huang H, et al. . Association between microbiota-dependent metabolite trimethylamine-N-oxide and type 2 diabetes. Am J Clin Nutr 2017;106:888–894 [DOI] [PubMed] [Google Scholar]
- 10.Gao X, Wang Y, Sun G. High dietary choline and betaine intake is associated with low insulin resistance in the Newfoundland population. Nutrition 2017;33:28–34 [DOI] [PubMed] [Google Scholar]
- 11.Sadre-Marandi F, Dahdoul T, Reed MC, Nijhout HF. Sex differences in hepatic one-carbon metabolism. BMC Syst Biol 2018;12:89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cai CX, Carlos S, Solaimani P, Trivedi BJ, Tran C, Castelino-Prabhu S. Nutritional and dietary interventions for nonalcoholic fatty liver disease. In Dietary Interventions in Liver Disease: Foods, Nutrients, and Dietary Supplements. Watson RR, Preedy VR, Eds. London, Academic Press, 2019, pp. 357–372 [Google Scholar]
- 13.The ARIC Study Investigators The decline of ischaemic heart disease mortality in the ARIC study communities. Int J Epidemiol 1989;18(Suppl. 1):S88–S98 [DOI] [PubMed] [Google Scholar]
- 14.Miyagawa N, Miura K, Okuda N, et al.; NIPPON DATA80 Research Group . Long-chain n-3 polyunsaturated fatty acids intake and cardiovascular disease mortality risk in Japanese: a 24-year follow-up of NIPPON DATA80. Atherosclerosis 2014;232:384–389 [DOI] [PubMed] [Google Scholar]
- 15.Gayoso-Diz P, Otero-González A, Rodriguez-Alvarez MX, et al. . Insulin resistance (HOMA-IR) cut-off values and the metabolic syndrome in a general adult population: effect of gender and age: EPIRCE cross-sectional study. BMC Endocr Disord 2013;13:47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sumner AE, Cowie CC. Ethnic differences in the ability of triglyceride levels to identify insulin resistance. Atherosclerosis 2008;196:696–703 [DOI] [PubMed] [Google Scholar]
- 17.American Diabetes Association Diagnosis and classification of diabetes mellitus [published correction appears in Diabetes Care 2010;33:e57] Diabetes Care 2010;33(Suppl. 1):S62–S69 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Willett WC, Sampson L, Stampfer MJ, et al. . Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:51–65 [DOI] [PubMed] [Google Scholar]
- 19.Zeisel SH, Mar MH, Howe JC, Holden JM. Concentrations of choline-containing compounds and betaine in common foods. J Nutr 2003;133:1302–1307 [DOI] [PubMed] [Google Scholar]
- 20.Bidulescu A, Chambless LE, Siega-Riz AM, Zeisel SH, Heiss G. Usual choline and betaine dietary intake and incident coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study. BMC Cardiovasc Disord 2007;7:20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shai I, Rimm EB, Hankinson SE, et al. . Lipoprotein (a) and coronary heart disease among women: beyond a cholesterol carrier? Eur Heart J 2005;26:1633–1639 [DOI] [PubMed] [Google Scholar]
- 22.Sotos-Prieto M, Bhupathiraju SN, Mattei J, et al. . Association of changes in diet quality with total and cause-specific mortality. N Engl J Med 2017;377:143–153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Institute of Medicine; Food and Nutrition Board; National Academy of Sciences Choline. In Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline. Washington, D.C., National Academies Press, 1998, pp. 390–422 [PubMed] [Google Scholar]
- 24.Zeisel SH, da Costa K-A. Choline: an essential nutrient for public health. Nutr Rev 2009;67:615–623 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wiedeman AM, Barr SI, Green TJ, Xu Z, Innis SM, Kitts DD. Dietary choline intake: current state of knowledge across the life cycle. Nutrients 2018;10:1513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Tang WH, Wang Z, Levison BS, et al. . Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med 2013;368:1575–1584 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zheng Y, Li Y, Rimm EB, et al. . Dietary phosphatidylcholine and risk of all-cause and cardiovascular-specific mortality among US women and men. Am J Clin Nutr 2016;104:173–180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Raubenheimer PJ, Nyirenda MJ, Walker BR. A choline-deficient diet exacerbates fatty liver but attenuates insulin resistance and glucose intolerance in mice fed a high-fat diet. Diabetes 2006;55:2015–2020 [DOI] [PubMed] [Google Scholar]
- 29.Wu G, Zhang L, Li T, Lopaschuk G, Vance DE, Jacobs RL. Choline deficiency attenuates body weight gain and improves glucose tolerance in ob/ob mice. J Obes 2012;2012:319172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ilcol YO, Gurun MS, Taga Y, Ulus IH. Choline increases serum insulin in rat when injected intraperitoneally and augments basal and stimulated aceylcholine release from the rat minced pancreas in vitro. Eur J Biochem 2003;270:991–999 [DOI] [PubMed] [Google Scholar]
- 31.Dicker-Brown A, Fonseca VA, Fink LM, Kern PA. The effect of glucose and insulin on the activity of methylene tetrahydrofolate reductase and cystathionine-beta-synthase: studies in hepatocytes. Atherosclerosis 2001;158:297–301 [DOI] [PubMed] [Google Scholar]
- 32.Huang T, Ren J, Huang J, Li D. Association of homocysteine with type 2 diabetes: a meta-analysis implementing Mendelian randomization approach. BMC Genomics 2013;14:867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Li Y, Wang D, Chiuve S, et al. . Major diet sources of gut microbiota metabolites and type 2 diabetes (121.8). FASEB J 2014;28(Suppl. 1):121.8 [Google Scholar]
- 34.Svingen G, Ueland PM, Schartum-Hansen H, Pedersen ER, Seifert R, Nygard OK. Metabolites in the choline oxidation pathway in relation to diabetes, glycemic control and insulin sensitivity among patients with angina pectoris. Atherosclerosis 2014;235:e51 [Google Scholar]
- 35.Yan J, Winter LB, Burns-Whitmore B, Vermeylen F, Caudill MA. Plasma choline metabolites associate with metabolic stress among young overweight men in a genotype-specific manner. Nutr Diabetes 2012;2:e49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Tiihonen K, Saarinen MT, Alhoniemi E, Mitsuya N, Yamaki G. Effect of dietary betaine on metabolic syndrome risk factors in Asian males with mild fatty liver. J Diabetes Metab 2016;7:692 [Google Scholar]
- 37.Ejaz A, Martinez-Guino L, Goldfine AB, et al. . Dietary betaine supplementation increases Fgf21 levels to improve glucose homeostasis and reduce hepatic lipid accumulation in mice. Diabetes 2016;65:902–912 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sadre-Marandi F, Dahdoul T, Reed MC, Nijhout HF. Sex differences in hepatic one-carbon metabolism. BMC Syst Biol 2018;12:89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Jacobs RL, Zhao Y, Koonen DP, et al. . Impaired de novo choline synthesis explains why phosphatidylethanolamine N-methyltransferase-deficient mice are protected from diet-induced obesity. J Biol Chem 2010;285:22403–22413 [DOI] [PMC free article] [PubMed] [Google Scholar]