Coffee Consumption Is Associated With Higher Plasma Adiponectin Concentrations in Women With or Without Type 2 Diabetes
Response to Williams et al.
- Kazuhiko Kotani, MD, PHD12,
- Shinji Fujiwara, MD, PHD3,
- Taku Hamada, PHD1,
- Kokoro Tsuzaki, BSC1 and
- Naoki Sakane, MD, PHD1
- 1Department of Preventive Medicine and Diabetes Education, Clinical Research Institute for Endocrine and Metabolic Disease, National Hospital Organization Kyoto Medical Center, Kyoto, Japan
- 2Division of Health Administration and Promotion, Faculty of Medicine, Tottori University, Yonago, Japan
- 3Division of Community Medicine and Family Practice, Jichi Medical University, Tochigi, Japan
- Corresponding author: Kazuhiko Kotani, MD, PHD, Division of Health Administration and Promotion, Faculty of Medicine, Tottori University, 86 Nishi-cho, Yonago 683-8503, Japan. E-mail: kakotani{at}grape.med.tottori-u.ac.jp
Williams et al. reported that women consuming ≥4 cups of coffee daily had higher adiponectin levels than those consuming lower amounts, regardless of the presence of diabetes (1). They found that regular coffee consumption has beneficial effects on low-grade systemic inflammation and type 2 diabetes incidence and that adiponectin may be involved in the mechanism (1). Given variances in different ethnicities' social cultures regarding coffee consumption and in adiponectin levels across ethnicity (2), Williams et al.'s data should be confirmed on various populations. Also up for discussion is whether the data are applicable to men because of the sex-based differences in adiponectin (3).
We studied 665 (205 men and 460 women) asymptomatic, rural community–dwelling Japanese volunteers (mean ± SD age 59 ± 14 years) without a history of cardiovascular disease, assessed using self-reported lifestyle parameters such as usual coffee consumption (categorized as none–few, 1–3 cups/day, and ≥4 cups/day), smoking, alcohol consumption, and exercise. In addition to overnight fasted routine measurements (i.e., blood pressure and serum cholesterol), serum adiponectin was measured using an ELISA (Otsuka Pharmaceutical, Tokyo, Japan).
This population included a low percentage of subjects that consumed ≥4 cups/day, with subjects who consumed more coffee tending to be younger. In men, none–few cups, n = 103 and age 65 ± 14 years; 1–3 cups, n = 88 and age 57 ± 15 years; and ≥4 cups, n = 14 and age 52 ± 17 years. In women, n = 145, age 65 ± 12 years; n = 292, age 55 ± 13 years; and n = 23, age 49 ± 14 years, respectively. There were nonsignificant differences in adiponectin among the groups of coffee-consumption categories in both sexes (P > 0.05 by ANOVA): 1) men, none–few cups, 7.6 ± 5.7 μg/ml; 1–3 cups, 6.0 ± 3.5 μg/ml; and ≥4 cups, 8.1 ± 4.9 μg/ml, and similarly, 2) women, 10.6 ± 5.4, 12.0 ± 6.9, and 10.6 ± 7.2 μg/ml, respectively. BMI and waist circumference did not differ among either group (data not shown). The association between coffee consumption and adiponectin was unchanged even through general linear model analyses (P > 0.05 in all models) on logarithm-transformed adiponectin with adjustments for multiple covariates such as age and BMI (model 1, F value of coffee consumption in men/women = 1.5/2.1), on model 1 variables plus other lifestyle parameters (model 2, F = 0.8/2.0), and on model 2 variables plus the presence of hypertension (defined as ≥140 mmHg of systolic and/or ≥90 mmHg of diastolic blood pressure) and hypercholesterolemia (defined as ≥5.72 mmol/l of total cholesterol) (F = 1.2/1.6).
Thus, the benefits of coffee consumption to adiponectin levels were not confirmed in this population. This may be partly explained by some environmental/genetic contributors to the ethnic differences in adiponectin. We should also consider reduced diabetes risks with high coffee consumption (3–6 cups/day) (1). Even if coffee consumption increases adiponectin, presently the appreciable benefits seem to be unperceived in a Japanese population. The communities studied themselves do not, based on their low percentage of heavy coffee drinkers, recognize its potential health value.
