ABSTRACT
Background
Coffee consumption has been linked to lower risk of various health outcomes. However, the biological pathways mediating the associations remain poorly understood.
Objectives
The aim of this study was to assess the association between coffee consumption and concentrations of plasma biomarkers in key metabolic and inflammatory pathways underlying common chronic diseases.
Methods
We investigated the associations of total, caffeinated, and decaffeinated coffee consumption with 14 plasma biomarkers, including C-peptide, insulin-like growth factor 1 (IGF-1), IGF binding protein (IGFBP) 1, IGFBP-3, estrone, total and free estradiol, total and free testosterone, sex hormone–binding globulin (SHBG), total adiponectin, high-molecular-weight (HMW) adiponectin, leptin, C-reactive protein (CRP), interleukin 6 (IL-6), and soluble tumor necrosis factor receptor 2 (sTNFR-2). Data were derived from 2 cohorts of 15,551 women (Nurses’ Health Study) and 7397 men (Health Professionals Follow-Up Study), who provided detailed dietary data before blood draw and were free of diabetes, cardiovascular disease, or cancer at the time of blood draw. Multivariable linear regression was used to calculate the percentage difference of biomarker concentrations comparing coffee drinkers with nondrinkers, after adjusting for a variety of demographic, clinical, and lifestyle factors.
Results
Compared with nondrinkers, participants who drank ≥4 cups of total coffee/d had lower concentrations of C-peptide (−8.7%), IGFBP-3 (−2.2%), estrone (−6.4%), total estradiol (−5.7%), free estradiol (−8.1%), leptin (−6.4%), CRP (−16.6%), IL-6 (−8.1%), and sTNFR-2 (−5.8%) and higher concentrations of SHBG (5.0%), total testosterone (7.3% in women and 5.3% in men), total adiponectin (9.3%), and HMW adiponectin (17.2%). The results were largely similar for caffeinated and decaffeinated coffee.
Conclusion
Our data indicate that coffee consumption is associated with favorable profiles of numerous biomarkers in key metabolic and inflammatory pathways. This trial was registered at clinicaltrials.gov as NCT03419455.
Keywords: coffee consumption, inflammation, adipokine, sex hormone, insulin
Introduction
Coffee is among the most commonly consumed beverages worldwide. According to the National Coffee Association's report in 2017, 62% of US adults drink coffee daily (1). Epidemiologic evidence indicates that coffee consumption is associated with a lower risk of various chronic diseases, including type 2 diabetes (T2D), cardiovascular disease (CVD), and certain types of cancer, as well as reduced overall mortality (2–4). Two recent umbrella reviews of meta-analyses further support the health benefits of coffee within usual amounts of intake (5, 6). However, the biological pathways underlying the associations remain largely unknown.
The development of T2D, CVD, and some cancers is closely related with states of hyperinsulinemia, insulin resistance, sex hormone dysregulation, and chronic inflammation. For example, circulating concentrations of insulin-related biomarkers [e.g., C-peptide, insulin-like growth factor 1 (IGF-1), IGF binding protein (IGFBP) 1 and 3], sex hormones [e.g., estrogens, testosterone, and sex hormone–binding globulin (SHBG)], adipokines (e.g., adiponectin and leptin), and inflammatory biomarkers [e.g., C-reactive protein (CRP), IL-6, and soluble TNF receptor 2 (sTNFR-2)] have been linked to T2D, CVD, and cancer risk (7–10). As a rich source of bioactive compounds (e.g., caffeine, chlorogenic acids, and diterpenes), coffee may elicit a multitude of physiologic effects and affect these biomarkers. A few studies have investigated coffee consumption in relation to plasma biomarkers of insulin secretion (11, 12), the IGF system (13, 14), sex hormones (15, 16), adiponectin (17, 18), and inflammation (18–21). However, the results remain inconsistent. Also, many of the studies are limited by their modest size (12, 13, 15), lack of repeated dietary assessment (11, 13–16, 18–20), or inadequate control for confounding by other lifestyle factors related to coffee intake (13, 14). Evidence from clinical trials suggests that coffee consumption may increase total testosterone (22) and adiponectin (23) but has no effect on SHBG (22), CRP, or IL-6 (24). However, these interventional studies are limited by their small sample sizes (<120), short duration (8 wk–3 mo), and highly selected participants with certain health conditions (e.g., overweight persons).
To address these limitations, we performed a comprehensive assessment of 14 plasma biomarkers of major pathways underlying coffee-related diseases, including the insulin/IGF system (C-peptide, IGF-1, IGFBP-1, IGFBP-3), sex hormones (estrone, total and free estradiol, total and free testosterone, and SHBG), adipokines [total adiponectin, high-molecular-weight (HMW) adiponectin, and leptin], and inflammation (CRP, IL-6, and sTNFR-2), within 2 large prospective cohort studies, the Nurses’ Health Study (NHS) and the Health Professionals Follow-Up Study (HPFS). Our primary hypothesis was that higher coffee intake is associated with a favorable profile of these biomarkers.
Methods
Study population
The NHS and HPFS are 2 ongoing prospective cohort studies that enrolled 121,700 female registered nurses aged 30–55 y at baseline in 1976 (25) and 51,529 male health professionals aged 40–75 y at baseline in 1986 (26), respectively. In each cohort, mailed questionnaires were administered biennially to collect data on numerous lifestyle and medical factors, with a follow-up rate >90% for each 2-y cycle. Between 1989 and 1990, 32,826 women from the NHS provided blood specimens on ice packs by overnight courier, as did 18,225 men from the HPFS between 1993 and 1995. As previously reported, participants who provided blood samples were similar to those who did not in terms of demographic characteristics, diet, and lifestyle (27). Detailed procedures for blood collection, handling, and storage have been described elsewhere (28).
In the current study, we used data on 17,881 women and 8748 men from previous nested case-control studies within the 2 cohorts that had measured plasma biomarkers of our interest. These studies were conducted for various outcomes, including T2D, CVD, and several common cancers (e.g., breast, colorectum, pancreas, endometrium, ovary, prostate, and multiple myeloma). Participants who provided blood samples in the cohorts were selected for these case-control studies using risk set sampling, in which cases who developed the outcome of study during follow-up were identified and matched with controls who were free of the outcome at the time of diagnosis for cases. For the current study, we excluded participants who had a history of diabetes, CVD, or cancer at the time of blood draw; had missing data on coffee consumption or erroneous records; or had biomarker concentrations considered as outliers; this left 15,551 women and 7397 men in the final analysis (Supplemental Figure 1). This study was approved by the institutional review boards at Brigham and Women's Hospital and Harvard TH Chan School of Public Health. Written informed consent was obtained from all participants.
Biomarker assessment
The biomarkers were chosen based on the following criteria: 1) commonly studied for interrogating the etiologic relevance of insulin, inflammation, and sex hormone pathways; 2) shown in epidemiologic studies to be related to T2D, CVD, or cancer; and 3) previously investigated within the NHS and HPFS. All biomarkers were measured using standard methods (Supplemental Table 1) (29–32). Quality-control samples were randomly interspersed among the case-control samples, and laboratory personnel were blinded to quality-control and case-control status for all assays. In quality-control samples, the intra-assay CV ranged from 1% to 20% for all biomarkers across batches. Because biomarkers were measured in multiple batches over time and there might be variation in mean biomarker concentrations due to differences in reagents, technicians, and laboratories, we recalibrated biomarker concentrations across batches within each cohort to the value of an “average batch” using the method of Rosner et al. (33) (Supplemental Methods 1). Free estradiol and testosterone were calculated using a validated algorithm based on total estradiol or testosterone, SHBG, an assumed constant representing the normal albumin concentration, and the association constants for the binding of estradiol and testosterone to SHBG and albumin (34).
Assessment of coffee consumption
Starting in 1986, a 131-item food-frequency questionnaire (FFQ) was administered every 4 y to collect updated dietary data. In each FFQ, participants were asked how often (ranging from “never or less than once per month” to “6 or more times per day”), on average, they consumed a standard portion size of each food item during the previous year. For coffee, the questionnaire inquired about the consumption frequency for caffeinated and decaffeinated coffee, separately, in a standard cup (8 ounces/237 mL). We calculated total coffee consumption as the sum of these 2 types of coffee. The validity and reproducibility of the FFQs have been reported previously (35), with a high correlation between coffee consumption assessed by the FFQ and repeated 1-wk diet records (r = 0.78). In the current study, to capture the habitual intake that is most relevant to biomarker concentrations and reduce within-person variability, we calculated the average of coffee intake using the 2 FFQs administered most proximately to blood draw (i.e., 1986 and 1990 for women and 1990 and 1994 for men). For those who provided coffee information in only 1 of the 2 FFQs (1806 women and 591 men), the intake reported in that available FFQ was used.
Statistical analysis
The outliers in each set of samples were identified by using the generalized extreme studentized deviate many-outlier procedure (36). We performed ln transformation for all biomarkers to improve the normality. We calculated the partial Spearman rank correlation coefficients between each 2 of the biomarkers after adjusting for sex and age at blood draw. Multivariate linear regression analysis was conducted to examine the associations between coffee consumption and plasma biomarker concentrations, with nondrinkers as the reference. Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalents of task-h/wk), smoking (never; past: <30; past: ≥30; current; <30; or current ≥30 pack-years), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), Alternate Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except for coffee (tea, cola beverage, and chocolate) (37, 38), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). The results of the multivariate regression analysis were presented as percentage differences in biomarker concentrations compared with the reference group by using the equation: [exp (β-coefficient) − 1] × 100%. Details about covariate assessments are provided in Supplemental Methods 2.
For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. To test for potential nonlinearity, we performed restricted cubic spline analyses and used a likelihood ratio test to compare the model with only the linear term of coffee consumption with the model with both the linear and spline terms. We did not find strong statistical evidence for nonlinearity and thus present the results based on the linear models.
Because the results were largely similar between the NHS and HPFS cohorts, we pooled the data for all biomarkers except for testosterone, for which we analyzed the data separately for women and men due to marked sex differences in its sources, circulating concentrations, and biological functions (39). We performed sex-specific analyses for all other biomarkers in the secondary analysis. To assess potential effect modification, we conducted stratified analyses by demographic and lifestyle factors (Supplemental Methods 3).
Finally, we conducted sensitivity analyses by restricting the analytic samples to the participants who provided fasting blood, were selected as controls in previous case-control studies, and reported no history of hypertension or hypercholesterolemia at blood draw. All statistical tests were 2-sided, and SAS version 9.4 (SAS Institute) was used for all analyses.
Results
Table 1 shows the characteristics of study participants according to total coffee consumption. Participants who drank more total coffee were more likely to be current smokers, drink more alcohol, have higher total caloric intake and lower caffeine intake from other sources (tea, cola beverage, and chocolate), and regularly use aspirin or nonsteroidal anti-inflammatory drugs. Similar patterns were observed among participants who drank more caffeinated coffee (Supplemental Table 2) and decaffeinated coffee (Supplemental Table 3).
TABLE 1.
Characteristics of participants at blood draw according to frequency of total coffee consumption in women from the Nurses’ Health Study and men from the Health Professionals Follow-Up Study1
| Women (n = 15,551) | Men (n = 7397) | |||||||
|---|---|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2–3 cups/d | ≥4 cups/d | Nondrinker | ≤1 cup/d | 2–3 cups/d | ≥4 cups/d |
| Participants, n (%) | 1354 (9) | 2690 (17) | 7217 (46) | 4290 (28) | 963 (13) | 1705 (23) | 2983 (40) | 1746 (24) |
| Age at blood draw, y | 55.8 ± 7.4 | 57.5 ± 7.2 | 57.8 ± 6.9 | 56.9 ± 6.7 | 60.8 ± 8.8 | 63.0 ± 8.8 | 62.9 ± 8.5 | 61.7 ± 7.9 |
| White, % | 98 | 96 | 98 | 99 | 94 | 92 | 94 | 94 |
| Total coffee,2 cups/d | 0 | 0.6 ± 0.3 | 2.2 ± 0.5 | 4.4 ± 1.1 | 0 | 0.6 ± 0.3 | 2.1 ± 0.5 | 4.4 ± 1.0 |
| Caffeinated coffee,2 cups/d | 0 | 0.3 ± 0.4 | 1.4 ± 1.0 | 3.0 ± 1.6 | 0 | 0.3 ± 0.3 | 1.5 ± 0.9 | 3.1 ± 1.6 |
| Decaffeinated coffee,2 cups/d | 0 | 0.3 ± 0.3 | 0.8 ± 0.9 | 1.4 ± 1.5 | 0 | 0.3 ± 0.3 | 0.7 ± 0.8 | 1.3 ± 1.5 |
| BMI,2 kg/m2 | 26.2 ± 5.3 | 25.9 ± 5.0 | 25.4 ± 4.5 | 25.2 ± 4.4 | 25.4 ± 3.2 | 25.4 ± 3.3 | 25.8 ± 3.2 | 26.1 ± 3.3 |
| Physical activity,2 MET-h/wk | 17.5 ± 17.9 | 17.6 ± 21.2 | 18.2 ± 18.8 | 17.7 ± 18.1 | 32.8 ± 25.6 | 32.8 ± 27.9 | 32.4 ± 25.5 | 30.8 ± 25.2 |
| Smoking status, % | ||||||||
| Never | 64 | 59 | 44 | 34 | 71 | 56 | 42 | 33 |
| Past | 29 | 35 | 45 | 42 | 26 | 42 | 53 | 54 |
| Current | 7 | 6 | 11 | 24 | 3 | 3 | 5 | 13 |
| Alcohol consumption,2 g/d | 2.9 ± 7.5 | 4.3 ± 8.2 | 6.8 ± 9.9 | 6.7 ± 9.7 | 5.4 ± 10.3 | 10.3 ± 13.7 | 13.4 ± 14.8 | 14.1 ± 14.8 |
| AHEI2 | 44.8 ± 9.8 | 45.9 ± 9.1 | 45.8 ± 9.0 | 45.8 ± 9.1 | 42.6 ± 9.0 | 42.8 ± 9.1 | 41.8 ± 9.1 | 40.8 ± 9.1 |
| Total energy intake,2 kcal/d | 1720 ± 446 | 1753 ± 442 | 1777 ± 443 | 1823 ± 458 | 1978 ± 510 | 1947 ± 549 | 1999 ± 552 | 2094 ± 562 |
| Caffeine intake from other sources,2,3 mg/d | 87.6 ± 80.2 | 58.1 ± 57.0 | 47.8 ± 44.4 | 40.7 ± 43.3 | 53.2 ± 67.0 | 44.8 ± 52.1 | 37.6 ± 42.6 | 32.7 ± 42.1 |
| Hypertension | 22 | 25 | 22 | 18 | 23 | 25 | 26 | 26 |
| Hypercholesterolemia | 34 | 42 | 39 | 36 | 30 | 35 | 36 | 35 |
| Postmenopausal | 73 | 75 | 74 | 75 | NA | NA | NA | NA |
| Current menopausal hormone use4 | 42 | 43 | 42 | 39 | NA | NA | NA | NA |
| Regular multivitamin use | 37 | 41 | 38 | 38 | 49 | 51 | 49 | 49 |
| Regular aspirin or NSAID use5 | 38 | 37 | 38 | 41 | 44 | 49 | 54 | 55 |
| Insulin/IGF markers,6 ng/mL | ||||||||
| C-peptide | 1.6 (1.1–2.5) | 1.6 (1.2–2.4) | 1.6 (1.1–2.3) | 1.5 (1.1–2.1) | 2.0 (1.4–3.3) | 2.0 (1.4–3.1) | 2.0 (1.4–3.1) | 2.0 (1.5–3.0) |
| IGF-1 | 151 (119–187) | 149 (117–194) | 151 (119–189) | 151 (121–191) | 140 (116–168) | 135 (113–161) | 137 (113–164) | 134 (113–160) |
| IGFBP-1 | 26.9 (14.2–43.5) | 26.8 (13.4–46.4) | 26.9 (14.2–44.7) | 28.4 (15.6–45.3) | 13.9 (6.9–25.0) | 15.6 (7.0–27.1) | 13.6 (6.9–23.8) | 15.3 (8.2–26.2) |
| IGFBP-3 | 4355 (3768–4930) | 4408 (3838–5105) | 4390 (3797–5013) | 4306 (3733–4942) | 3912 (3402–4454) | 3831 (3276–4289) | 3821 (3324–4329) | 3800 (3317–4265) |
| Sex hormones6 | ||||||||
| Estrone, pg/mL | 25.3 (17.8–34.6) | 26.1 (18.9–39.0) | 25.1 (18.2–34.8) | 23.4 (17.2–33.3) | 30.3 (24.2–35.5) | 28.6 (24.3–34.1) | 30.3 (23.4–37.3) | 27.7 (22.5–34.1) |
| Total estradiol, pg/mL | 6.2 (4.1–9.4) | 6.3 (4.4–10.5) | 5.7 (4.1–8.7) | 5.4 (3.8–8.6) | 23.1 (20.3–27.9) | 23.0 (19.1–26.8) | 23.1 (19.0–27.7) | 23.2 (20.1–28.1) |
| Free estradiol, pg/mL | 0.09 (0.05–0.12) | 0.09 (0.05–0.16) | 0.08 (0.05–0.12) | 0.07 (0.05–0.11) | 0.42 (0.36–0.49) | 0.41 (0.35–0.49) | 0.43 (0.34–0.52) | 0.42 (0.34–0.50) |
| SHBG, nmol/L | 67.5 (46.3–106.8) | 68.3 (44.2–111.0) | 69.4 (46.2–102.8) | 71.6 (49.3–102.9) | 29.1 (22.1–38.4) | 29.4 (23.0–37.5) | 29.2 (22.6–37.1) | 30.2 (24.4–37.9) |
| Total testosterone, ng/dL | 18.8 (12.4–24.6) | 18.9 (13.6–26.4) | 18.9 (13.6–26.5) | 19.6 (14.2–27.7) | 448 (335–552) | 432 (353–536) | 440 (340–553) | 463 (358–551) |
| Free testosterone, ng/dL | 0.13 (0.08–0.20) | 0.14 (0.08–0.21) | 0.14 (0.09–0.21) | 0.14 (0.09–0.21) | 8.1 (6.0–9.9) | 8.0 (6.0–9.9) | 8.2 (6.4–10.1) | 8.2 (6.5–10.0) |
| Adipokines and inflammatory markers6 | ||||||||
| Total adiponectin, ng/mL | 9769 (6780–12,487) | 10,236 (7361–12,570) | 10,436 (7799–13,184) | 10,771 (8242–13,463) | 5323 (3857–7439) | 5788 (4182–8133) | 5838 (4232–7895) | 6111 (4316–8456) |
| HMW adiponectin, ng/mL | 4314 (2688–7104) | 4913 (2829–7446) | 5332 (3446–8049) | 5825 (3939–8286) | 1836 (1422–2941) | 1923 (1249–3136) | 2408 (1436–3554) | 2325 (1515–3833) |
| Leptin, ng/mL | 26.4 (14.5–40.1) | 25.7 (15.1–38.6) | 22.2 (12.8–36.8) | 21.1 (12.0–34.5) | 6.7 (4.3–10.1) | 6.5 (4.1–10.3) | 6.9 (4.7–10.6) | 7.0 (4.4–11.2) |
| CRP, mg/L | 2.1 (0.8–4.5) | 2.0 (0.9–4.2) | 1.8 (0.8–3.8) | 1.5 (0.7–3.3) | 0.9 (0.5–2.0) | 1.0 (0.5–2.1) | 1.0 (0.5–2.1) | 1.1 (0.6–2.2) |
| IL-6, pg/mL | 1.3 (0.8–2.0) | 1.2 (0.8–1.9) | 1.2 (0.8–1.7) | 1.1 (0.7–1.7) | 1.1 (0.7–1.5) | 1.0 (0.7–1.5) | 1.0 (0.7–1.6) | 1.1 (0.7–1.6) |
| sTNFR-2, pg/mL | 2694 (2283–3216) | 2597 (2194–3099) | 2512 (2133–2969) | 2491 (2123–2916) | 2625 (2231–3029) | 2522 (2121–3050) | 2466 (2108–2922) | 2459 (2111–2893) |
Values are means ± SDs for continuous variables and percentages for categorical variables unless otherwise specified. All variables are standardized by age at blood draw squared except for age and biomarkers. AHEI, Alternative Healthy Eating Index; CRP, C-reactive protein; HMW, high-molecular-weight; IGF-1, insulin-like growth factor 1; IGFBP, insulin-like growth factor binding protein; MET, metabolic equivalent of task; NSAID, nonsteroidal anti-inflammatory drug; SHBG, sex hormone-binding globulin; sTNFR-2, soluble TNF receptor 2.
Cumulative average assessments before blood draw were used to derive the values.
Calculated by summing the caffeine contents from tea, cola beverage, and chocolate.
Defined in menopausal women only.
Regular users are defined as ≥2 tablets of aspirin (325 mg/tablet) or NSAIDs/wk.
The ln-transformed biomarker concentrations were back transformed and are presented as median values (IQRs).
Most of the biomarkers were weakly or moderately correlated (rs ranged from −0.54 to 0.68), except for estrone and total estradiol (rs = 0.82 in women and 0.70 in men), and for total and HMW adiponectin (rs = 0.91 in women and 0.85 in men) (Supplemental Table 4).
Tables 2 and 3 show the associations between coffee consumption and biomarker concentrations of the insulin/IGF system. Comparing participants who drank ≥4 cups of total coffee/d with nondrinkers, the percentage differences (95% CIs) after adjusting for demographic, lifestyle, and medical variables were −8.7% (−12.0%, −5.2%) and −2.2% (−3.7%, −0.6%) for C-peptide and IGFBP-3, respectively. Similar associations with C-peptide were observed for caffeinated and decaffeinated coffee, whereas only caffeinated coffee showed an inverse association with IGFBP-3 (percentage difference for ≥4 cups/d compared with none: −3.7%; 95% CI: −5.1%, −2.3%), and only decaffeinated coffee showed a positive association with IGF-1 (percentage difference for ≥3 cups/d compared with none: 2.8%; 95% CI: 0.7%, 4.9%).
TABLE 2.
Percentage differences (95% CIs) in concentrations of plasma biomarkers of the insulin/IGF system according to total and caffeinated coffee consumption among participants from the Nurses’ Health Study and the Health Professionals Follow-Up Study1
| Categories of coffee consumption | ||||||
|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2–3 cups/d | ≥4 cups/d | Per cup increase | P-trend |
| Total coffee | ||||||
| C-peptide | ||||||
| n | 986 | 1948 | 4462 | 2434 | ||
| Model 1 | 0 (ref) | −2.6 (−6.6, 1.7) | −3.6 (−7.2, 0.3) | −6.5 (−10.3, −2.5) | −1.1 (−1.8, −0.4) | 0.001 |
| Model 2 | 0 (ref) | −2.3 (−5.9, 1.4) | −4.3 (−7.5, −0.9) | −8.7 (−12.0, −5.2) | −1.9 (−2.5, −1.3) | <0.001 |
| IGF-1 | ||||||
| n | 1080 | 2020 | 4483 | 2587 | ||
| Model 1 | 0 (ref) | 1.8 (−0.6, 4.2) | 2.0 (−0.2, 4.3) | 2.0 (−0.4, 4.4) | 0.1 (−0.3, 0.4) | 0.81 |
| Model 2 | 0 (ref) | 2.3 (−0.1, 4.7) | 2.5 (0.3, 4.7) | 2.3 (−0.03, 4.7) | 0.01 (−0.4, 0.4) | 0.96 |
| IGFBP-1 | ||||||
| n | 681 | 1384 | 3003 | 1629 | ||
| Model 1 | 0 (ref) | −2.7 (−10.4, 5.6) | −5.2 (−12.0, 2.1) | −0.1 (−7.8, 8.2) | 0.4 (−1.0, 1.7) | 0.60 |
| Model 2 | 0 (ref) | −0.8 (−7.5, 6.4) | −2.5 (−8.6, 4.1) | 2.8 (−4.3, 10.3) | 0.9 (−0.3, 2.1) | 0.15 |
| IGFBP-3 | ||||||
| n | 1110 | 2089 | 4692 | 2662 | ||
| Model 1 | 0 (ref) | 0.9 (−0.7, 2.5) | 0.2 (−1.2, 1.7) | −1.1 (−2.7, 0.4) | −0.6 (−0.8, −0.3) | <0.001 |
| Model 2 | 0 (ref) | 0.1 (−1.5, 1.7) | −1.2 (−2.6, 0.3) | −2.2 (−3.7, −0.6) | −0.7 (−1.0, −0.4) | <0.001 |
| Caffeinated coffee | ||||||
| C-peptide | ||||||
| n | 2216 | 2877 | 3458 | 1279 | ||
| Model 1 | 0 (ref) | −1.2 (−4.3, 1.9) | −2.7 (−5.6, 0.2) | −3.7 (−7.3, 0.1) | −0.9 (−1.6, −0.1) | 0.03 |
| Model 2 | 0 (ref) | −1.3 (−3.9, 1.4) | −4.3 (−6.8, −1.7) | −8.0 (−11.2, −4.8) | −1.9 (−2.6, −1.2) | <0.001 |
| IGF-1 | ||||||
| n | 2320 | 2910 | 3528 | 1412 | ||
| Model 1 | 0 (ref) | −0.1 (−1.9, 1.6) | −0.1 (−1.8, 1.6) | −1.1 (−3.3, 1.0) | −0.2 (−0.7, 0.2) | 0.25 |
| Model 2 | 0 (ref) | 0.1 (−1.6, 1.8) | −0.02 (−1.7, 1.7) | −1.3 (−3.4, 0.9) | −0.3 (−0.8, 0.1) | 0.13 |
| IGFBP-1 | ||||||
| n | 1560 | 1968 | 2318 | 851 | ||
| Model 1 | 0 (ref) | −3.5 (−9.1, 2.4) | −4.1 (−9.5, 1.6) | 0.3 (−7.0, 8.1) | −0.1 (−1.6, 1.4) | 0.91 |
| Model 2 | 0 (ref) | −2.4 (−7.3, 2.7) | −0.6 (−5.5, 4.5) | 4.3 (−2.4, 11.5) | 0.8 (−0.6, 2.2) | 0.25 |
| IGFBP-3 | ||||||
| n | 2398 | 3030 | 3677 | 1448 | ||
| Model 1 | 0 (ref) | 0.4 (−0.7, 1.6) | −1.2 (−2.3, −0.1) | −3.3 (−4.7, −1.9) | −0.8 (−1.1, −0.6) | <0.001 |
| Model 2 | 0 (ref) | −0.1 (−1.3, 1.0) | −2.1 (−3.2, −1.0) | −3.7 (−5.1, −2.3) | −1.0 (−1.2, −0.7) | <0.001 |
Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), cumulative average levels of BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalent of task-h/wk), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), smoking (never; past: <30; past: ≥30; current: <30; or current: ≥30 pack-years), Alternative Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except coffee (tea, cola beverage, and chocolate), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. IGF, insulin-like growth factor; IGFBP, insulin-like growth factor binding protein; ref, reference.
TABLE 3.
Percentage differences (95% CIs) in concentrations of plasma biomarkers of the insulin/IGF system according to decaffeinated coffee consumption among participants from the Nurses’ Health Study and the Health Professionals Follow-Up Study1
| Categories of decaffeinated coffee consumption | ||||||
|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2 cups/d | ≥3 cups/d | Per cup increase | P-trend |
| C-peptide | ||||||
| n | 3421 | 3829 | 1258 | 1322 | ||
| Model 1 | 0 (ref) | −6.0 (−8.4, −3.6) | −6.4 (−9.8, −3.0) | −7.1 (−10.4, −3.7) | −1.9 (−2.9, −0.8) | <0.001 |
| Model 2 | 0 (ref) | −3.5 (−5.6, −1.3) | −5.9 (−8.8, −2.8) | −5.8 (−8.7, −2.8) | −1.9 (−2.8, −1.0) | <0.001 |
| IGF-1 | ||||||
| n | 3616 | 3948 | 1268 | 1338 | ||
| Model 1 | 0 (ref) | 2.9 (1.4, 4.4) | 2.5 (0.4, 4.6) | 2.8 (0.7, 4.9) | 0.7 (0.1, 1.3) | 0.03 |
| Model 2 | 0 (ref) | 3.1 (1.7, 4.6) | 2.5 (0.4, 4.6) | 2.8 (0.7, 4.9) | 0.7 (0.1, 1.3) | 0.01 |
| IGFBP-1 | ||||||
| n | 2339 | 2608 | 844 | 906 | ||
| Model 1 | 0 (ref) | 6.2 (1.0, 11.6) | 10.4 (2.9, 18.5) | 4.7 (−2.4, 12.2) | 1.5 (−0.6, 3.6) | 0.16 |
| Model 2 | 0 (ref) | 2.0 (−2.3, 6.4) | 7.7 (1.4, 14.5) | 1.3 (−4.6, 7.6) | 1.0 (−0.8, 2.9) | 0.26 |
| IGFBP-3 | ||||||
| n | 3745 | 4100 | 1320 | 1388 | ||
| Model 1 | 0 (ref) | 0.3 (−0.6, 1.3) | 0.5 (−0.9, 1.9) | 0.3 (−1.0, 1.7) | 0.03 (−0.4, 0.4) | 0.88 |
| Model 2 | 0 (ref) | 0.1 (−0.9, 1.0) | −0.3 (−1.6, 1.1) | −0.2 (−1.6, 1.1) | −0.2 (−0.6, 0.2) | 0.43 |
Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), cumulative average levels of BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalent of task-h/wk), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), smoking (never; past: <30; past: ≥30; current: <30; or current: ≥30 pack-years), Alternative Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except coffee (tea, cola beverage, and chocolate), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. IGF, insulin-like growth factor; IGFBP, insulin-like growth factor binding protein; ref, reference.
Tables 4 and 5 show the results for sex hormones. Compared with nondrinkers, the percentage differences (95% CIs) in biomarker concentrations for those drinking ≥4 cups/d of total coffee were as follows—estrone: −6.4% (−11.9%, −0.5%); total estradiol: −5.7% (−10.9%, −0.1%); free estradiol: −8.1% (−14.4%, −1.4%); SHBG: 5.0% (0.2%, 10.1%); and total testosterone: 7.3% (1.2%, 13.8%) for women and 5.3% (−1.4%, 12.5%) for men. The results were similar for caffeinated and decaffeinated coffee.
TABLE 4.
Percentage differences (95% CIs) in concentrations of plasma biomarkers of sex hormones according to total and caffeinated coffee consumption among participants from the Nurses’ Health Study and the Health Professionals Follow-Up Study1
| Categories of coffee consumption | ||||||
|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2–3 cups/d | ≥4 cups/d | Per cup increase | P-trend |
| Total coffee | ||||||
| Estrone | ||||||
| n | 317 | 692 | 1659 | 992 | ||
| Model 1 | 0 (ref) | 2.1 (−4.5, 9.2) | 0.5 (−5.5, 6.8) | −4.7 (−10.6, 1.6) | −1.1 (−2.2, −0.1) | 0.03 |
| Model 2 | 0 (ref) | 0.6 (−5.4, 7.0) | −0.9 (−6.4, 4.9) | −6.4 (−11.9, −0.5) | −1.4 (−2.3, −0.4) | 0.01 |
| Total estradiol | ||||||
| n | 388 | 834 | 1948 | 1159 | ||
| Model 1 | 0 (ref) | 4.1 (−2.4, 10.9) | −1.9 (−7.4, 4.0) | −5.7 (−11.3, 0.3) | −1.7 (−2.7, −0.7) | <0.001 |
| Model 2 | 0 (ref) | 2.1 (−3.7, 8.1) | −2.1 (−7.2, 3.2) | −5.7 (−10.9, −0.1) | −1.5 (−2.5, −0.6) | 0.001 |
| Free estradiol | ||||||
| n | 324 | 661 | 1516 | 888 | ||
| Model 1 | 0 (ref) | 6.1 (−2.0, 14.8) | −1.6 (−8.4, 5.7) | −8.5 (−15.2, −1.3) | −2.8 (−4.0, −1.5) | <0.001 |
| Model 2 | 0 (ref) | 3.3 (−3.7, 10.9) | −2.6 (−8.8, 3.9) | −8.1 (−14.4, −1.4) | −2.5 (−3.7, −1.3) | <0.001 |
| SHBG | ||||||
| n | 554 | 1138 | 2652 | 1571 | ||
| Model 1 | 0 (ref) | −0.6 (−6.0, 5.0) | 0.7 (−4.2, 5.9) | 4.4 (−1.0, 10.0) | 1.1 (0.3, 2.0) | 0.01 |
| Model 2 | 0 (ref) | −0.6 (−5.2, 4.1) | 0.8 (−3.5, 5.3) | 5.0 (0.2, 10.1) | 1.5 (0.7, 2.3) | <0.001 |
| Total testosterone (women) | ||||||
| n | 380 | 924 | 2433 | 1447 | ||
| Model 1 | 0 (ref) | 5.9 (−0.3, 12.5) | 7.3 (1.5, 13.3) | 9.9 (3.8, 16.4) | 1.6 (0.7, 2.5) | <0.001 |
| Model 2 | 0 (ref) | 5.9 (−0.3, 12.4) | 6.5 (0.8, 12.6) | 7.3 (1.2, 13.8) | 1.0 (0.1, 1.9) | 0.04 |
| Free testosterone (women) | ||||||
| n | 271 | 679 | 1802 | 1054 | ||
| Model 1 | 0 (ref) | 6.9 (−2.6, 17.3) | 7.6 (−1.1, 17.1) | 8.9 (−0.3, 18.9) | 1.2 (−0.1, 2.5) | 0.08 |
| Model 2 | 0 (ref) | 8.2 (−0.5, 17.6) | 8.0 (0.01, 16.6) | 7.3 (−1.1, 16.5) | 0.4 (−0.9, 1.7) | 0.53 |
| Total testosterone (men) | ||||||
| n | 205 | 355 | 575 | 344 | ||
| Model 1 | 0 (ref) | −1.0 (−7.2, 5.6) | −0.4 (−6.2, 5.7) | 3.0 (−3.4, 9.9) | 1.1 (−0.1, 2.3) | 0.07 |
| Model 2 | 0 (ref) | 0.04 (−6.0, 6.5) | 1.9 (−4.0, 8.2) | 5.3 (−1.4, 12.5) | 1.6 (0.4, 2.8) | 0.01 |
| Free testosterone (men) | ||||||
| n | 202 | 348 | 568 | 335 | ||
| Model 1 | 0 (ref) | −0.2 (−6.3, 6.4) | 2.3 (−3.6, 8.5) | 2.4 (−4.0, 9.2) | 0.9 (−0.3, 2.0) | 0.15 |
| Model 2 | 0 (ref) | 0.2 (−6.0, 6.8) | 3.7 (−2.4, 10.2) | 3.6 (−3.1, 10.7) | 1.2 (−0.1, 2.4) | 0.06 |
| Caffeinated coffee | ||||||
| Estrone | ||||||
| n | 772 | 1053 | 1298 | 537 | ||
| Model 1 | 0 (ref) | 1.5 (−3.1, 6.4) | −0.7 (−5.1, 3.9) | −3.8 (−9.1, 1.8) | −1.0 (−2.1, 0.1) | 0.08 |
| Model 2 | 0 (ref) | 1.0 (−3.3, 5.4) | −1.0 (−5.1, 3.3) | −5.9 (−10.8, −0.7) | −1.3 (−2.3, −0.2) | 0.02 |
| Total estradiol | ||||||
| n | 932 | 1243 | 1528 | 626 | ||
| Model 1 | 0 (ref) | 0.8 (−3.7, 5.4) | −1.2 (−5.4, 3.1) | −3.7 (−8.8, 1.7) | −1.0 (−2.0, 0.1) | 0.08 |
| Model 2 | 0 (ref) | 0.5 (−3.4, 4.7) | −0.8 (−4.7, 3.3) | −4.1 (−8.9, 0.9) | −0.9 (−1.9, 0.1) | 0.08 |
| Free estradiol | ||||||
| n | 737 | 998 | 1176 | 478 | ||
| Model 1 | 0 (ref) | 1.7 (−3.9, 7.6) | −1.4 (−6.7, 4.1) | −8.5 (−14.6, −2.0) | −2.2 (−3.5, −0.8) | 0.002 |
| Model 2 | 0 (ref) | 1.4 (−3.6, 6.7) | −1.6 (−6.4, 3.4) | −8.2 (−13.9, −2.1) | −2.1 (−3.4, −0.8) | 0.002 |
| SHBG | ||||||
| n | 1318 | 1693 | 2059 | 845 | ||
| Model 1 | 0 (ref) | −0.7 (−4.5, 3.3) | −0.04 (−3.7, 3.8) | 3.5 (−1.3, 8.5) | 0.7 (−0.2, 1.7) | 0.14 |
| Model 2 | 0 (ref) | −0.1 (−3.4, 3.3) | 2.4 (−0.9, 5.8) | 7.0 (2.5, 11.6) | 1.6 (0.7, 2.4) | <0.001 |
| Total testosterone (women) | ||||||
| n | 1118 | 1427 | 1877 | 762 | ||
| Model 1 | 0 (ref) | 2.9 (−1.1, 7.1) | 5.0 (1.2, 9.1) | 9.9 (4.8, 15.3) | 2.3 (1.3, 3.3) | <0.001 |
| Model 2 | 0 (ref) | 3.0 (−1.0, 7.1) | 4.3 (0.4, 8.3) | 6.2 (1.1, 11.5) | 1.5 (0.5, 2.6) | 0.003 |
| Free testosterone (women) | ||||||
| n | 808 | 1059 | 1376 | 563 | ||
| Model 1 | 0 (ref) | 6.4 (0.2, 13.0) | 7.0 (1.0, 13.4) | 9.3 (1.7, 17.5) | 1.8 (0.4, 3.3) | 0.01 |
| Model 2 | 0 (ref) | 6.1 (0.5, 12.0) | 3.6 (−1.8, 9.3) | 3.1 (−3.6, 10.4) | 0.4 (−1.0, 1.8) | 0.59 |
| Total testosterone (men) | ||||||
| n | 358 | 471 | 463 | 187 | ||
| Model 1 | 0 (ref) | 1.0 (−4.0, 6.4) | 0.8 (−4.2, 6.2) | 5.1 (−1.7, 12.3) | 0.9 (−0.4, 2.2) | 0.18 |
| Model 2 | 0 (ref) | 1.4 (−3.5, 6.5) | 2.4 (−2.6, 7.8) | 7.4 (0.5, 14.8) | 1.3 (−0.04, 2.7) | 0.06 |
| Free testosterone (men) | ||||||
| n | 354 | 463 | 454 | 182 | ||
| Model 1 | 0 (ref) | 0.7 (−4.3, 6.0) | 2.0 (−3.1, 7.3) | 3.6 (−3.0, 10.6) | 0.7 (−0.6, 2.0) | 0.30 |
| Model 2 | 0 (ref) | 0.7 (−4.2, 5.9) | 3.2 (−2.0, 8.7) | 4.7 (−2.1, 12.0) | 1.0 (−0.4, 2.3) | 0.17 |
Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), cumulative average levels of BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalent of task-h/wk), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), smoking (never; past: <30; past: ≥30; current: <30; or current: ≥30 pack-years), Alternative Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except coffee (tea, cola beverage, and chocolate), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. ref, reference; SHBG, sex hormone–binding globulin.
TABLE 5.
Percentage differences (95% CIs) in concentrations of plasma biomarkers of sex hormones according to decaffeinated coffee consumption among participants from the Nurses’ Health Study and the Health Professionals Follow-Up Study1
| Categories of decaffeinated coffee consumption | ||||||
|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2 cups/d | ≥3 cups/d | Per cup increase | P-trend |
| Estrone | ||||||
| n | 1267 | 1414 | 468 | 511 | ||
| Model 1 | 0 (ref) | 1.4 (−2.5, 5.4) | 2.3 (−3.1, 8.0) | −5.4 (−10.3, −0.2) | −1.4 (−2.9, 0.2) | 0.08 |
| Model 2 | 0 (ref) | 1.0 (−2.6, 4.6) | 2.2 (−2.8, 7.4) | −5.5 (−10.1, −0.8) | −1.5 (−2.9, −0.1) | 0.04 |
| Total estradiol | ||||||
| n | 1499 | 1694 | 538 | 598 | ||
| Model 1 | 0 (ref) | 0.7 (−2.9, 4.5) | −5.1 (−9.9, 0.1) | −9.4 (−14.0, −4.6) | −3.3 (−4.8, −1.9) | <0.001 |
| Model 2 | 0 (ref) | 1.1 (−2.2, 4.6) | −3.6 (−8.1, 1.1) | −7.8 (−12.0, −3.4) | −2.8 (−4.1, −1.4) | <0.001 |
| Free estradiol | ||||||
| n | 1175 | 1339 | 413 | 462 | ||
| Model 1 | 0 (ref) | −2.1 (−6.6, 2.6) | −7.3 (−13.3, −0.8) | −12.5 (−18.1, −6.6) | −4.2 (−6.0, −2.3) | <0.001 |
| Model 2 | 0 (ref) | −1.1 (−5.2, 3.2) | −4.9 (−10.4, 1.0) | −11.0 (−16.1, −5.5) | −3.4 (−5.1, −1.7) | <0.001 |
| SHBG | ||||||
| n | 2009 | 2348 | 716 | 842 | ||
| Model 1 | 0 (ref) | 5.7 (2.3, 9.2) | 5.9 (1.1, 10.9) | 8.4 (3.7, 13.4) | 2.2 (0.9, 3.5) | 0.001 |
| Model 2 | 0 (ref) | 2.1 (−0.7, 5.0) | 3.3 (−0.7, 7.6) | 5.8 (1.8, 10.0) | 1.6 (0.4, 2.7) | 0.01 |
| Total testosterone (women) | ||||||
| n | 1659 | 2030 | 677 | 818 | ||
| Model 1 | 0 (ref) | −0.9 (−4.1, 2.4) | 1.4 (−3.1, 6.2) | −0.8 (−5.0, 3.7) | 0.1 (−1.1, 1.4) | 0.84 |
| Model 2 | 0 (ref) | 0.5 (−2.8, 3.9) | 2.3 (−2.3, 7.1) | −0.7 (−5.0, 3.7) | −0.1 (−1.4, 1.2) | 0.86 |
| Free testosterone (women) | ||||||
| n | 1208 | 1513 | 492 | 593 | ||
| Model 1 | 0 (ref) | −5.1 (−9.8, −0.2) | −0.04 (−6.8, 7.2) | −4.7 (−10.9, 2.0) | −0.6 (−2.5, 1.3) | 0.55 |
| Model 2 | 0 (ref) | 0.9 (−3.6, 5.7) | 3.7 (−2.7, 10.5) | −0.4 (−6.4, 5.9) | 0.1 (−1.6, 1.9) | 0.90 |
| Total testosterone (men) | ||||||
| n | 572 | 604 | 147 | 156 | ||
| Model 1 | 0 (ref) | 1.3 (−2.9, 5.7) | 6.0 (−0.9, 13.5) | 5.1 (−1.6, 12.4) | 1.4 (−0.5, 3.4) | 0.15 |
| Model 2 | 0 (ref) | 2.5 (−1.7, 6.8) | 6.0 (−0.6, 13.1) | 8.4 (1.6, 15.7) | 2.1 (0.2, 4.0) | 0.03 |
| Free testosterone (men) | ||||||
| n | 561 | 595 | 145 | 152 | ||
| Model 1 | 0 (ref) | 1.1 (−3.1, 5.5) | 5.2 (−1.6, 12.5) | 5.9 (−0.9, 13.2) | 1.2 (−0.7, 3.2) | 0.21 |
| Model 2 | 0 (ref) | 1.7 (−2.5, 6.1) | 4.7 (−1.9, 11.8) | 7.9 (1.0, 15.2) | 1.6 (−0.3, 3.6) | 0.10 |
Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), cumulative average levels of BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalent of task-h/wk), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), smoking (never; past: <30; past: ≥30; current: <30; or current: ≥30 pack-years), Alternative Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except coffee (tea, cola beverage, and chocolate), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. ref, reference; SHBG, sex hormone–binding globulin.
Tables 6 and 7 show the results for adipokines and inflammatory markers. Total, caffeinated, and decaffeinated coffee all showed positive associations with concentrations of total and HMW adiponectin and inverse associations with concentrations of CRP, IL-6, and sTNFR-2. Total and caffeinated coffee also showed inverse associations with leptin. The percentage differences (95% CIs) comparing participants who drank ≥4 cups of total coffee/d with nondrinkers were as follows— total adiponectin: 9.3% (6.1%, 12.5%); HMW adiponectin: 17.2% (8.2%, 26.8%); leptin: −6.4% (−11.3%, −1.2%); CRP: −16.6% (−21.6%, −11.4%); IL-6: −8.1% (−12.4%, −3.7%); and sTNFR-2: −5.8% (−7.7%, −4.0%).
TABLE 6.
Percentage differences (95% CIs) in concentrations of plasma biomarkers of adipokines and inflammatory markers according to total and caffeinated coffee consumption among participants from the Nurses’ Health Study and the Health Professionals Follow-Up Study1
| Categories of coffee consumption | ||||||
|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2–3 cups/d | ≥4 cups/d | Per cup increase | P-trend |
| Total coffee | ||||||
| Total adiponectin | ||||||
| n | 1195 | 2406 | 5790 | 3404 | ||
| Model 1 | 0 (ref) | 1.5 (−1.6, 4.7) | 5.8 (2.8, 8.8) | 11.5 (8.3, 14.9) | 2.2 (1.7, 2.7) | <0.001 |
| Model 2 | 0 (ref) | 0.3 (−2.6, 3.3) | 3.0 (0.3, 5.9) | 9.3 (6.1, 12.5) | 2.0 (1.6, 2.5) | <0.001 |
| HMW adiponectin | ||||||
| n | 359 | 777 | 1856 | 1060 | ||
| Model 1 | 0 (ref) | −0.5 (−8.6, 8.4) | 12.6 (4.3, 21.7) | 24.5 (14.7, 35.0) | 5.2 (3.9, 6.5) | <0.001 |
| Model 2 | 0 (ref) | −1.7 (−9.2, 6.4) | 7.4 (−0.2, 15.5) | 17.2 (8.2, 26.8) | 4.1 (2.8, 5.4) | <0.001 |
| Leptin | ||||||
| n | 662 | 1261 | 2682 | 1480 | ||
| Model 1 | 0 (ref) | −4.5 (−10.9, 2.4) | −6.2 (−11.9, −0.1) | −9.2 (−15.2, −2.8) | −1.9 (−3.0, −0.7) | 0.001 |
| Model 2 | 0 (ref) | −4.1 (−9.1, 1.1) | −3.1 (−7.8, 1.7) | −6.4 (−11.3, −1.2) | −1.1 (−2.0, −0.2) | 0.01 |
| CRP | ||||||
| n | 1428 | 2803 | 6358 | 3850 | ||
| Model 1 | 0 (ref) | −1.9 (−8.5, 5.0) | −9.0 (−14.5, −3.1) | −12.8 (−18.4, −6.9) | −2.9 (−3.9, −1.8) | <0.001 |
| Model 2 | 0 (ref) | −1.9 (−7.8, 4.4) | −8.8 (−13.9, −3.5) | −16.6 (−21.6, −11.4) | −4.1 (−5.1, −3.1) | <0.001 |
| IL-6 | ||||||
| n | 869 | 1709 | 4116 | 2404 | ||
| Model 1 | 0 (ref) | −3.1 (−7.9, 2.0) | −7.0 (−11.2, −2.7) | −6.5 (−10.9, −1.9) | −0.9 (−1.7, −0.2) | 0.02 |
| Model 2 | 0 (ref) | −2.4 (−7.0, 2.5) | −6.3 (−10.4, −2.1) | −8.1 (−12.4, −3.7) | −1.7 (−2.4, −0.9) | <0.001 |
| sTNFR-2 | ||||||
| n | 990 | 1952 | 4670 | 2737 | ||
| Model 1 | 0 (ref) | −4.8 (−6.8, −2.9) | −7.5 (−9.2, −5.7) | −7.4 (−9.2, −5.6) | −1.1 (−1.4, −0.8) | <0.001 |
| Model 2 | 0 (ref) | −3.6 (−5.5, −1.7) | −5.4 (−7.1, −3.7) | −5.8 (−7.7, −4.0) | −1.0 (−1.3, −0.7) | <0.001 |
| Caffeinated coffee | ||||||
| Total adiponectin | ||||||
| n | 2822 | 3640 | 4507 | 1826 | ||
| Model 1 | 0 (ref) | 1.0 (−1.2, 3.2) | 4.6 (2.4, 6.8) | 9.0 (6.2, 12.0) | 1.9 (1.4, 2.5) | <0.001 |
| Model 2 | 0 (ref) | 0.2 (−1.9, 2.3) | 3.0 (1.0, 5.2) | 8.8 (6.0, 11.7) | 1.9 (1.3, 2.4) | <0.001 |
| HMW adiponectin | ||||||
| n | 891 | 1138 | 1447 | 576 | ||
| Model 1 | 0 (ref) | −3.4 (−9.0, 2.5) | 7.6 (1.6, 13.9) | 18.0 (9.8, 26.7) | 4.4 (2.9, 5.9) | <0.001 |
| Model 2 | 0 (ref) | −4.1 (−9.3, 1.3) | 3.4 (−2.0, 9.2) | 13.7 (6.0, 22.0) | 3.4 (1.9, 4.9) | <0.001 |
| Leptin | ||||||
| n | 1397 | 1809 | 2110 | 769 | ||
| Model 1 | 0 (ref) | −1.9 (−6.9, 3.3) | −5.4 (−10.1, −0.6) | −7.4 (−13.2, −1.1) | −2.0 (−3.2, −0.7) | 0.003 |
| Model 2 | 0 (ref) | −3.1 (−6.8, 0.8) | −4.9 (−8.5, −1.1) | −5.9 (−10.6, −1.0) | −1.4 (−2.4, −0.4) | 0.01 |
| CRP | ||||||
| n | 3232 | 4120 | 5005 | 2082 | ||
| Model 1 | 0 (ref) | −2.9 (−7.6, 2.1) | −9.0 (−13.2, −4.5) | −8.4 (−13.7, −2.7) | −2.1 (−3.2, −0.9) | <0.001 |
| Model 2 | 0 (ref) | −3.4 (−7.7, 1.0) | −10.7 (−14.6, −6.7) | −15.9 (−20.5, −11.1) | −3.8 (−4.8, −2.7) | <0.001 |
| IL-6 | ||||||
| n | 2037 | 2539 | 3220 | 1302 | ||
| Model 1 | 0 (ref) | −1.6 (−5.1, 2.0) | −3.7 (−7.0, −0.3) | 0.9 (−3.4, 5.4) | −0.2 (−1.0, 0.7) | 0.70 |
| Model 2 | 0 (ref) | −1.7 (−5.0, 1.8) | −4.1 (−7.3, −0.9) | −4.3 (−8.3, −0.04) | −1.3 (−2.2, −0.4) | 0.003 |
| sTNFR-2 | ||||||
| n | 2333 | 2916 | 3631 | 1469 | ||
| Model 1 | 0 (ref) | −2.3 (−3.7, −0.8) | −5.0 (−6.3, −3.6) | −4.1 (−5.7, −2.3) | −1.1 (−1.4, −0.7) | <0.001 |
| Model 2 | 0 (ref) | −1.6 (−3.0, −0.2) | −3.6 (−4.9, −2.3) | −3.8 (−5.5, −2.1) | −1.0 (−1.4, −0.6) | <0.001 |
Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), cumulative average levels of BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalent of task-h/wk), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), smoking (never; past: <30; past: ≥30; current: <30; or current: ≥30 pack-years), Alternative Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except coffee (tea, cola beverage, and chocolate), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. CRP, C-reactive protein; HMW, high-molecular-weight; ref, reference; sTNFR-2, soluble TNF receptor 2.
TABLE 7.
Percentage differences (95% CIs) in concentrations of plasma biomarkers of adipokines and inflammatory markers according to decaffeinated coffee consumption among participants from the Nurses’ Health Study and the Health Professionals Follow-Up Study1
| Categories of decaffeinated coffee consumption | ||||||
|---|---|---|---|---|---|---|
| Variable | Nondrinker | ≤1 cup/d | 2 cups/d | ≥3 cups/d | Per cup increase | P-trend |
| Total adiponectin | ||||||
| n | 4443 | 4931 | 1589 | 1832 | ||
| Model 1 | 0 (ref) | 2.7 (0.8, 4.6) | 9.2 (6.4, 12.0) | 8.4 (5.7, 11.1) | 2.8 (2.0, 3.5) | <0.001 |
| Model 2 | 0 (ref) | 0.8 (−1.0, 2.5) | 6.5 (3.9, 9.1) | 6.4 (3.9, 8.9) | 2.3 (1.6, 3.1) | <0.001 |
| HMW adiponectin | ||||||
| n | 1434 | 1548 | 501 | 569 | ||
| Model 1 | 0 (ref) | 6.7 (1.6, 12.1) | 20.1 (12.0, 28.7) | 21.9 (13.9, 30.4) | 6.5 (4.4, 8.6) | <0.001 |
| Model 2 | 0 (ref) | 3.4 (−1.2, 8.3) | 13.2 (6.1, 20.9) | 16.3 (9.1, 24.0) | 5.1 (3.1, 7.1) | <0.001 |
| Leptin | ||||||
| n | 2205 | 2360 | 733 | 787 | ||
| Model 1 | 0 (ref) | −4.8 (−8.8, −0.6) | −6.8 (−12.4, −0.8) | −6.2 (−11.7, −0.2) | −1.7 (−3.4, 0.1) | 0.06 |
| Model 2 | 0 (ref) | −0.8 (−4.0, 2.6) | −1.6 (−6.1, 3.2) | −1.4 (−5.9, 3.4) | −0.6 (−1.9, 0.8) | 0.39 |
| CRP | ||||||
| n | 5134 | 5542 | 1762 | 2001 | ||
| Model 1 | 0 (ref) | −7.5 (−11.3, −3.6) | −11.4 (−16.5, −6.0) | −15.7 (−20.4, −10.8) | −4.5 (−6.1, −2.9) | <0.001 |
| Model 2 | 0 (ref) | −3.0 (−6.6, 0.7) | −8.3 (−13.0, −3.3) | −14.2 (−18.5, −9.7) | −4.8 (−6.2, −3.3) | <0.001 |
| IL-6 | ||||||
| n | 3188 | 3475 | 1165 | 1270 | ||
| Model 1 | 0 (ref) | −7.4 (−10.1, −4.6) | −9.6 (−13.3, −5.7) | −9.9 (−13.5, −6.1) | −2.4 (−3.6, −1.3) | <0.001 |
| Model 2 | 0 (ref) | −4.6 (−7.3, −1.8) | −6.7 (−10.3, −2.8) | −7.8 (−11.4, −4.1) | −2.1 (−3.2, −0.9) | <0.001 |
| sTNFR-2 | ||||||
| n | 3598 | 3969 | 1309 | 1473 | ||
| Model 1 | 0 (ref) | −3.0 (−4.1, −1.8) | −2.2 (−3.9, −0.5) | −4.9 (−6.5, −3.3) | −1.1 (−1.6, −0.7) | <0.001 |
| Model 2 | 0 (ref) | −2.1 (−3.2, −0.9) | −1.2 (−2.8, 0.5) | −3.8 (−5.4, −2.2) | −0.9 (−1.3, −0.4) | <0.001 |
Model 1 was adjusted for sex and age at blood draw (linear and quadratic terms). Model 2 was additionally adjusted for race (white, nonwhite), fasting status (yes or no), cumulative average levels of BMI (in kg/m2; <23.0, 23.0–24.9, 25.0–27.4, 27.5–29.9, or ≥30.0), physical activity (<3.0, 3.0–8.9, 9.0–17.9, 18.0–26.9, or ≥27.0 metabolic equivalent of task-h/wk), alcohol consumption (0, 0.1–4.9, 5.0–9.9, 10.0–14.9, or ≥15.0 g/d), smoking (never; past: <30; past: ≥30; current: <30; or current: ≥30 pack-years), Alternative Healthy Eating Index dietary score (quintiles), total energy intake (quintiles), caffeine intake (quintiles) calculated from dietary items except coffee (tea, cola beverage, and chocolate), regular multivitamin use (yes or no), regular aspirin or nonsteroidal anti-inflammatory drug use (yes or no), hypertension (yes or no), hypercholesterolemia (yes or no), and for women, menopausal status (premenopausal, postmenopausal, or unknown) and menopausal hormone therapy (never, past, or current use). For the analysis of caffeinated and decaffeinated coffee, mutual adjustment was performed. CRP, C-reactive protein; HMW, high-molecular-weight; ref, reference; sTNFR-2, soluble TNF receptor 2.
Stratified analyses by sex, age, smoking, alcohol consumption, physical activity, BMI, menopausal status, and hormonal use are shown in Supplemental Figure 2. The associations of total coffee consumption with biomarker concentrations were generally consistent across subgroups, with a few exceptions. A positive association of coffee intake with IGFBP-1 was present only in men (P-interaction = 0.01) and the inverse association with CRP was stronger in women than in men (P-interaction <0.001). In addition, an inverse association with IGF was observed among participants with BMI <23 and the positive association with SHBG remained statistically significant in postmenopausal women who were not currently using hormones.
Sensitivity analysis showed that the aforementioned associations were robust among participants who provided fasting blood (Supplemental Table 5), were selected as controls in previous case-control studies (Supplemental Table 6), and were free of hypertension or hypercholesterolemia at blood draw (Supplemental Table 7).
Discussion
In this large study in women and men, we found that coffee consumption was associated with lower concentrations of C-peptide, IGFBP-3, estrogen, leptin, and inflammatory biomarkers and with higher concentrations of SHBG, total testosterone, and adiponectin. The results were largely similar for caffeinated and decaffeinated coffee. Comprehensive stratified and sensitivity analyses supported the robustness of the observed associations. Therefore, our findings suggest that coffee might exert beneficial effects through modulation of metabolic and inflammatory pathways. Although the association of high coffee consumption with some biomarkers appeared relatively small (e.g., for IGFBP-3 and SHBG), given the potential pleiotropic benefit of coffee on multiple pathways the collective effect of coffee on individual pathways may contribute to a clinically significant impact on disease risk (2–4).
Insulin/IGF system
The insulin/IGF system is a major regulator of energy balance and cell growth. In the circulation, high concentrations of C-peptide, a marker of insulin secretion, and IGF-1 have been associated with an increased risk of breast, colorectal, and prostate cancers (10).
An interventional trial in 20 healthy nonobese participants showed a reduction in C-peptide concentrations at 60 min after caffeinated coffee ingestion (40). In our previous study in the NHS cohort, both caffeinated and decaffeinated coffee were associated with lower C-peptide, particularly in obese and overweight women (41). The current study with a larger sample size confirmed the associations in both sexes, supporting the potential benefit of coffee consumption on lowering insulin secretion. Moreover, our findings of a similar association for caffeinated and decaffeinated coffee imply that bioactive ingredients of coffee, other than caffeine, might mediate the benefit. For example, chlorogenic acid, a strong antioxidant in both caffeinated and decaffeinated coffee, has been shown to improve insulin sensitivity in patients with T2D and reduce insulin concentrations in rats, possibly through decreased oxidative stress (42).
Previous studies on coffee consumption and IGF-1 concentrations reported either an inverse (13) or null (14, 43) association. These studies are limited by their small sample size or lack of separate assessment of different coffee types. In the current study, we found that caffeinated coffee tended to be associated with lower concentrations of IGF-1, whereas decaffeinated coffee was linked to higher concentrations of IGF-1. Experimental evidence supports that circulating IGF-1 may be reduced by caffeine exposure in rats (44). In addition, the circulating half-life of IGF-1 can be greatly prolonged by IGFBP-3, which provides most of the IGF-binding capacity (45). These data are in line with our findings, suggesting that, unlike decaffeinated coffee, caffeinated coffee might reduce IGFBP-3, thereby lowering bioavailable IGF-1.
Sex hormones
Circulating estrone and estradiol have been linked to a higher risk of breast, endometrial, and ovarian cancers in women (8) and prostate cancer in men (46). In contrast, testosterone has shown antiproliferative and proapoptotic effects on breast cancer (47). SHBG is a carrier protein for estrogen and androgen, acting as a modulator of hormonal activity. Lower concentrations of SHBG have been associated with increased risk of metabolic diseases and hormone-related cancers (48).
Two small studies reported that caffeine intake was not associated with estrone in women (49, 50). In the current study, we found that total coffee consumption was associated with lower concentrations of estrone. Caffeinated and decaffeinated coffee showed similar results. For estradiol, several studies reported a null association between caffeine intake and estradiol in women (51, 52) or men (16, 53), whereas other studies showed an inverse association (15, 49, 54). Consistent with the latter studies, we found that coffee consumption was associated with lower concentrations of total and free estradiol, providing a possible explanation for the strong association of coffee consumption with lower risk of endometrial cancer (6). Animal models indicated that caffeine might interfere with estrogen metabolism via inhibition of aromatase, the key enzyme mediating the conversion of androgen to estrogen (55). The evidence supports our observed association between caffeinated coffee and higher concentrations of total testosterone. Consistently, a randomized controlled trial found that male total testosterone concentrations were significantly increased after 4 wk consumption of caffeinated coffee (22). Experimental studies also suggested that caffeine administration in male animals could increase plasma testosterone through sympathetic stimulation of Leydig cells in the testis (56, 57). For women, our findings of a positive association between caffeinated coffee and total testosterone concentrations contrasted with the null associations reported in previous studies (54, 58). Although the exact reasons for this discrepancy remain unclear, previous studies had relatively small sample sizes and thus may have been underpowered to identify a modest association between caffeinated coffee and testosterone concentrations.
With regard to SHBG, several studies have reported a positive association with coffee or caffeine consumption in women (51, 54, 58), whereas the results were less consistent in men (16, 59, 60). In the current study, caffeinated coffee was associated with higher concentrations of SHBG in both sexes. Because caffeine is primarily metabolized in the liver where SHBG is primarily synthesized and metabolized, it is possible that caffeine consumption might influence SHBG metabolism and elevate SHBG concentrations (58). Our findings also suggest that although coffee intake may increase total testosterone, most of the testosterone is likely to bind to increased SHBG, leading to a nonsignificant association between coffee intake and free testosterone.
Adipokines and inflammatory biomarkers
Adiponectin and leptin are adipokines primarily secreted by adipose tissue (61). Evidence has suggested that HMW adiponectin may be the most bioactive isoform of circulating adiponectin, which is inversely associated with insulin resistance, obesity, T2D, and CVD (62). In contrast, chronically elevated leptin has been linked to obesity and cancer risk (10).
A positive association between caffeinated coffee and total adiponectin was reported in a previous study in the NHS cohort (17). The current study has extended these findings by also including men and reporting a particularly strong association for HMW adiponectin in both women and men. Our findings are consistent with other reports of a positive association between coffee consumption and total adiponectin in men (63–65) and women (66), and with the observation of an adiponectin-elevating effect of a 3-mo intervention of coffee drinking (24).
A few studies have investigated the relation between coffee consumption and leptin concentrations but reported null (20, 67) or inverse associations (64, 68–70). The current study found that total and caffeinated coffee consumption was associated with lower leptin concentrations. These findings are in line with the report of a clinical trial that high caffeine consumption decreased leptin concentrations, possibly through elevated thermogenesis and fat oxidation (71). Moreover, animal experiments indicated that caffeine and chlorogenic acid in combination might affect hepatic lipid metabolism and lower leptin concentrations (72).
The anti-inflammatory effect of coffee consumption has been widely studied. Consistent with most of the cross-sectional studies (18, 20, 21, 68), we observed inverse associations of coffee consumption with plasma CRP, IL-6, and sTNFR-2 concentrations. Furthermore, generally similar associations for caffeinated and decaffeinated coffee imply that bioactive constituents other than caffeine might have anti-inflammatory effects (20). Indeed, chlorogenic acids may reduce the production of inflammatory mediators by inhibiting protein tyrosine phosphatase 1B, lowering the expression of proinflammatory cytokine genes, and modulating nuclear factor-κB activation (6).
Strengths and limitations
The current study has several strengths, including the large sample size, inclusion of both women and men, repeated assessment of caffeinated and decaffeinated coffee consumption before blood draw, and comprehensive assessment of various biomarkers implicated in the development of chronic diseases. In addition, we used a standardized approach to account for potential batch effects and collected detailed data on covariates that allow for robust confounding control and assessment of potential effect modification.
Some limitations need to be considered as well. First, our study design was cross-sectional, thus limiting the ability to make causal inferences. Prospective studies are warranted to investigate whether coffee consumption is associated with changes in biomarker concentrations over time. Second, our study participants were predominantly white health professionals, which limited the generalizability of the findings; however, the homogeneity of the study population minimized the likelihood of residual confounding. Third, the assessed biomarkers are interrelated in a complex manner, which makes it difficult to disentangle their independent associations with coffee intake. However, by using the percentage of change in biomarker concentrations per 1-cup coffee intake as the standard outcome measure, we found that, among all the studied biomarkers, CRP, HMW adiponectin, and estrogen appeared to be most strongly associated with coffee, whereas a relatively modest association was found for C-peptide. Taken together with the central role of inflammation in metabolic disturbances (73) and the known effect of sex hormones on inflammation (74), it is possible that the beneficial associations observed in the current study may be predominantly driven by reduction in inflammation, with an additional effect by modulation of sex hormone metabolism. Further studies that account for the complex biomarker interrelations are needed to better understand the predominant pathways underlying the benefit of coffee intake.
In conclusion, our study indicates that coffee consumption is associated with a favorable profile of plasma biomarkers of metabolic and inflammatory pathways. Future prospective and interventional studies are warranted to confirm our findings.
Supplementary Material
ACKNOWLEDGEMENTS
The authors’ responsibilities were as follows—MS and ELG: conceived and designed the study; DH: performed the statistical analysis, drafted the manuscript, and is the guarantor; DH and MS: 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; ASK, WM, and YH: acquired the data; FKT, HN, ZH, HS, LAM, and ATC: critically revised the manuscript for important intellectual content; and all authors: read and approved the final manuscript. None of the authors reported a conflict of interest related to the study.
Notes
Supported by NIH grants K99 CA207736 (to FKT), K99 CA215314 (to MS), R00 CA215314, UM1 CA186107, R01 CA49449, and UM1 CA167552, American Cancer Society grant MRSG-17-220-01–NEC (to MS), and National Natural Science Foundation of China grant 81502873 (to DH). The funders had no role in design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; or the decision to submit the manuscript for publication.
Supplemental Figures 1–2, Supplemental Tables 1–7, and Supplemental Methods 1–3 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/ajcn/.
Abbreviations used: CRP, C-reactive protein; CVD, cardiovascular disease; FFQ, food-frequency questionnaire; HMW, high-molecular-weight; HPFS, Health Professionals Follow-Up Study; IGF, insulin-like growth factor; IGF-1, insulin-like growth factor 1; IGFBP, insulin-like growth factor binding protein; NHS, Nurses’ Health Study; SHBG, sex hormone–binding globulin; sTNFR-2, soluble TNF receptor 2; T2D, type 2 diabetes.
References
- 1. National Coffee Association. National coffee drinking trends. New York: National Coffee Association; 2017. [Google Scholar]
- 2. Ding M, Satija A, Bhupathiraju SN, Hu Y, Sun Q, Han J, Lopez-Garcia E, Willett W, van Dam RM, Hu FB. Association of coffee consumption with total and cause-specific mortality in 3 large prospective cohorts. Circulation. 2015;132(24):2305–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Gunter MJ, Murphy N, Cross AJ, Dossus L, Dartois L, Fagherazzi G, Kaaks R, Kuhn T, Boeing H, Aleksandrova K et al.. Coffee drinking and mortality in 10 European countries: a multinational cohort study. Ann Intern Med. 2017;167(4):236–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Park SY, Freedman ND, Haiman CA, Le Marchand L, Wilkens LR, Setiawan VW. Association of coffee consumption with total and cause-specific mortality among nonwhite populations. Ann Intern Med. 2017;167(4):228–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Poole R, Kennedy OJ, Roderick P, Fallowfield JA, Hayes PC, Parkes J. Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes. BMJ. 2017;359:j5024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Grosso G, Godos J, Galvano F, Giovannucci EL. Coffee, caffeine, and health outcomes: an umbrella review. Annu Rev Nutr. 2017;37:131–56. [DOI] [PubMed] [Google Scholar]
- 7. Rajpathak SN, He M, Sun Q, Kaplan RC, Muzumdar R, Rohan TE, Gunter MJ, Pollak M, Kim M, Pessin JE et al.. Insulin-like growth factor axis and risk of type 2 diabetes in women. Diabetes. 2012;61(9):2248–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Chuffa LG, Lupi-Junior LA, Costa AB, Amorim JP, Seiva FR. The role of sex hormones and steroid receptors on female reproductive cancers. Steroids. 2017;118:93–108. [DOI] [PubMed] [Google Scholar]
- 9. Mattu HS, Randeva HS. Role of adipokines in cardiovascular disease. J Endocrinol. 2013;216(1):T17–36. [DOI] [PubMed] [Google Scholar]
- 10. Nimptsch K, Pischon T. Obesity biomarkers, metabolism and risk of cancer: an epidemiological perspective. Recent Results Cancer Res. 2016;208:199–217. [DOI] [PubMed] [Google Scholar]
- 11. Pham NM, Nanri A, Kochi T, Kuwahara K, Tsuruoka H, Kurotani K, Akter S, Kabe I, Sato M, Hayabuchi H et al.. Coffee and green tea consumption is associated with insulin resistance in Japanese adults. Metabolism. 2014;63(3):400–8. [DOI] [PubMed] [Google Scholar]
- 12. Shaposhnikov S, Hatzold T, Yamani NE, Stavro PM, Lorenzo Y, Dusinska M, Reus A, Pasman W, Collins A. Coffee and oxidative stress: a human intervention study. Eur J Nutr. 2018;57(2):533–44. [DOI] [PubMed] [Google Scholar]
- 13. Landin-Wilhelmsen K, Wilhelmsen L, Lappas G, Rosen T, Lindstedt G, Lundberg PA, Bengtsson BA. Serum insulin-like growth factor I in a random population sample of men and women: relation to age, sex, smoking habits, coffee consumption and physical activity, blood pressure and concentrations of plasma lipids, fibrinogen, parathyroid hormone and osteocalcin. Clin Endocrinol (Oxf). 1994;41(3):351–7. [DOI] [PubMed] [Google Scholar]
- 14. Maruyama K, Iso H, Ito Y, Watanabe Y, Inaba Y, Tajima K, Nakachi K, Tamakoshi A; JACC Study Group. Associations of food and nutrient intakes with serum IGF-I, IGF-II, IGFBP-3, TGF-b1, total SOD activity and sFas levels among middle-aged Japanese: the Japan Collaborative Cohort study. Asian Pac J Cancer Prev. 2009;10(Suppl):7–22. [PubMed] [Google Scholar]
- 15. Schliep KC, Schisterman EF, Mumford SL, Pollack AZ, Zhang C, Ye A, Stanford JB, Hammoud AO, Porucznik CA, Wactawski-Wende J. Caffeinated beverage intake and reproductive hormones among premenopausal women in the BioCycle study. Am J Clin Nutr. 2012;95(2):488–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Frey T, Platz EA, Kanarek N, Bradwin G, Dobs AS, Rohrmann S. Consumption of caffeinated beverages and serum concentrations of sex steroid hormones in US men. Cancer Causes Control. 2018;29(1):157–66. [DOI] [PubMed] [Google Scholar]
- 17. Williams CJ, Fargnoli JL, Hwang JJ, van Dam RM, Blackburn GL, Hu FB, Mantzoros CS. Coffee consumption is associated with higher plasma adiponectin concentrations in women with or without type 2 diabetes: a prospective cohort study. Diabetes Care. 2008;31(3):504–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Jacobs S, Kroger J, Floegel A, Boeing H, Drogan D, Pischon T, Fritsche A, Prehn C, Adamski J, Isermann B et al.. Evaluation of various biomarkers as potential mediators of the association between coffee consumption and incident type 2 diabetes in the EPIC-Potsdam Study. Am J Clin Nutr. 2014;100(3):891–900. [DOI] [PubMed] [Google Scholar]
- 19. Zampelas A, Panagiotakos DB, Pitsavos C, Chrysohoou C, Stefanadis C. Associations between coffee consumption and inflammatory markers in healthy persons: the ATTICA study. Am J Clin Nutr. 2004;80(4):862–7. [DOI] [PubMed] [Google Scholar]
- 20. Loftfield E, Shiels MS, Graubard BI, Katki HA, Chaturvedi AK, Trabert B, Pinto LA, Kemp TJ, Shebl FM, Mayne ST et al.. Associations of coffee drinking with systemic immune and inflammatory markers. Cancer Epidemiol Biomarkers Prev. 2015;24(7):1052–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Lopez-Garcia E, van Dam RM, Qi L, Hu FB. Coffee consumption and markers of inflammation and endothelial dysfunction in healthy and diabetic women. Am J Clin Nutr. 2006;84(4):888–93. [DOI] [PubMed] [Google Scholar]
- 22. Wedick NM, Mantzoros CS, Ding EL, Brennan AM, Rosner B, Rimm EB, Hu FB, van Dam RM. The effects of caffeinated and decaffeinated coffee on sex hormone-binding globulin and endogenous sex hormone levels: a randomized controlled trial. Nutr J. 2012;11:86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kempf K, Kolb H, Gartner B, Bytof G, Stiebitz H, Lantz I, Lang R, Hofmann T, Martin S . Cardiometabolic effects of two coffee blends differing in content for major constituents in overweight adults: a randomized controlled trial. Eur J Nutr. 2015;54(5):845–54. [DOI] [PubMed] [Google Scholar]
- 24. Kempf K, Herder C, Erlund I, Kolb H, Martin S, Carstensen M, Koenig W, Sundvall J, Bidel S, Kuha S et al.. Effects of coffee consumption on subclinical inflammation and other risk factors for type 2 diabetes: a clinical trial. Am J Clin Nutr. 2010;91(4):950–7. [DOI] [PubMed] [Google Scholar]
- 25. Colditz GA, Manson JE, Hankinson SE. The Nurses’ Health Study: 20-year contribution to the understanding of health among women. J Womens Health. 1997;6(1):49–62. [DOI] [PubMed] [Google Scholar]
- 26. Rimm EB, Giovannucci EL, Willett WC, Colditz GA, Ascherio A, Rosner B, Stampfer MJ. Prospective study of alcohol consumption and risk of coronary disease in men. Lancet. 1991;338(8765):464–8. [DOI] [PubMed] [Google Scholar]
- 27. Hunter DJ, Hankinson SE, Hough H, Gertig DM, Garcia-Closas M, Spiegelman D, Manson JE, Colditz GA, Willett WC, Speizer FE et al.. A prospective study of NAT2 acetylation genotype, cigarette smoking, and risk of breast cancer. Carcinogenesis. 1997;18(11):2127–32. [DOI] [PubMed] [Google Scholar]
- 28. Hankinson SE, Willett WC, Manson JE, Hunter DJ, Colditz GA, Stampfer MJ, Longcope C, Speizer FE. Alcohol, height, and adiposity in relation to estrogen and prolactin levels in postmenopausal women. J Natl Cancer Inst. 1995;87(17):1297–302. [DOI] [PubMed] [Google Scholar]
- 29. Wu K, Feskanich D, Fuchs CS, Chan AT, Willett WC, Hollis BW, Pollak MN, Giovannucci E. Interactions between plasma levels of 25-hydroxyvitamin D, insulin-like growth factor (IGF)-1 and C-peptide with risk of colorectal cancer. PLoS One. 2011;6(12):e28520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lin JH, Zhang SM, Rexrode KM, Manson JE, Chan AT, Wu K, Tworoger SS, Hankinson SE, Fuchs C, Gaziano JM et al.. Association between sex hormones and colorectal cancer risk in men and women. Clin Gastroenterol Hepatol. 2013;11(4):419–24.. e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Song M, Zhang X, Wu K, Ogino S, Fuchs CS, Giovannucci EL, Chan AT. Plasma adiponectin and soluble leptin receptor and risk of colorectal cancer: a prospective study. Cancer Prev Res (Phila). 2013;6(9):875–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Pai JK, Pischon T, Ma J, Manson JE, Hankinson SE, Joshipura K, Curhan GC, Rifai N, Cannuscio CC, Stampfer MJ et al.. Inflammatory markers and the risk of coronary heart disease in men and women. N Engl J Med. 2004;351(25):2599–610. [DOI] [PubMed] [Google Scholar]
- 33. Rosner B, Cook N, Portman R, Daniels S, Falkner B. Determination of blood pressure percentiles in normal-weight children: some methodological issues. Am J Epidemiol. 2008;167(6):653–66. [DOI] [PubMed] [Google Scholar]
- 34. Sodergard R, Backstrom T, Shanbhag V, Carstensen H. Calculation of free and bound fractions of testosterone and estradiol-17 beta to human plasma proteins at body temperature. J Steroid Biochem. 1982;16(6):801–10. [DOI] [PubMed] [Google Scholar]
- 35. Salvini S, Hunter DJ, Sampson L, Stampfer MJ, Colditz GA, Rosner B, Willett WC. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;18(4):858–67. [DOI] [PubMed] [Google Scholar]
- 36. Rosner B. Percentage points for a generalized ESD many-outlier procedure. Technometrics. 1983;25(2):165–72. [Google Scholar]
- 37. Winkelmayer WC, Stampfer MJ, Willett WC, Curhan GC. Habitual caffeine intake and the risk of hypertension in women. JAMA. 2005;294(18):2330–5. [DOI] [PubMed] [Google Scholar]
- 38. Grobbee DE, Rimm EB, Giovannucci E, Colditz G, Stampfer M, Willett W. Coffee, caffeine, and cardiovascular disease in men. N Engl J Med. 1990;323(15):1026–32. [DOI] [PubMed] [Google Scholar]
- 39. Wylie K, Rees M, Hackett G, Anderson R, Bouloux PM, Cust M, Goldmeier D, Kell P, Terry T, Trinick T et al.. Androgens, health and sexuality in women and men. Maturitas. 2010;67(3):275–89. [DOI] [PubMed] [Google Scholar]
- 40. Buscemi S, Verga S, Batsis JA, Donatelli M, Tranchina MR, Belmonte S, Mattina A, Re A, Cerasola G. Acute effects of coffee on endothelial function in healthy subjects. Eur J Clin Nutr. 2010;64(5):483–9. [DOI] [PubMed] [Google Scholar]
- 41. Wu T, Willett WC, Hankinson SE, Giovannucci E. Caffeinated coffee, decaffeinated coffee, and caffeine in relation to plasma C-peptide levels, a marker of insulin secretion, in U.S. women. Diabetes Care. 2005;28(6):1390–6. [DOI] [PubMed] [Google Scholar]
- 42. Tajik N, Tajik M, Mack I, Enck P. The potential effects of chlorogenic acid, the main phenolic components in coffee, on health: a comprehensive review of the literature. Eur J Nutr. 2017;56(7):2215–44. [DOI] [PubMed] [Google Scholar]
- 43. Kaklamani VG, Linos A, Kaklamani E, Markaki I, Mantzoros C. Age, sex, and smoking are predictors of circulating insulin-like growth factor 1 and insulin-like growth factor-binding protein 3. J Clin Oncol. 1999;17(3):813–17. [DOI] [PubMed] [Google Scholar]
- 44. Choi YY, Choi Y, Kim J, Choi H, Shin J, Roh J. Peripubertal caffeine exposure impairs longitudinal bone growth in immature male rats in a dose- and time-dependent manner. J Med Food. 2016;19(1):73–84. [DOI] [PubMed] [Google Scholar]
- 45. Pollak MN, Schernhammer ES, Hankinson SE. Insulin-like growth factors and neoplasia. Nat Rev Cancer. 2004;4(7):505–18. [DOI] [PubMed] [Google Scholar]
- 46. Nelles JL, Hu WY, Prins GS. Estrogen action and prostate cancer. Expert Rev Endocrinol Metab. 2011;6(3):437–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Somboonporn W, Davis SR; National Health and Medical Research Council. Testosterone effects on the breast: implications for testosterone therapy for women. Endocr Rev. 2004;25(3):374–88. [DOI] [PubMed] [Google Scholar]
- 48. Thaler MA, Seifert-Klauss V, Luppa PB. The biomarker sex hormone-binding globulin – from established applications to emerging trends in clinical medicine. Best Pract Res Clin Endocrinol Metab. 2015;29(5):749–60. [DOI] [PubMed] [Google Scholar]
- 49. London S, Willett W, Longcope C, McKinlay S. Alcohol and other dietary factors in relation to serum hormone concentrations in women at climacteric. Am J Clin Nutr. 1991;53(1):166–71. [DOI] [PubMed] [Google Scholar]
- 50. Cooper C, Atkinson EJ, Wahner HW, O'Fallon WM, Riggs BL, Judd HL, Melton LJ 3rd. Is caffeine consumption a risk factor for osteoporosis?. J Bone Miner Res. 1992;7(4):465–71. [DOI] [PubMed] [Google Scholar]
- 51. Ferrini RL, Barrett-Connor E. Caffeine intake and endogenous sex steroid levels in postmenopausal women: the Rancho Bernardo Study. Am J Epidemiol. 1996;144(7):642–4. [DOI] [PubMed] [Google Scholar]
- 52. Nagata C, Kabuto M, Shimizu H. Association of coffee, green tea, and caffeine intakes with serum concentrations of estradiol and sex hormone-binding globulin in premenopausal Japanese women. Nutr Cancer. 1998;30(1):21–4. [DOI] [PubMed] [Google Scholar]
- 53. Ferrini RL, Barrett-Connor E. Sex hormones and age: a cross-sectional study of testosterone and estradiol and their bioavailable fractions in community-dwelling men. Am J Epidemiol. 1998;147(8):750–4. [DOI] [PubMed] [Google Scholar]
- 54. Kotsopoulos J, Eliassen AH, Missmer SA, Hankinson SE, Tworoger SS. Relationship between caffeine intake and plasma sex hormone concentrations in premenopausal and postmenopausal women. Cancer. 2009;115(12):2765–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Simpson ER. Sources of estrogen and their importance. J Steroid Biochem Mol Biol. 2003;86(3–5):225–30. [DOI] [PubMed] [Google Scholar]
- 56. Ezzat AR, el-Gohary ZM. Hormonal and histological effects of chronic caffeine administration on the pituitary-gonadal and pituitary-adrenocortical axes in male rabbits. Funct Dev Morphol. 1994;4(1):45–50. [PubMed] [Google Scholar]
- 57. Pollard I. Increases in plasma concentrations of steroids in the rat after the administration of caffeine: comparison with plasma disposition of caffeine. J Endocrinol. 1988;119(2):275–80. [DOI] [PubMed] [Google Scholar]
- 58. Goto A, Song Y, Chen BH, Manson JE, Buring JE, Liu S. Coffee and caffeine consumption in relation to sex hormone-binding globulin and risk of type 2 diabetes in postmenopausal women. Diabetes. 2011;60(1):269–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Svartberg J, Midtby M, Bonaa KH, Sundsfjord J, Joakimsen RM, Jorde R. The associations of age, lifestyle factors and chronic disease with testosterone in men: the Tromso Study. Eur J Endocrinol. 2003;149(2):145–52. [DOI] [PubMed] [Google Scholar]
- 60. Mantzoros CS, Georgiadis EI. Body mass and physical activity are important predictors of serum androgen concentrations in young healthy men. Epidemiology. 1995;6(4):432–5. [DOI] [PubMed] [Google Scholar]
- 61. Jung UJ, Choi MS. Obesity and its metabolic complications: the role of adipokines and the relationship between obesity, inflammation, insulin resistance, dyslipidemia and nonalcoholic fatty liver disease. Int J Mol Sci. 2014;15(4):6184–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62. Whitehead JP, Richards AA, Hickman IJ, Macdonald GA, Prins JB. Adiponectin – a key adipokine in the metabolic syndrome. Diabetes, Obes Metab. 2006;8(3):264–80. [DOI] [PubMed] [Google Scholar]
- 63. Imatoh T, Tanihara S, Miyazaki M, Momose Y, Uryu Y, Une H. Coffee consumption but not green tea consumption is associated with adiponectin levels in Japanese males. Eur J Nutr. 2011;50(4):279–84. [DOI] [PubMed] [Google Scholar]
- 64. Pham NM, Nanri A, Yasuda K, Kurotani K, Kuwahara K, Akter S, Sato M, Hayabuchi H, Mizoue T. Habitual consumption of coffee and green tea in relation to serum adipokines: a cross-sectional study. Eur J Nutr. 2015;54(2):205–14. [DOI] [PubMed] [Google Scholar]
- 65. Mure K, Maeda S, Mukoubayashi C, Mugitani K, Iwane M, Kinoshita F, Mohara O, Takeshita T. Habitual coffee consumption inversely associated with metabolic syndrome-related biomarkers involving adiponectin. Nutrition. 2013;29(7–8):982–7. [DOI] [PubMed] [Google Scholar]
- 66. Lee CB, Yu SH, Kim NY, Kim SM, Kim SR, Oh SJ, Jee SH, Lee JE. Association between coffee consumption and circulating levels of adiponectin and leptin. J Med Food. 2017;20(11):1068–75. [DOI] [PubMed] [Google Scholar]
- 67. Marques-Vidal P, Bochud M, Paccaud F, Mooser V, Waeber G, Vollenweider P. Distribution of plasma levels of adiponectin and leptin in an adult Caucasian population. Clin Endocrinol (Oxf). 2010;72(1):38–46. [DOI] [PubMed] [Google Scholar]
- 68. Yamashita K, Yatsuya H, Muramatsu T, Toyoshima H, Murohara T, Tamakoshi K. Association of coffee consumption with serum adiponectin, leptin, inflammation and metabolic markers in Japanese workers: a cross-sectional study. Nutr Diabetes. 2012;2:e33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69. Imatoh T, Kamimura S, Miyazaki M. Coffee but not green tea consumption is associated with prevalence and severity of hepatic steatosis: the impact on leptin level. Eur J Clin Nutr. 2015;69(9):1023–7. [DOI] [PubMed] [Google Scholar]
- 70. Cornelis MC, Gustafsson S, Arnlov J, Elmstahl S, Soderberg S, Sundstrom J, Michaelsson K, Lind L, Ingelsson E. Targeted proteomic analysis of habitual coffee consumption. J Intern Med. 2018;283(2):200–11. [DOI] [PubMed] [Google Scholar]
- 71. Westerterp-Plantenga MS, Lejeune MP, Kovacs EM. Body weight loss and weight maintenance in relation to habitual caffeine intake and green tea supplementation. Obes Res. 2005;13(7):1195–204. [DOI] [PubMed] [Google Scholar]
- 72. Zheng G, Qiu Y, Zhang QF, Li D. Chlorogenic acid and caffeine in combination inhibit fat accumulation by regulating hepatic lipid metabolism-related enzymes in mice. Br J Nutr. 2014;112(6):1034–40. [DOI] [PubMed] [Google Scholar]
- 73. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest. 2006;116(7):1793–801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74. Bouman A, Heineman MJ, Faas MM. Sex hormones and the immune response in humans. Hum Reprod Update. 2005;11(4):411–23. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.