Associations between Caffeine Intake and the Risk of Mortality in Adults with Prediabetes: Evidence from NHANES 1999-2018

doi:10.21203/rs.3.rs-3402030/v1

Download PDF

Research Article

Associations between Caffeine Intake and the Risk of Mortality in Adults with Prediabetes: Evidence from NHANES 1999-2018

https://doi.org/10.21203/rs.3.rs-3402030/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Prediabetes is a pre-clinical disease state, in which individuals have a higher risk of developing diabetes, cardiovascular disease, and death compared to the general population. Recent studies have shown that consuming caffeine may decrease mortality risk in people with diabetes. However, it is not clear if the same effect exists in individuals with prediabetes.

Objectives

In this study, we investigated the potential impact of caffeine intake on the risk of all-cause mortality and cardiovascular mortality in adults with prediabetes.

Methods

we analyzed data from the National Health and Nutrition Examination Survey (NHANES) 1999–2018, and the vital status of participants was tracked until December 31, 2019 using the National Death Index (NDI). Caffeine intake was measured at baseline via 24-hour dietary recall. Multivariable Cox regression analysis and subgroup analysis were conducted to examine the independent relationships among caffeine intake and mortality.

Results

After approximately 20 years of follow-up, 11,581 adults with prediabetes were included, and there were 1,892 deaths, including 497 deaths due to cardiovascular diseases. The results revealed that patients in the highest quintile of caffeine intake had significantly lower hazard ratios (HRs) and 95% confidence intervals (CIs) compared to those in the lowest quintile. Specifically, the HR for all-cause mortality was 0.70 (with a 95% CI of 0.58–0.84), and the HR for cardiovascular mortality was 0.68 (with a 95% CI of 0.46–0.99).

Conclusions

Our findings indicate a positive effect of caffeine intake on reducing the risk of all-cause and cardiovascular mortality in adults with prediabetes.

caffeine intake

prediabetes

all-cause mortality

cardiovascular mortality

NHANES

Prediabetes is a pre-clinical stage between normoglycemia and diabetes, characterized by impaired glucose tolerance (IGT) and/or impaired fasting glucose (IFG) ¹. The pathophysiological defects of prediabetes are similar to those of diabetes, including insulin resistance, pancreatic cells dysfunction, inflammation, oxidative stress, and endothelial dysfunction, etc. ^{2, 3}. Studies have shown that prediabetes is likely to develop type 2 diabetes (T2D) and is associated with early complications of T2D, such as early forms of nephropathy, retinopathy and neuropathy ⁴. According to the epidemiological evidence, prediabetes affects 51% of adults in the US and 720 million people worldwide. People with prediabetes have higher risks of all-cause mortality and cardiovascular disease (CVD), ranging from 13–32% ⁵, and a recent meta-analysis updated this data to 39% ⁶. Despite the high prevalence and hazard, prediabetes is easily overlooked due to the lack of clinical symptoms. Fortunately, prediabetes is reversible and it is reported that lifestyle modification, such as dietary management and exercise are effective measures ⁷. Therefore, it is necessary to focus on this disease with the hope that lifestyle intervention can slow down the development of prediabetes, and reduce the incidence of T2D, CVD, and mortality in prediabetic populations.

Caffeine presents in lots of regular foods, including coffee, tea, beverages, and even some drugs. In the US, people consume caffeine mainly through drinking coffee ⁸. Coffee contains a wide range of biologically active substances, and there have been many studies on the biological activity of caffeine. Studies have shown that caffeine exerts anti-inflammatory and antioxidant effects by antagonizing adenosine receptors ^{9, 10}. A recent experimental animal study revealed that caffeine was able to decrease levels of pro-inflammatory cytokines and increase the expression of anti-inflammatory cytokines ¹¹. Another study proved that caffeine could increase the level of adiponectin, reduce the expression of reactive oxygen species (ROS) ¹². Studies have reported that caffeine intake can not only delay the occurrence of prediabetes in normal glucose people, but also prevent the progression of T2D in prediabetes population^13–16. In addition, consuming caffeine may decrease mortality risk in people with diabetes. For example, a prospective study found a negative association between caffeine consumption and mortality in individuals with T2D ¹⁷. Another study demonstrated that caffeine intake had a positive impact on mortality in women with T2D ¹⁸. In fact, far more people are currently diagnosed with prediabetes than diabetes ¹⁹. However, the relationship between caffeine and mortality in people with prediabetes is not yet fully understood.

To fill this gap, we conducted a study based on the data from the National Health and Nutrition Examination Survey (NHANES) 1999–2018. We hypothesized that caffeine would be a protective factor to reduce the mortality in patients with prediabetes.

2.1 Study Population

The NHANES program is a comprehensive study of the health and nutrition of people in the United States. It uses a combination of interviews and physical assessments to collect data. The interviews cover a wide range of topics, such as demographic information, socioeconomic status, diet, and health-related questions. The physical examinations take place in mobile examination centers (MECs) and encompass various medical, dental, and physiological measurements, as well as laboratory tests.

We conducted a prospective study using data from the NHANES 1999–2018, combined with mortality data from the National Death Index (NDI). We focused on participants aged ≥ 20 years with prediabetes, resulting in a sample of 13,268 subjects. Then excluded individuals with missing data on dietary caffeine intake (n = 833), participants who were pregnant at baseline (n = 78), those with extreme energy intake (n = 755), and participants with missing follow-up data (n = 21). Finally, our study included 11,581 participants with prediabetes (Fig. 1).

2.2 Ascertainment of Prediabetes

According to the American Diabetes Association (ADA) guidelines, prediabetes is characterized by not having diabetes mellitus, but meeting certain criteria, including: 1) fasting plasma glucose levels ranging from 100–125 mg/dL (5.6–6.9 mmol/L); 2) 2-hour plasma glucose levels during the oral glucose tolerance test (OGTT) ranging from 140–199 mg/dL (7.8–11.0 mmol/L); or 3) hemoglobin A1c (HbA1c) levels ranging from 5.7% − 6.4% ²⁰.

2.3 Assessment of Caffeine Intake

In NHANES 1999–2002, all participants underwent a 24-hour dietary recall interview. Since 2003, two interviews were included, the first was the same as before, and then after 3 to 10 days, the second interview was conducted by telephone. Consequently, for the 1999–2002 cycle, information from a single 24-hour dietary recall was used, while for the 2003–2018 period, the average of two 24-hour dietary recalls was used ²¹. In order to record the quantity of food more accurately, each MEC was equipped with a standard measuring tool, and the data were recorded through an automated multiple-pass method ²². The study participants were categorized into four quintiles (Q1–Q4), based on their daily caffeine intake levels: < 29.5 mg/d, 29.5–101.0 mg/d, 101.1–210.0 mg/d, and > 210.0 mg/d.

2.4 Ascertainment of Mortality

In this study, we focused on two main outcomes: all-cause mortality and cardiovascular mortality. To obtain information on mortality, we linked survey data from the National Center for Health Statistics (NCHS) to the NDI death certificates. This linked mortality file (LMF) was updated until December 31, 2019 ²³. To identify the specific cause of death, we referenced disease codes from the International Classification of Diseases, 10th Revision (ICD-10), which I00-I09, I11, I13, I20-I51, and I60-I69 represent deaths due to CVD, and C00-C97 represent deaths due to cancer. The days between the first examination date and the last survival date represents the follow-up time for each participant.

2.5 Assessment of Covariates

We collected information on age, gender, race or ethnicity, education level, family income to poverty ratio (PIR), smoking status, drinking status, physical activity and energy intake from standardized questionnaires. Race or ethnicity was classified into: Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black, and others. Education level was classified as below high school, high school, and above high school. Family PIR was stratified into three groups, based on eligibility for federal food assistance programs and considering the number of individuals in the household ²⁴. Smoking status was divided into: nonsmoker, former smoker, and current smoker. Drinking status was also divided into: nondrinker, low-to-moderate drinker, and heavy drinker. Based on self-reported weekly leisure-time activities and metabolic equivalent (MET), physical activity was divided into: inactive group (no leisure activity), insufficiently active group (moderate activity 1–5 times weekly with MET between 3–6 or vigorous activity 1–3 times weekly with MET > 6), and active group (moderate or vigorous activity more than above) ²⁵. Energy intake was determined from the 24-hour dietary recall. Hypertension was defined as previously diagnosed with hypertension or taking anti-hypertensive drugs. CVD was defined as having a previous diagnosis of coronary heart disease (CHD), myocardial infarction, heart failure or stroke. Cancer was defined as those with a previous diagnosis of cancer or any type of malignancy.

In addition, we collected data on physical examinations, including body mass index (BMI), fasting blood glucose (FBG), HbA1c, HOMA-IR, total cholesterol (TC), triglyceride (TG), C-reactive protein (CRP), systolic blood pressure (SBP) and diastolic blood pressure (DBP). BMI was calculated by dividing weight in kilograms by height in meters squared and categorized as < 25.0, 25.0-29.9, and > 29.9 kg/m². HOMA-IR was calculated by multiplying insulin and glucose and dividing by 22.5. Measurements of other biochemical indicators were available in the NHANES Laboratory Procedures Manual ²⁶.

2.6 Statistical Analysis

Sample weights, strata, and primary sampling units were included in all analyses. The appropriate weights were selected for each study variable, as recommended by the US Centers for Disease Control and Prevention.

The baseline characteristics of study participants were categorized into continuous and categorical variables. Continuous variables were represented as means and standard deviations (SDs), or medians and interquartile ranges (IQRs). Categorical variables were presented as percentages. The study populations were divided into four quintiles (Q1-Q4) based on the daily caffeine intake levels.

The effect of caffeine intake on all-cause and cardiovascular death was evaluated by weighted Cox proportional hazards regression models, with following covariates: age, gender, and race/ethnicity (model 1); model 1 plus education level, family PIR, smoking status, drinking status, physical activity, total energy intake, and BMI (model 2). The results were expressed as hazard ratios (HRs) and 95% confidence intervals (CIs).

Stratified analyses were performed in the strata of age, gender, race/ethnicity (non-Hispanic White and others), family PIR (≤ 1.0, 1.1–3.0, and > 3.0), smoking status (nonsmoker, former smoker, and current smoker), drinking status (nondrinker, low-to-moderate drinker, and heavy drinker.), physical activity (inactive, insufficiently active and active), BMI (< 25.0, 25.0-29.9, and > 29.9 kg/m²), and self-reported hypertension (yes or no). P values were also calculated for the interaction between caffeine intake levels and stratified variables. To investigate dose-response associations between caffeine intake levels with all-cause and cardiovascular mortality, we applied a restricted cubic spline (RCS) model.

In addition, we carried out sensitivity analyses to validate the reliability of the findings. Firstly, in order to reduce interference of reverse causality, we excluded participants who passed away within the first 2 years. Secondly, individuals with a history of CVD were excluded. Thirdly, those who had cancer history were also excluded. Finally, we adjusted for several cardiovascular risk factors to explore the relationship between caffeine intake and the risk of mortality in individuals with prediabetes.

The statistical analyses for this study were performed using R software, version 4.2.2. We considered two-sided P values less than 0.05 as statistically significant.

In this study, 11,581 adults with prediabetes were enrolled (52.15%, men) from NHANES 1999–2018 (Table S1). According to the quartiles of caffeine intake levels, the baseline characteristics were listed in Table 1. It showed that individuals who consumed more caffeine tended to be older, predominantly men, non-Hispanic white, smokers and drinkers. They had higher levels of education and family income, and were more likely to be obese with higher levels of TC. Moreover, they had a higher proportion of insufficient physical activity, consumed more energy, and had lower levels of HbA1c (P < 0.05). However, caffeine intake seemed to have weak association with blood pressure.

Table 1

Baseline characteristics of adults with prediabetes according to quartiles of caffeine intake levels in NHANES 1999–2018.
Characteristics	Quartiles of caffeine intake levels, mg/day				P value
Characteristics	< 29.5 (n = 2916)	29.5–101.0 (n = 2877)	101.1–210.0 (n = 2898)	> 210.0 (n = 2890)	P value
Age, years					< 0.001
20–39	701 (28.84)	651 (25.58)	555 (22.81)	388 (13.65)
40–59	938 (34.64)	1021 (41.17)	1034 (40.02)	1249 (49.34)
≥ 60	1277 (36.52)	1205 (33.25)	1309 (37.17)	1253 (37.01)
Male, %	1414 (47.50)	1381 (45.21)	1489 (48.10)	1756 (61.07)	< 0.001
Race/ethnicity, %					< 0.001
Mexican American	535 (11.24)	584 (11.11)	485 (7.26)	336 (4.20)
Other Hispanic	263 (6.36)	295 (7.08)	267 (5.63)	192 (3.23)
non-Hispanic White	883 (51.69)	942 (56.09)	1346 (70.81)	1889 (82.87)
non-Hispanic Black	938 (21.76)	718 (15.44)	527 (9.29)	260 (4.03)
Other races	297 (8.94)	338 (10.29)	273 (7.01)	213 (5.67)
Education level, %					< 0.001
Below high school	930 (22.43)	904 (20.62)	767 (16.64)	640 (15.16)
High school	633 (23.18)	665 (26.36)	683 (25.02)	739 (26.27)
Above high school	1353 (54.40)	1308 (53.02)	1448 (58.35)	1511 (58.57)
Family PIR, %					< 0.001
≤ 1.0	673 (18.48)	633 (15.48)	498 (12.33)	421 (9.22)
1.1–3.0	1282 (39.06)	1267 (40.32)	1258 (36.10)	1174 (33.73)
> 3.0	961 (42.46)	977 (44.21)	1142 (51.57)	1295 (57.06)
Smoking status, %					< 0.001
Nonsmoker	1840 (63.37)	1721 (60.73)	1442 (50.57)	1008 (36.40)
Former smoker	662 (22.30)	720 (24.44)	893 (31.86)	996 (34.88)
Current smoker	414 (14.33)	436 (14.83)	563 (17.57)	886 (28.73)
Drinking status, %					< 0.001
Nondrinker	939 (29.35)	774 (23.80)	618 (18.19)	433 (13.17)
Low-to-moderate drinker	1756 (62.31)	1898 (67.89)	2098 (73.75)	2170 (75.55)
Heavy drinker	221 (8.34)	205 (8.31)	182 (8.06)	287 (11.28)
BMI, %					< 0.001
< 25.0 kg/m²	709 (24.17)	675 (23.13)	585 (19.24)	568 (18.49)
25.0–29.9 kg/m²	959 (30.59)	1003 (33.02)	1068 (35.74)	1082 (39.30)
> 29.9 kg/m²	1248 (45.24)	1199 (43.85)	1245 (45.01)	1240 (42.21)
Physical activity, %					0.012
Inactive	865 (25.64)	862 (26.71)	826 (24.28)	733 (21.27)
Insufficiently active	968 (35.17)	998 (36.22)	1041 (37.78)	1080 (39.94)
Active	1083 (39.19)	1017 (37.07)	1031 (37.94)	1077 (38.79)
TC, mg/dL	194.00 (168.00,221.00)	199.00 (174.00,229.00)	200.00 (174.00,228.00)	202.00 (177.00,229.00)	< 0.001
Energy intakes, kcal/day	1806.00 (1387.00,2294.50)	1865.50 (1455.00,2358.50)	1963.50 (1540.00,2473.50)	2124.00 (1695.00,2679.50)	< 0.001
Systolic BP, mmHg	124.00 (114.00,136.00)	123.33 (113.33,136.67)	124.00 (114.67,136.00)	123.33 (114.00,134.67)	0.711
Diastolic BP, mmHg	72.00 (64.67,79.33)	72.67 (65.33,79.33)	72.67 (64.67,80.00)	72.00 (65.33,79.33)	0.877
HbA1c, %	5.80 (5.40,5.90)	5.80 (5.40,5.90)	5.70 (5.40,5.90)	5.70 (5.40,5.90)	0.019
Self-reported hypertension, %	1253 (40.94)	1168 (39.67)	1175 (38.11)	1167 (39.30)	0.494
Caffeine intakes, mg/day	4.00 (0.00, 15.00)	65.00 (47.36, 83.00)	150.50 (126.50,178.00)	321.00 (256.00,441.00)	< 0.001
Abbreviations: PIR, income to poverty ratio; BMI, body mass index; TC, total cholesterol; BP, blood pressure; HbA1c; hemoglobin A1c. Normally distributed continuous variables are described as means ± SDs, and continuous variables without a normal distribution are presented as medians and interquartile ranges. Categorical variables are presented as percentages. N reflect the study sample while percentages reflect the survey-weighted.

During the 20 years follow-up, 1,892 deaths occurred (16.3% of the study population), of which 497 (4.3% of the study population) were due to CVD. The HRs for all-cause and cardiovascular mortality based on quartiles of daily caffeine intake levels were listed in Table 2. After adjusted for various factors such as age, gender, race, education level, family income, smoking and drinking status, physical activity, total energy intakes, BMI, TC and self-reported hypertension, we discovered that there was a protective effect of caffeine intake on all-cause mortality in individuals with prediabetes. Additionally, we revealed an inverse relationship between coffee intake and cardiovascular mortality. Compared to those with the lowest quintile of caffeine intake, those in the highest quintile had HRs (95% CIs) of 0.70 (0.58–0.84) for all-cause mortality and 0.68 (0.46–0.99) for cardiovascular mortality.

Table 2

HRs (95% CIs) of all-cause and cardiovascular mortality according to quartiles of caffeine intake levels among adults with prediabetes in NHANES 1999–2018.
	Quartiles of caffeine intake levels, mg/day				P_trend
	< 29.5	29.5–101.0	101.1–210.0	> 210.0	P_trend
All-cause mortality
No. deaths/total	504/2916	450/2877	455/2898	483/2890
Crude	1.00 [Reference]	0.82 (0.70–0.97)	0.75 (0.63–0.90)	0.78 (0.65–0.93)	0.009
Model 1 ^a	1.00 [Reference]	0.88 (0.74–1.05)	0.71 (0.60–0.84)	0.69 (0.58–0.82)	< 0.001
Model 2 ^b	1.00 [Reference]	0.87 (0.74–1.03)	0.73 (0.61–0.88)	0.70 (0.58–0.84)	< 0.001
Cardiovascular mortality
No. deaths/total	134/2916	145/2877	110/2898	108/2890
Crude	1.00 [Reference]	1.04 (0.76–1.43)	0.69 (0.51–0.94)	0.69 (0.48–0.98)	0.007
Model 1 ^a	1.00 [Reference]	1.16 (0.84–1.61)	0.65 (0.47–0.90)	0.60 (0.42–0.85)	< 0.001
Model 2 ^b	1.00 [Reference]	1.17 (0.84–1.61)	0.70 (0.50–0.98)	0.68 (0.46–0.99)	0.007
Data are presented as HR (95% CI) unless indicated otherwise;
^a Adjusted for age (20–39, 40–59, or ≥ 60), gender (male or female), and race/ethnicity (Mexican American, Other Hispanic, non-Hispanic White, non-Hispanic Black or Other);
^b Further adjusted for education level (below high school, high school, or above high school), family PIR (≤ 1.0, 1.1-3.0, or > 3.0), drinking status (nondrinker, low-to-moderate drinker, or heavy drinker), smoking status (nonsmoker, former smoker, or current smoker), BMI (< 25.0, 25.0-29.9, or > 29.9), physical activity (inactive, insufficiently active, or active), total energy intakes (in quartiles), total cholesterol (in quartiles), self-reported hypertension (yes or no).

In order to exclude the interaction between confounding factors on caffeine intake and all-cause mortality. We performed stratified analyses that considered age, gender, race/ethnicity, family PIR, smoking and drinking status, physical activity, BMI and hypertension (Table 3). The results revealed that there was no significant interaction between the factors we examined on the relationship between caffeine intake and the risk of all-cause mortality in patients with prediabetes.

Table 3

Stratified analyses of the associations between quartiles of caffeine intake levels and all-cause mortality among adults with prediabetes in NHANES 1999–2018.
Subgroups	N	Quartiles of caffeine intake levels, mg/day				P_trend	P_interaction
Subgroups	N	< 29.5	29.5–101.0	101.1–210.0	> 210.0	P_trend	P_interaction
Age, years
20–39	2295	1.00 (Ref.)	0.30 (0.09–0.99)	0.69 (0.26–1.82)	0.49 (0.16–1.51)	0.293	0.107
40–59	4242	1.00 (Ref.)	0.93 (0.54–1.58)	0.62 (0.35–1.12)	0.75 (0.44–1.31)	0.252
≥ 60	5044	1.00 (Ref.)	0.90 (0.75–1.07)	0.75 (0.62–0.91)	0.70 (0.57–0.86)	< 0.001
Gender, %
Male	6040	1.00 (Ref.)	0.86 (0.66–1.12)	0.81 (0.63–1.04)	0.71 (0.54–0.93)	0.01	0.486
Female	5541	1.00 (Ref.)	0.90 (0.71–1.13)	0.68 (0.51–0.91)	0.70 (0.53–0.91)	0.003	0.486
Race/ethnicity, %
non-Hispanic White	5060	1.00 (Ref.)	0.79 (0.59–1.05)	0.83 (0.61–1.13)	0.67 (0.47–0.96)	0.051	0.559
Others	6521	1.00 (Ref.)	0.90 (0.73–1.10)	0.71 (0.56–0.89)	0.69 (0.56–0.86)	< 0.001	0.559
Family PIR, %
≤ 1.0	2225	1.00 (Ref.)	0.87 (0.62–1.22)	0.77 (0.50–1.18)	0.89 (0.64–1.24)	0.417	0.738
1.1–3.0	4981	1.00 (Ref.)	0.88 (0.69–1.11)	0.76 (0.59–0.97)	0.69 (0.52–0.91)	0.005
> 3.0	4375	1.00 (Ref.)	0.91 (0.67–1.23)	0.65 (0.47–0.89)	0.63 (0.45–0.89)	0.003
Smoking status, %
Nonsmoker	6011	1.00 (Ref.)	0.90 (0.70–1.15)	0.70 (0.53–0.92)	0.63 (0.45–0.88)	0.002	0.380
Former smoker	3271	1.00 (Ref.)	0.86 (0.65–1.13)	0.85 (0.66–1.09)	0.78 (0.61–0.99)	0.048
Current smoker	2299	1.00 (Ref.)	0.76 (0.46–1.26)	0.60 (0.38–0.94)	0.60 (0.40–0.90)	0.013
Drinking status, %
Nondrinker	2764	1.00 (Ref.)	0.94 (0.72–1.22)	0.79 (0.57–1.09)	0.79 (0.57–1.12)	0.122	0.091
Low-to-moderate drinker	7922	1.00 (Ref.)	0.87 (0.69–1.10)	0.77 (0.60–0.98)	0.77 (0.59-1.00)	0.039	0.091
Heavy drinker	895	1.00 (Ref.)	0.71 (0.40–1.24)	0.38 (0.18–0.77)	0.32 (0.17–0.60)	< 0.001
BMI, kg/m²
< 25.0	2537	1.00 (Ref.)	0.95 (0.69–1.31)	0.82 (0.60–1.14)	0.72 (0.51–1.02)	0.048		0.335
25.0–29.9	4112	1.00 (Ref.)	0.78 (0.58–1.05)	0.61 (0.45–0.83)	0.59 (0.44–0.78)	< 0.001
> 29.9	4932	1.00 (Ref.)	0.90 (0.66–1.25)	0.83 (0.60–1.13)	0.83 (0.59–1.15)	0.259
Physical activity, %
Inactive	3286	1.00 (Ref.)	0.90 (0.71–1.14)	0.81 (0.61–1.09)	0.69 (0.52–0.92)	0.011		0.610
Insufficiently active	4087	1.00 (Ref.)	0.90 (0.67–1.22)	0.69 (0.51–0.94)	0.78 (0.58–1.05)	0.061
Active	4208	1.00 (Ref.)	0.74 (0.53–1.05)	0.69 (0.49–0.96)	0.59 (0.41–0.87)	0.009
Self-reported hypertension, %
Yes	4763	1.00 (Ref.)	0.92 (0.73–1.16)	0.72 (0.59–0.88)	0.64 (0.51–0.81)	< 0.001		0.065
No	6818	1.00 (Ref.)	0.82 (0.62–1.09)	0.75 (0.56–1.02)	0.77 (0.57–1.03)	0.09		0.065
Data are presented as HR (95% CI) unless indicated otherwise; Analyses were adjusted for covariates age (20–39, 40–59, or ≥ 60), gender (male or female), race/ethnicity (Mexican American, Other Hispanic, non-Hispanic White, non-Hispanic Black or Other), education level (below high school, high school, or above high school), family PIR (≤ 1.0, 1.1-3.0, or > 3.0), drinking status (nondrinker, low-to-moderate drinker, or heavy drinker), smoking status (nonsmoker, former smoker, or current smoker), BMI (< 25.0, 25.0-29.9, or > 29.9), physical activity (inactive, insufficiently active, or active), total energy intakes (in quartiles), total cholesterol (in quartiles), self-reported hypertension (yes or no) when they were not the strata variables.

We took sensitivity analyses to ensure the reliability of our findings. Initially, we excluded patients who passed away during the first two years of follow-up, and this did not substantially alter the negative associations between caffeine intake and the risk of death (Table S2). Furthermore, we excluded patients with a history of CVD or cancer at baseline, and the results remained consistent for both all-cause mortality and cardiovascular mortality (Table S3 and Table S4). Finally, we made adjustments for various cardiovascular risk factors, including SBP, DBP, TG, CRP, FBG, HbA1c and HOMR-IR, and still obtained similar results (Table S5). These findings suggest that higher levels of caffeine intake were associated with a lower risk of death from any cause and cardiovascular causes in prediabetic patients.

In addition, we utilized RCS model to examine the associations between caffeine intake and the risk of mortality (Fig. 2). It indicated a dose-dependent relationship, suggesting that as the levels of caffeine intake increased, the possibility of all-cause mortality decreased in people with prediabetes (P for nonlinearity = 0.051). Furthermore, caffeine was a protective factor for cardiovascular mortality. When caffeine intake > 15.07 mg/d, the curve showed a substantial decrease, which implied a significant reduction in cardiovascular mortality (non-linear P < 0.001).

Based on our analysis of data from NHANES 1998–2018, we demonstrated that caffeine acts as a protective factor for both all-cause mortality and cardiovascular mortality in prediabetic populations. Results were independent of confounding factors, including demographic characteristics, diet and lifestyle, disease history and cardiovascular risk factors. Subsequently, we conducted stratified analyses and sensitivity analyses to ensure the reliability of our findings.

There have been a large number of studies looking at how caffeine relates to mortality. However, results in the general population are not entirely consistent. For example, a previous study found that consuming caffeine was linked to lower mortality, such as death due to CVD, respiratory disease, infections, injuries, and accidents ²⁷. Another study also found a significant connection between caffeine intake and lower mortality rates from CVD, neurologic diseases, and suicide ²⁸. Conversely, a study using data from NHANES 1999–2010 showed that caffeine intake was not related to the cardiovascular mortality ²⁹.

Caffeine has various important biological functions, mainly including anti-inflammatory, antioxidant, improving glucose metabolism, and weight loss ^30–33. A recent mendelian randomization study found that some people have a genetic tendency to metabolize caffeine slowly. This leads to higher levels of caffeine in their blood plasma, as well as a lower BMI and a reduced risk of T2D ³⁴. Another study found that moderate coffee drinkers have a 10–15% lower risk of CVD compared to people who never drink coffee. It may due to the positive effects of caffeine on glucose metabolism and inflammation reduction ³⁵. Additionally, evidence suggests that higher coffee consumption is associated with lower death rates and a decreased incidence of cancer, possibly through the ROS-induced apoptosis pathway ³⁶. In experimental animal research, Yamauchi et al. found that caffeine has anti-diabetic properties, which could improve insulin sensitivity, reduce the expression of inflammatory cytokines, and inhibit hyperglycemia in diabetic mice ³⁷. A recent study revealed that caffeine can improve the metabolic syndrome in obese mice by modulating the gut microbiota, lipid metabolism, bile acid metabolism and energy metabolism ³⁸. However, the exact mechanisms by which caffeine affects mortality are still unclear. Nonetheless, it is reasonable to assume that the mechanisms described above could be some of the potential mechanisms involved. Therefore, further studies are needed to elucidate the underlying mechanisms linking caffeine consumption and mortality across a broad population.

There are also studies in the diabetes population. A prospective cohort study of 3,497 men diagnosed with diabetes found that regular caffeine intake would not increase the hazard of CVD or mortality ³⁹. Similarly, another study involving 7,170 women with T2D indicated that caffeine intake was not related to the increased incidence of CVD and mortality ⁴⁰. However, a large prospective study of 3,837 Finnish patients demonstrated that caffeine intake was actually linked to a reduction in all-cause, CVD, and CHD mortality among those with T2D ¹⁷. The variations between these studies may be attributed to differences in sample size, duration of follow-up, and baseline characteristics of the study populations.

Currently, there are limited knowledge regarding the effect of caffeine on mortality in individuals with prediabetes. However, it has been observed that the mortality risk in prediabetic populations is higher than general populations. It is believed that lifestyle interventions have the potential to slow down the progression of prediabetes. Thus, our aim is to investigate whether caffeine intake could be beneficial for patients with prediabetes. Through a long-term follow-up, this prospective cohort study will be the first of its kind to explore the connections between caffeine intake and mortality in adults with prediabetes. One of the main strengths of our study is the utilization of comprehensive and up-to-date data from the NHANES database, which is extensive in scale. Additionally, we have taken into account a wide range of potential confounding factors based on survey data collected from NHANES between the years 1999 and 2018. Finally, we have performed various sensitivity analyses to ensure the reliability of the results.

Furthermore, our study encountered several limitations. Firstly, due to the observational nature, we were unable to establish a causal relationship despite the large sample size and long tracking time. Secondly, the assessment of caffeine intake was based on one or two 24-hour dietary recalls, which may not reflect the participants' real dietary patterns or allow for the assessment of the relationship between interest and time. Thirdly, the data obtained from NHANES only represents the general characteristics of US adults with prediabetes and may not be applicable to other populations. Lastly, although we made efforts to adjust for potential confounding factors, there may still be residual confounding factors that were not accounted for.

In this study, we discovered that consuming caffeine was linked to a lower risk of all-cause mortality in a group of US adults with prediabetes. This association was found to be dependent on the amount of caffeine consumed. Additionally, caffeine intake acted as an independent protective factor on cardiovascular mortality. Based on these findings, we concluded that consuming caffeine could be beneficial for individuals with prediabetes. Adding coffee to their daily diet could be considered as a lifestyle intervention, as the study had reported previously ⁴¹. However, further research is still required, specifically large-scale studies involving multiple centers, to investigate the connection between caffeine intake and mortality in patients with prediabetes.

ADA, American Diabetes Association; BMI, body mass index; CVD, cardiovascular diseases; CHD, coronary heart disease; CI, confidence interval; CRP, C-reactive protein; DBP, diastolic blood pressure; FBG, fasting blood glucose; HR, hazard ratio; HbA1c, hemoglobin A1c; IGT, impaired glucose tolerance; IFG, impaired fasting glucose; ICD, International Classification of Diseases; IQR, interquartile range; LDL, low-density lipoprotein; LMF, linked mortality file; MEC, mobile examination center; MET, metabolic equivalent; NCHS, National Center for Health Statistics; NDI, National Death Index; NHANES, National Health and Nutrition Examination Survey; NGT, normal glucose tolerance; OGTT, oral glucose tolerance test; PIR, family income to poverty ratio; ROS, reactive oxygen species; RCS, restricted cubic spline; SBP, Systolic blood pressure; SD, standard deviation; T2D, type 2 diabetes; TC, total cholesterol; TG, triglyceride.

Acknowledgements

Author Contributions: Dr H. Zhu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Y. Pan designed this cohort study and wrote the original draft manuscript. Y. Wu obtained and interpreted data. S. Wu and R. Li conducted the statistical analysis. J. Hong and W. Liu revised the manuscript for important intellectual content. All authors have read and approved the final manuscript.

Competing interests: Authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding: Dr H. Zhu acknowledged the financial support from the National Natural Science Foundation of China (82370829) and the Zhejiang Provincial Medical Science and Technology Plan (2022491161).

Ethics approval and consent to participate: The ethics protocol was approved by the Research Ethics Review Board of National Center for Health Statistics (https://www.cdc.gov/nchs/nhanes/irba98.htm), and informed consent was signed by all recruited participants.

Availability of data and materials: The data used in this study are freely available for download by the public at: https://wwwn.cdc.gov/nchs/nhanes/Default.aspx.

Additional Contributions: We are grateful to the participants and to the people involved in the National Health and Nutrition Examination Survey study.

American Diabetes A. 2. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes-2020. Diabetes Care. Jan 2020;43(Suppl 1):S14-S31. doi: 10.2337/dc20-S002
Huang Z, Chen C, Li S, Kong F, Shan P, Huang W. Serum Markers of Endothelial Dysfunction and Inflammation Increase in Hypertension with Prediabetes Mellitus. Genet Test Mol Biomarkers. Jun 2016;20(6):322-327. doi:10.1089/gtmb.2015.0255
Abdul-Ghani MA, Tripathy D, DeFronzo RA. Contributions of beta-cell dysfunction and insulin resistance to the pathogenesis of impaired glucose tolerance and impaired fasting glucose. Diabetes Care. May 2006;29(5):1130-1139. doi:10.2337/diacare.2951130
Tabak AG, Herder C, Rathmann W, Brunner EJ, Kivimaki M. Prediabetes: a high-risk state for diabetes development. Lancet. Jun 16 2012;379(9833):2279-2290. doi:10.1016/S0140-6736(12)60283-9
Echouffo-Tcheugui JB, Selvin E. Prediabetes and What It Means: The Epidemiological Evidence. Annu Rev Public Health. Apr 1 2021;42:59-77. doi:10.1146/annurev-publhealth-090419-102644
Schlesinger S, Neuenschwander M, Barbaresko J, et al. Prediabetes and risk of mortality, diabetes-related complications and comorbidities: umbrella review of meta-analyses of prospective studies. Diabetologia. Feb 2022;65(2):275-285. doi:10.1007/s00125-021-05592-3
Glechner A, Keuchel L, Affengruber L, et al. Effects of lifestyle changes on adults with prediabetes: A systematic review and meta-analysis. Prim Care Diabetes. Oct 2018;12(5):393-408. doi:10.1016/j.pcd.2018.07.003
Frary CD, Johnson RK, Wang MQ. Food sources and intakes of caffeine in the diets of persons in the United States. J Am Diet Assoc. Jan 2005;105(1):110-113. doi:10.1016/j.jada.2004.10.027
Barcelos RP, Lima FD, Carvalho NR, Bresciani G, Royes LF. Caffeine effects on systemic metabolism, oxidative-inflammatory pathways, and exercise performance. Nutr Res. Aug 2020;80:1-17. doi:10.1016/j.nutres.2020.05.005
Koloverou E, Panagiotakos DB, Pitsavos C, et al. The evaluation of inflammatory and oxidative stress biomarkers on coffee-diabetes association: results from the 10-year follow-up of the ATTICA Study (2002-2012). Eur J Clin Nutr. Nov 2015;69(11):1220-1225. doi:10.1038/ejcn.2015.98
Eichwald T, Solano AF, Souza J, et al. Anti-Inflammatory Effect of Caffeine on Muscle under Lipopolysaccharide-Induced Inflammation. Antioxidants (Basel). Feb 23 2023;12(3):554. doi:10.3390/antiox12030554 2023;12(3):554
Surma S, Sahebkar A, Banach M. Coffee or tea: Anti-inflammatory properties in the context of atherosclerotic cardiovascular disease prevention. Pharmacol Res. Jan 2023;187:106596. doi:10.1016/j.phrs.2022.106596
Mirmiran P, Carlstrom M, Bahadoran Z, Azizi F. Long-term effects of coffee and caffeine intake on the risk of pre-diabetes and type 2 diabetes: Findings from a population with low coffee consumption. Nutr Metab Cardiovasc Dis. Dec 2018;28(12):1261-1266. doi:10.1016/j.numecd.2018.09.001
Smith B, Wingard DL, Smith TC, Kritz-Silverstein D, Barrett-Connor E. Does coffee consumption reduce the risk of type 2 diabetes in individuals with impaired glucose? Diabetes Care. Nov 2006;29(11):2385-2390. doi:10.2337/dc06-1084
Jiang X, Zhang D, Jiang W. Coffee and caffeine intake and incidence of type 2 diabetes mellitus: a meta-analysis of prospective studies. Eur J Nutr. Feb 2014;53(1):25-38. doi:10.1007/s00394-013-0603-x
Ding M, Bhupathiraju SN, Chen M, van Dam RM, Hu FB. Caffeinated and decaffeinated coffee consumption and risk of type 2 diabetes: a systematic review and a dose-response meta-analysis. Diabetes Care. Feb 2014;37(2):569-586. doi:10.2337/dc13-1203
Bidel S, Hu G, Qiao Q, Jousilahti P, Antikainen R, Tuomilehto J. Coffee consumption and risk of total and cardiovascular mortality among patients with type 2 diabetes. Diabetologia. Nov 2006;49(11):2618-2626. doi:10.1007/s00125-006-0435-9
Neves JS, Leitao L, Magrico R, et al. Caffeine Consumption and Mortality in Diabetes: An Analysis of NHANES 1999-2010. Front Endocrinol (Lausanne). 2018;9:547. doi:10.3389/fendo.2018.00547
Vatcheva KP, Fisher-Hoch SP, Reininger BM, McCormick JB. Sex and age differences in prevalence and risk factors for prediabetes in Mexican-Americans. Diabetes Res Clin Pract. Jan 2020;159:107950. doi:10.1016/j.diabres.2019.107950
Sonne DP, Hemmingsen B. Comment on American Diabetes Association. Standards of Medical Care in Diabetes-2017. Diabetes Care 2017;40(Suppl. 1):S1-S135. Diabetes Care. Jul 2017;40(7):e92-e93. doi:10.2337/dc17-0299
Zipf G, Chiappa M, Porter KS, Ostchega Y, Lewis BG, Dostal J. National health and nutrition examination survey: plan and operations, 1999-2010. Vital Health Stat 1. Aug 2013(56):1-37.
Moshfegh AJ, Rhodes DG, Baer DJ, et al. The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. Am J Clin Nutr. Aug 2008;88(2):324-332. doi:10.1093/ajcn/88.2.324
Li B, Chen L, Hu X, et al. Association of Serum Uric Acid With All-Cause and Cardiovascular Mortality in Diabetes. Diabetes Care. Feb 1 2023;46(2):425-433. doi:10.2337/dc22-1339
Brauchla M, Juan W, Story J, Kranz S. Sources of Dietary Fiber and the Association of Fiber Intake with Childhood Obesity Risk (in 2-18 Year Olds) and Diabetes Risk of Adolescents 12-18 Year Olds: NHANES 2003-2006. J Nutr Metab. 2012;2012:736258. doi:10.1155/2012/736258
Beddhu S, Baird BC, Zitterkoph J, Neilson J, Greene T. Physical activity and mortality in chronic kidney disease (NHANES III). Clin J Am Soc Nephrol. Dec 2009;4(12):1901-1906.
National Health and Nutrition Examination Survey (NHANES)-MEC-Laboratory-Procedures-Manual-2020. doi:10.2215/CJN.01970309
Freedman ND, Park Y, Abnet CC, Hollenbeck AR, Sinha R. Association of coffee drinking with total and cause-specific mortality. N Engl J Med. May 17 2012;366(20):1891-1904. doi:10.1056/NEJMoa1112010
Ding M, Satija A, Bhupathiraju SN, et al. Association of Coffee Consumption With Total and Cause-Specific Mortality in 3 Large Prospective Cohorts. Circulation. Dec 15 2015;132(24):2305-2315. doi:10.1161/CIRCULATIONAHA.115.017341
Tsujimoto T, Kajio H, Sugiyama T. Association Between Caffeine Intake and All-Cause and Cause-Specific Mortality: A Population-Based Prospective Cohort Study. Mayo Clin Proc. Aug 2017;92(8):1190-1202. doi:10.1016/j.mayocp.2017.03.010
Reis CEG, Dorea JG, da Costa THM. Effects of coffee consumption on glucose metabolism: A systematic review of clinical trials. J Tradit Complement Med. Jul 2019;9(3):184-191. doi:10.1016/j.jtcme.2018.01.001
Gavrieli A, Karfopoulou E, Kardatou E, et al. Effect of different amounts of coffee on dietary intake and appetite of normal-weight and overweight/obese individuals. Obesity (Silver Spring). Jun 2013;21(6):1127-1132. doi:10.1002/oby.20190
Paiva C, Beserra B, Reis C, Dorea JG, Da Costa T, Amato AA. Consumption of coffee or caffeine and serum concentration of inflammatory markers: A systematic review. Crit Rev Food Sci Nutr. 2019;59(4):652-663. doi:10.1080/10408398.2017.1386159
Yashin A, Yashin Y, Wang JY, Nemzer B. Antioxidant and Antiradical Activity of Coffee. Antioxidants (Basel). Oct 15 2013;2(4):230-245. doi:10.3390/antiox2040230
Larsson SC, Woolf B, Gill D. Appraisal of the causal effect of plasma caffeine on adiposity, type 2 diabetes, and cardiovascular disease: two sample mendelian randomisation study. BMJ Med. 2023;2(1):1-8. doi:10.1136/bmjmed-2022-000335
Ding M, Bhupathiraju SN, Satija A, van Dam RM, Hu FB. Long-term coffee consumption and risk of cardiovascular disease: a systematic review and a dose-response meta-analysis of prospective cohort studies. Circulation. Feb 11 2014;129(6):643-659. doi:10.1161/CIRCULATIONAHA.113.005925
Safe S, Kothari J, Hailemariam A, Upadhyay S, Davidson LA, Chapkin RS. Health Benefits of Coffee Consumption for Cancer and Other Diseases and Mechanisms of Action. Int J Mol Sci. Jan 31 2023;24(3). doi:10.3390/ijms24032706
Yamauchi R, Kobayashi M, Matsuda Y, et al. Coffee and caffeine ameliorate hyperglycemia, fatty liver, and inflammatory adipocytokine expression in spontaneously diabetic KK-Ay mice. J Agric Food Chem. May 12 2010;58(9):5597-5603. doi:10.1021/jf904062c
Chen L, Wang XJ, Chen JX, et al. Caffeine ameliorates the metabolic syndrome in diet-induced obese mice through regulating the gut microbiota and serum metabolism. Diabetol Metab Syndr. Mar 8 2023;15(1):37. doi:10.1186/s13098-023-00993-3
Zhang W, Lopez-Garcia E, Li TY, Hu FB, van Dam RM. Coffee consumption and risk of cardiovascular diseases and all-cause mortality among men with type 2 diabetes. Diabetes Care. Jun 2009;32(6):1043-1045. doi:10.2337/dc08-2251
Zhang WL, Lopez-Garcia E, Li TY, Hu FB, van Dam RM. Coffee consumption and risk of cardiovascular events and all-cause mortality among women with type 2 diabetes. Diabetologia. May 2009;52(5):810-817. doi:10.1007/s00125-009-1311-1
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. Nov 22 2017;359:j5024. doi:10.1136/bmj.j5024

No competing interests reported.

SupplementaryOnlineContent9.10.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Associations between Caffeine Intake and the Risk of Mortality in Adults with Prediabetes: Evidence from NHANES 1999-2018

Status:

Version 1

Abstract

Background

Objectives

Methods

Results

Conclusions

Figures

1. Introduction

2. Methods

2.1 Study Population

2.2 Ascertainment of Prediabetes

2.3 Assessment of Caffeine Intake

2.4 Ascertainment of Mortality

2.5 Assessment of Covariates

2.6 Statistical Analysis

3. Results

4. Discussion

5. Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1