https://orcid.org/0000-0002-4690-7105
Figures
Abstract
Background
Despite the substantial evidence of the relationship between diet and mortality, the role of beverage consumption patterns is not well known. The aim of this study was to assess the association of the adherence to a Healthy Beverage Score (HBS) and all-cause mortality in a representative sample of the Spanish adult population.
Methods and findings
We conducted an observational cohort study using data from the Study on Nutrition and Cardiovascular Risk in Spain (ENRICA), which included 12,161 community-dwelling individuals aged ≥18 years recruited in 2008 to 2010 and followed until January 2022. At baseline, food consumption was collected using a validated diet history. The HBS consists of 7 items, each of which is scored from 1 to 4 (highest adherence). The HBS ranges from 7 to 28 points with a higher score representing a healthier pattern. Adherence was assigned as a higher consumption of low-fat milk, and coffee and tea, a lower consumption of whole-fat milk, no consumption of fruit juice, artificially sweetened beverages, or sugar-sweetened beverages, and no or moderate consumption of alcohol. Total mortality was ascertained by linkage to the Spanish National Death Index. Statistical analyses were performed with Cox models and adjusted for the main confounders, including sociodemographic, lifestyle, dietary variables, and morbidity.
After a mean follow-up of 12.5 years (SD: 1.7; range: 0.5 to 12.9), a total of 967 deaths occurred. For all-cause mortality, the fully adjusted hazard ratio (HR) for the highest versus lowest sex-specific quartiles of HBS was 0.72 (95% confidence interval [0.57, 0.91], p linear-trend = 0.015), corresponding to an 8.3% reduction in the absolute risk of death. A linear relationship between the risk of death and the adherence to the HBS was observed using restricted cubic splines. The results were robust to sensitivity analyses. The main limitation was that repeated measurements on beverage consumption were not available and beverage consumption could have changed during follow-up.
Author summary
Why was this study done?
- Most dietary patterns focus solely on solid foods, and the role of beverages as a whole has received little attention.
- Our aim was to assess the relationship between a Healthy Beverage Score (HBS) and mortality in a representative sample of community-dwelling individuals from Spain.
- Our hypothesis was that high adherence to the HBS would be associated with lower mortality.
What did the researchers do and find?
- We included a representative sample of 12,161 adults (18 years and older) from Spain who were recruited in 2008 to 2010 and followed up until 2022. A total of 967 deaths occurred.
- Participants were categorized according to their adherence to the HBS.
- A higher total score was achieved with a higher consumption of low-fat milk, and coffee and tea, no consumption of whole-fat milk, fruit juice, artificially sweetened beverages, sugar-sweetened beverages, and no consumption or moderate consumption of alcohol.
- Each HBS item scored from 1 (minimum adherence) to 4 points (maximum adherence) and the HBS ranged from 7 to 28 points. The higher the HBS, the healthier.
- When comparing extreme categories, higher adherence to the HBS was associated with lower all-cause mortality in the Spanish adult population, with an 8.3% reduction in the absolute risk of death.
Citation: Rodríguez-Ayala M, Donat-Vargas C, Moreno-Franco B, Mérida DM, Ramón Banegas J, Rodríguez-Artalejo F, et al. (2024) Association of a healthy beverage score with total mortality in the adult population of Spain: A nationwide cohort study. PLoS Med 21(1): e1004337. https://doi.org/10.1371/journal.pmed.1004337
Received: July 12, 2023; Accepted: December 21, 2023; Published: January 23, 2024
Copyright: © 2024 Rodríguez-Ayala et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data are freely available upon request by contacting Esther López-García at the Department of Preventive Medicine and Public Health, Faculty of Medicine, Universidad Autónoma de Madrid (UAM)-IdiPaz, CIBERESP (CIBER of Epidemiology and Public Health), 28029, Madrid, Spain. E-mail address: esther.lopez@uam.es. UAM website: https://www.uam.es/ss/Satellite/Medicina/es/1242658444664/subhome/Departamento_de_Medicina_Preventiva_y_Salud_Publica_y_Microbiologia.htm.
Funding: This work was supported by FIS grants 17/1709, and 20/144 from the Carlos III Health Institute, the Secretary of R+D+I, and the European Regional Development Fund/European Social Fund (to P.G-C); by the National Plan on Drugs grant 2020/17, Spanish Ministry of Health, Spain (to F.R-A); by the FACINGLCOVID-CM project, Comunidad de Madrid and European Regional Development Fund (ERDF), European Union (to F.R-A); and by the REACT EU Program, Comunidad de Madrid and European Regional Development Fund (ERDF), European Union (to F.R-A). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: BMI, body mass index; CI, confidence interval; ENRICA, Study on Nutrition and Cardiovascular Risk in Spain; HBS, Healthy Beverage Score; HR, hazard ratio; METs-hour/week, metabolic equivalents in hours per week; SD, standard deviation
Introduction
The influence of unhealthy dietary factors on adverse outcomes, including premature mortality, is a public health concern [1]. Thus, the association between diet and mortality has been examined by using several approaches such as analyzing individual nutrients, food and, more recently, assessing dietary patterns and indexes [2–4]. Most of the indexes include mainly solid food (e.g., meat, poultry, fish, as well as fruit and vegetables) [5], although some of them also comprise beverages [6].
The mechanisms by which beverages influence health are complex and are not only based on the nutritional quality of their components (such as energy provided, macronutrients, fiber, minerals, and vitamins) [7], but also rely on other factors such as satiety mechanisms and factors that affect the assimilation of beverages such as rapid gastric emptying and intestinal absorption [8]. Moreover, the addition of artificial sweeteners influences mortality [9]. On the other hand, some beverages are also an important source of other additives (e.g., phosphates) as well as contaminants from packaging or processing (e.g., organophosphate esters, phthalates) [10].
In 2015, a healthy beverage index, based on commonly consumed drinks, was developed to evaluate the role of beverage quality on cardiometabolic risk in adult Americans. A low adherence to this index was associated with several detrimental cardiometabolic markers [11]. Consistent results were obtained in another US study where the association between the adherence to a healthy beverage pattern and total mortality was evaluated. Data were obtained from a cohort of 2,283 adults, aged ≥21 years, with a previous diagnosis of mild to moderate chronic renal insufficiency. A healthier beverage index was inversely associated with the progression of chronic kidney disease and all-cause mortality [12].
Although the association of specific beverages has been studied previously, the role of a healthy beverage index and its association with mortality has not been assessed in the general population yet. We hypothesized that higher adherence to a 7-item Healthy Beverage Score (HBS), previously proposed by Hu and colleagues [12] and adapted to the Spanish beverage consumption, could be associated with lower mortality. Therefore, the aim of this study is to assess the association between the HBS and all-cause mortality in a representative cohort of Spanish adults.
Methods
Study design and participants
Data were obtained from the Study on Nutrition and Cardiovascular Risk in Spain (ENRICA) whose methods have been reported elsewhere [13]. In brief, 13,105 individuals aged ≥18 years were recruited from 2008 to 2010. A stratified cluster sampling based on the census sections of Spain was performed to guarantee the representativeness of the sample. Sample weights were based on the size of municipalities, sex, and age. Three sequential stages were followed for data collection. First, sociodemographic, lifestyle characteristics, and morbidity information was obtained through a telephone interview. Second, blood and urine samples were collected on a first home visit. Third, a physical examination and a face-to-face dietary history (DH-ENRICA) were completed during a second home visit. The response rate was 51% and the main reasons for non-participation were refusal to provide a blood sample (51.7%) and not being interested in the study (37.8%).
From the initial sample (13,105 individuals), 944 participants were excluded: 60 (0.5%) without information on diet and 884 (6.8%) with implausible values for total energy intake (<800 kcal/day or >5,000 kcal/day in males; <500 kcal/day or >4,000 kcal/day in females). Therefore, a total of 12,161 participants were included in the analysis (S1 Fig).
The Clinical Research Ethics Committee of La Paz University Hospital in Madrid provided ethical approval. All participants from the ENRICA Study gave written informed consent after explaining the details of the study.
Study variables
Dietary history.
Information on diet was obtained through a computerized dietary history (DH-ENRICA), conducted by trained and certified nonmedical interviewers. The DH-ENRICA collected information on 861 items of food, with 82 beverages included. Participants informed about all items of food and beverages consumed at least once every 2 weeks in the previous year. Food consumed during weekdays and weekends were considered. A total of 127 sets of digitalized photos, household measurements, as well as the usual proportion sizes of food from typical Spanish recipes were used to estimate portion sizes in grams per day. Regarding beverages, a total of 14 digitized photos and 23 household measurements were used to later estimate beverage consumption in milliliters per day. In addition, for alcoholic beverages, the consumption of ethanol in grams per day was calculated using Spanish food composition tables [14]. The validity correlation coefficients in HD-ENRICA for beverages were: 0.71 for coffee, 0.69 for milk, 0.40 for soft drinks, and 0.64 for alcoholic beverages [15].
The Healthy Beverage Score (HBS).
A Healthy Beverage Score (HBS) was previously described by Hu and colleagues [12]. Based on the HBS, we built a 7-item HBS modifying its cut-off points to fit with the beverage consumption of a representative sample of the Spanish adult population. Each item of the HBS scored from 1 (minimal adherence) to 4 points (maximal adherence) based on sex-specific categories of consumption. Thus, the HBS ranged from 7 (low adherence) to 28 points (high adherence). The higher the HBS, the healthier the pattern. Items were grouped in 2 main components: adequacy and moderation (Table 1). Two beverages were considered as adequacy components: low-fat milk as well as coffee and tea consumption. For these 2 components, the higher the score, the healthier the pattern. No low-fat milk consumption scored 1, while the remaining sample was divided into tertiles among consumers; coffee and tea consumption was grouped into quartiles. Five items were included as moderation components: whole-fat milk, fruit juice, artificially sweetened beverages, sugar-sweetened beverages, and alcohol. For these 5 items, the higher the consumption, the lower the score, with a specific classification for alcohol consumption. Whole-fat milk and sugar-sweetened beverages scored 4 for no consumption and the remaining sample was divided into tertiles among consumers. Fruit juice and artificially sweetened beverages consumption scored 4 for no consumption and 1 for any consumption. The scoring of fruit juices and artificially sweetened beverages was decided on the basis of the lack of a wide range of consumption, and to maintain the relative weight of these items in the score, as previously described by Hu and colleagues [12]. Finally, for alcohol consumption, participants with no consumption or moderate drinkers (<40 g/day for males and <24 g/day for females) scored 4, and heavy drinkers (≥40 g/day for males and ≥24 g/day for females) scored 1.
Mortality assessment.
All-cause mortality was ascertained through a computerized linkage with the Spanish National Death Index. Participants were followed from baseline in 2008 to 2010 to January 31, 2022. Follow-up was censored at the date of death or at the end of follow-up, whichever occurred first.
Confounders.
Participants provided information regarding age, sex, educational level, and smoking status which were obtained through the computer-assisted telephone interview performed at baseline. On the second home visit and following standardized procedures, blood pressure, and weight and height were measured. Body mass index (BMI) was calculated as weight divided by the square of the height in meters (kg/m2). Leisure time and household physical activity were evaluated with the EPIC short questionnaire, collecting information on 17 activities. Then, each activity was multiplied by their respective energy expenditure rate in metabolic equivalents in hours per week (METs-hour/week) [16] and total energy expenditure was obtained by summing up all activities. Hours spent watching television were used to account for sedentary activities. In order to control for good dietary quality, analyses were adjusted for total energy intake, fiber, fruit and vegetable consumption, as well as the Mediterranean Diet Adherence defined by Trichopoulou and colleagues [17] without including alcohol. Blood samples collected on the first home visit were centrally analyzed in the CORE laboratory of La Paz University Hospital in Madrid. A colorimetric enzymatic method with lipase and glycerol kinase (for triglycerides) and a colorimetric enzymatic method with cholesterol-oxidase, esterase, and peroxidase (for cholesterol) were used. To define hypertriglyceridemia, we used a threshold of ≥150 mg/d in fasting plasma triglycerides levels, and for hypercholesterolemia, a fasting plasma total cholesterol level of ≥200 mg/dL or prescribed lipid-lowering medications. Hypertension was defined as ≥140/90 mmHg or taking antihypertensive medication. Analyses were also controlled for the number of chronic conditions (chronic obstructive pulmonary disease, coronary heart disease, stroke, heart failure, osteoarthritis, cancer, depression diagnosed by a physician, and diabetes), as well as the number of prescribed medications to consider prevalent morbidity.
Independent variables with missing values were imputed by using multiple imputation [18]: educational level (<1%), smoking status (<1%), BMI (1.5%), number of television hours (<1%), hypertriglyceridemia (<1%), hypercholesterolemia (<1%), and high blood pressure (1.1%). The validity of imputed data was examined against analyses performed with variables that contained full information.
Statistical analysis
Across sex-specific quartiles of adherence to the HBS, age-adjusted baseline characteristics of participants were computed using marginals. Age-adjusted means for continuous variables and age-adjusted proportions (%) for categorical variables were provided. To estimate hazard ratios (HR) and their 95% confidence intervals (CIs), Cox proportional hazards regression models were built and age was considered as the underlying time metric. The survey command was applied to account for the complex sampling design. The lowest category of adherence to the HBS was used as reference.
Three sequential Cox models were used. Model 1 was an unadjusted model. Model 2 was additionally adjusted for age, sex, educational level, smoking status, ex-drinker status, BMI, physical activity in leisure time, total energy intake, fruit and vegetable consumption, total fiber intake, hypertriglyceridemia, hypercholesterolemia, hypertension, number of self-reported chronic conditions, and number of medications. Finally, Model 3 was adjusted for the Mediterranean index by Trichopoulou excluding alcohol (maximum score = 8), but excluding fruit, vegetable, and fiber consumption. Schoenfeld residuals were plotted against time to assess proportional hazards assumptions and visually no violations were found. To test for linear trend, categories of the HBS were modeled as a continuous variable. The dose–response relationship was assessed by using restricted cubic splines with 3 knots (at the 10th, 50th, and 90th percentiles). Interactions between the HBS and age (<65 years versus ≥65 years), sex (male versus female), BMI (<30 kg/m2 versus ≥30 kg/m2), physical activity (≤median 61.5 METs-hour/week versus >median 61.5 METs-hour/week), vegetable consumption (≤median 183.5 g/d versus >median 183.5 g/d), adherence to the Mediterranean diet without including alcohol (≤median 4 versus >median 4), and the prevalence of chronic conditions (yes/no) were tested by including multiplicative terms in Model 3. Sensitivity analyses were conducted excluding deaths in the first 3 years of follow-up to account for the effect of subclinical conditions at baseline. Also, individual items of the HBS were assessed according to Model 3 plus adjustment for the remaining items that are part of the score.
This was a preplanned study and data analysis was conducted according to a prespecified plan (S1 Text). The HBS was previously used to assess the association between adherence to HBS and age-related frailty in a sample of Spanish older adults [19]. A minor modification was introduced in alcohol consumption classification. For older adults, moderate alcohol consumption was considered the healthy option due to their high cardiovascular risk [19]. However, in the current analysis, which involves the adult general population (aged ≥18 years) with lower cardiovascular risk, the category of alcohol considered healthy was “no consumption or moderate consumption.”
This study was reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 STROBE Checklist).
A two-tailed p value less than 0.05 was considered statistically significant. All analyses were performed with Stata, version 17.0 (StataCorp, College Station, Texas, United States of America).
Results
The median age of participants (N = 12,161) was 46 years old (interquartile range 35 to 61) and 52.6% were females. Compared with those in quartile 1 (less healthy) of adherence to the HBS, participants in quartile 4 (healthier) were older, more frequently females, with a higher level of education and with a less sedentary lifestyle, and were more physically active. Also, those in quartile 4 had lower energy intake, consumed more fiber, fruit and vegetables, showed a higher adherence to the Mediterranean diet, and had more frequently hypercholesterolemia (Table 2). Sex-specific cut-off points for individual items of the HBS are shown in S1 Table.
In accordance with the rules for the construction of the HBS, compared with those in quartile 1 (less healthy), participants in quartile 4 (healthier) consumed more low-fat milk, coffee and tea, and alcohol, but consumed less whole-fat milk, fruit juice, artificially sweetened beverages, and sugar-sweetened beverages (Table 3).
After a mean follow-up of 12.5 years (SD: 1.7; range: 0.5 to 12.9) and 151,459 person-years of follow-up, a total of 967 deaths occurred. The HR for all-cause mortality when comparing extreme quartiles of the adherence to the HBS was 0.72 (95% CI, 0.57 to 0.91, p for linear trend = 0.015) in the fully adjusted model (Table 4). The decrease in absolute risk of death was 4.3% for quartile 2, 6.3% for quartile 3, and 8.3% for quartile 4. No significant interactions were found for age, sex, BMI, physical activity, vegetable consumption, or adherence to the Mediterranean diet (without including alcohol). However, a statistically significant interaction was found when stratifying for the presence of at least 1 chronic condition (p = 0.030). Among those with at least 1 chronic condition, higher adherence to the HBS was associated with lower mortality. No association was observed among those with no chronic conditions (Table 5). After excluding the first 3 years of follow-up, the inverse association between the adherence to the HBS and total mortality remained similar (S2 Table). When assessing dose–response, a linear relationship was observed using restricted cubic splines (p value for non-linearity = 0.010) (Fig 1).
Lines are restricted cubic splines, showing the dose–response association of the HBS with mortality. The solid line represents the HR, and the dashed lines indicate the lower and upper 95% CIs. The knots were located at the 10th, 50th, and 90th percentiles (corresponding to HBS scores 17, 22, and 25, respectively); p for non-linearity = 0.010. Adjusted as in Model 3. Data were adjusted for age (years, continuous), sex (male, female), educational level (primary or less, secondary, university), smoking (non-smoker, former smoker, current smoker), ex-drinker (yes/no), BMI (<25, ≥25 and ≤30, >30 kg/m2), time watching TV (hours, continuous), physical activity (METs-hour/week, continuous), energy intake (kcal/day, continuous), hypertriglyceridemia (yes/no), hypercholesterolemia (yes/no), hypertension (yes/no), number of chronic conditions (0, 1, and ≥2), number of medications (0, 1–3, >3), and adherence to the Mediterranean diet without including alcohol (maximum score = 8). Age was the underlying time metric.
When individual HBS items were analyzed using Model 3 for adjustment, a higher consumption of coffee and tea, and no consumption of fruit juices and artificially sweetened beverages contributed most to the association with lower all-cause mortality (Fig 2). Unadjusted results are also shown (S3 Fig).
Adjusted as in Model 3. Data were adjusted for age (years, continuous), sex (male, female), educational level (primary or less, secondary, university), smoking (non-smoker, former smoker, current smoker), ex-drinker (yes/no), BMI (<25, ≥25 and ≤30, >30 kg/m2), time watching TV (hours, continuous), physical activity (METs-hour/week, continuous), energy intake (kcal/day, continuous), hypertriglyceridemia (yes/no), hypercholesterolemia (yes/no), hypertension (yes/no), number of chronic conditions (0, 1, and ≥2), number of medications (0, 1–3, >3), adherence to the Mediterranean diet without including alcohol (maximum score = 8), and for the rest of items of the HBS (as appropriate). Age was the underlying time metric. CI, confidence interval; HBS, Healthy Beverage Score; HR, hazard ratio.
Discussion
In this large population-based study of Spanish adults with a mean follow-up of 12.5 years, a higher adherence to the HBS was inversely associated with total mortality, after adjusting for potential confounders. Those with higher adherence to the HBS had an 8.3% reduction in the absolute risk of death compared to those with lower adherence. The association was linear and robust. It may also be of particular interest to people with preexisting chronic conditions, as they had lower mortality, although these findings need to be confirmed in future research.
Regarding to items of the HBS, 2 recent prospective studies performed in Spain found an inverse association between coffee consumption and all-cause mortality [20,21]. Results from the EPIC study (with 500,000 participants from 10 European countries) [22] and from the UK Biobank study were also similar to the findings in this study [23]. Our results are also in line with meta-analyses comprising cross-sectional studies, longitudinal cohorts, as well as interventional studies [24–26]. The beneficial effect of coffee might rely, among others, on the antioxidant and anti-inflammatory activity exhibited by its bioactive components, mainly melanoidins, chlorogenic acids, and caffeine [27]. These compounds reduce oxidative stress and inflammation [28], enhance endothelial function [29], and counteract carcinogenesis on in vitro studies [30]. Coffee also increases the metabolic rate [31], improves the glucose metabolism [32], and lowers long-term blood pressure [33]. Moreover, coffee could reduce mortality even in those with impaired caffeine metabolism [34] and independently to the addition of sweeteners [35]. However, high coffee consumption has been associated with an increase in serum levels of total cholesterol, LDL-cholesterol, and triglycerides [36]. On the other hand, evidence suggests that coffee consumption above 4 cups/day is not associated with further lower mortality [25].
Spain is included among the European countries with the lowest tea consumption [37] and we did not find studies that evaluated its relationship with mortality among Spanish adults. Therefore, it is unlikely that tea consumption accounts for our results. However, in literature, both all-cause and cardiovascular mortality were reduced among tea consumers [38].
The effect of milk on health has been widely studied due to its fatty acid composition [39]. Two recent cohort studies showed that low-fat milk consumption was associated with lower all-cause mortality when compared to whole-fat milk consumption [40,41]. Whole-fat milk has a higher content of saturated fats that has been related to an increase in LDL-cholesterol and atherosclerosis [42]. As a result, whole-fat milk consumption could be particularly harmful among individuals with known cardiovascular risk. However, in a clinical trial among normocholesterolemic individuals, whole-fat milk consumption showed no impairment in lipid profile nor in glucose-insulin metabolism when compared to low-fat milk consumption [43]. It is of note that, milk could modulate satiety mechanisms [44] and also has several components with potential beneficial effects, such as caseins with antioxidant properties [45]. Milk is also rich in minerals, vitamins, and other bioactive compounds involved in anti-inflammatory and immune regulation [46]. A recent meta-analysis showed that whole-fat milk consumption was associated with a higher risk of all-cause, cardiovascular disease and cancer mortality; however, low-fat milk showed a protective but nonsignificant association [47].
Results on fruit juice consumption and mortality mostly depend on the distinction between processed or fresh fruit juice [48]. A recent meta-analysis showed that processed fruit juice consumption was associated with a higher risk of type 2 diabetes and total mortality [49]. Evidence on fresh fruit juice consumption and health, however, is insufficient to draw conclusions. Results of a cohort study from the US with 13,440 participants showed that a higher consumption of 100% fruit juice was associated with a higher mortality [50]. Conversely, a study with 198,285 individuals from the UK found a positive association between sugar-sweetened beverages and mortality, but not for 100% fruit juice consumption [51]. Similarly, a recent meta-analysis of prospective cohorts concluded that there was no association between 100% fruit juice consumption and all-cause mortality [52]. Several studies have also found that, compared with 100% fruit juice, sugary or processed fruit beverages produce harmful glucose levels after ingestion, mainly due to the higher content of free sugars [53]. In our study, bottled, sweetened, as well as fresh fruit juices were analyzed together as a unique item because of their rapid absorption [8] and similar effect on postprandial glucose levels [54]. Fructose intake, particularly from sugar-sweetened beverages at any dose, or from fruit juice at higher doses, contributes a rapid extra dietary energy source that could explain its detrimental effect on health [55]. However, food-based dietary guidelines from various countries from Europe Union, including Spain, consent to replace occasionally 1 daily portion of fruit with fresh fruit juice [56,57].
In order to lower calorie intake and control body weight, artificially sweetened beverages could be adequate short-term substitutes [58]. However, when considering the long-term influence of artificially sweetened beverage consumption, several studies have found associations with higher obesity, hypertension, type 2 diabetes, stroke, cardiovascular disease incidence and mortality, and all-cause mortality [59–61]. Since these beverages contain few to no calories nor sugars [51], some investigations have related them with weight gain as a result of an increased consumption of sweet food due to a greater affinity for sweet flavors or the perception of eating fewer calories [62]. In addition, their flavoring components have been associated with the formation of advanced glycation end-products [63], which are involved in the development of metabolic diseases [64]. Moreover, some sweeteners such as sucralose and saccharin could induce glucose intolerance and alterations in gut microbiota [65] that are linked to obesity [66].
In literature, a low to moderate alcohol consumption is related to a reduction in all-cause mortality [67]. Biological explanations for this protective role on health are based on lipid regulation, insulin response, and endothelial function [68] resulting from the modulation of some anti-inflammatory biomarkers [69]. However, at high doses, alcohol is detrimental to cardiovascular health and is related to several types of cancer [70]. A harmful alcohol consumption is associated to neurodegenerative processes [71], microbial dysbiosis [70], and an increased intestinal permeability that leads to a permanent hepatic exposure to bacterial translocation, oxidative stress, and other inflammatory components [72]. In addition, alcohol use could result in hepatic steatosis and de novo lipogenesis, and also could reduce the utilization of lipids [73]. Lastly, alcohol may injure myocardium with potential cardiomyopathy and heart failure [74], and increase the risk of hypertension [75]. On the other hand, the beneficial association of alcohol consumption with mortality found in some studies may rely on abstinence bias, insufficient adjustment for covariates, or consumption changes due to disease detection [76,77]. However, recent studies from an epigenetic perspective have proposed that alcohol at restricted doses could be particularly beneficial in older adults based on changes in alcohol metabolism related to age [78]. In this regard, a meta-analysis for the Global Burden of Disease Study has proposed a change from sex-specific to age-specific recommendations on alcohol consumption [79]. Then, low alcohol consumption could be beneficial among older adults, but not for younger adults. For our analyses, we considered that being a heavy drinker was harmful because of its well-established association.
In a previous study, a 10-item Healthy Beverage Index was constructed using an a priori approach based on the US recommendations for beverage consumption [11]. Similar to this index, Hu and colleagues described the HBS as a more suitable score for use in large epidemiological studies [12]. The HBS excluded water consumption as well as 2 items on total energy from beverages and calculations of daily fluid intake. In our study, the same scoring weights (from 1 to 4) were maintained for all items in the HBS. However, in contrast to Hu and colleagues, we considered both no alcohol consumption and moderate alcohol consumption as healthy. Additionally, we also modified the HBS cut-off points of the items to fit with the beverage consumption of the Spanish adult population. As a result, the HBS used in our study retained the same items as originally described by Hu and colleagues, as well as the relative weight of the items.
The use of the HBS as an overall measure of beverage consumption has several advantages. First, the use of this score overcomes the limitations of analyses of relationships between individual beverages and diseases, as beverage consumption may be correlated, and an increase in consumption of one beverage may be associated with a decrease in the others. Secondly, the HBS could be a complementary tool for assessing adherence to dietary patterns that include only solid foods, in order to assess dietary quality as a whole. Thus, the HBS could serve as a simple and rapid screener to obtain information on the quality of beverage consumption from the general population, similar to other indexes used to assess adherence to certain diets, such as the Mediterranean diet. Also, the use of this 7-item pattern may be an optimal choice when dealing with patients in time-constrained clinical settings. Finally, the HBS includes items on commonly consumed beverages and could be easily adapted to other populations with only minor modifications to account for their specific beverage consumption.
We have used the HBS in the general population and caution should be exercised in deriving beverage consumption recommendations from this score in specific populations, especially those with restricted fluid requirements, long-term liquid diets, and other preexisting conditions involving fluid consumption. Further studies are also needed in specific population subgroups. In addition, a future study could consider intercorrelations and specific population-based patterns of beverage consumption using an a posteriori approach.
This study has some limitations. First, when measuring diet, non-differential misclassification of the exposure is always possible, in general, resulting in an underestimation of the associations found. Second, no information concerning behavioral changes or repeated measurements on beverage consumption were available, and beverage consumption could have changed during follow-up. Third, water consumption was not available in this study. Water is universally recommended as a safe beverage and as the main source of hydration. As water does not provide energy, macronutrients or micronutrients, its consumption is considered free for the general population.
There are also some strengths. To our knowledge, this is the first examination on the relation between a healthy beverage score and all-cause mortality among the Spanish adult population. In addition, we used a dietary history that allowed us to collect information on beverages with validity and reproducibility in a Spanish population. Also, the national vital statistics records, accessed through linkage to the Spanish National Death Index, ensured an extensive follow-up of the cohort for mortality assessment. Finally, several confounders were considered in more adjusted models.
In conclusion, in this representative study of the Spanish adult population, higher adherence to the HBS was associated with a reduction in total mortality. As the consumption of a healthy solid diet should be encouraged, adherence to a healthy beverage consumption pattern may also play an important role in the prevention of premature mortality as part of public health nutrition prevention strategies.
Supporting information
S1 Text. Prespecified analysis plan and modifications.
https://doi.org/10.1371/journal.pmed.1004337.s002
(DOCX)
S1 Table. Sex-specific cut-off points for individual items of the Healthy Beverage Score (HBS) in the ENRICA Study (2008–2010) (N = 12,161).
https://doi.org/10.1371/journal.pmed.1004337.s003
(DOCX)
S2 Table. Mortality risk according to quartiles of the adherence to the Healthy Beverage Score (HBS) in the ENRICA Study from baseline (2008–2010) to January 2022 (N = 12,161) excluding the first 3 years of follow-up.
https://doi.org/10.1371/journal.pmed.1004337.s004
(DOCX)
S2 Fig. Unadjusted restricted cubic splines of the association of the Healthy Beverage Score (HBS) with mortality risk in the ENRICA Study from baseline (2008-2010) to January 2022 (N = 12,161).
Lines are restricted cubic splines, showing the dose-response association of the Healthy Beverage Score (HBS) with mortality. The solid line represents the hazard ratio (HR), and the dashed lines indicate the lower and upper 95% confidence intervals. The knots were located at the 10th, 50th, and 90th percentiles (corresponding to HBS scores 17, 22 and 25, respectively). p for non-linearity = 0.003.
https://doi.org/10.1371/journal.pmed.1004337.s006
(TIF)
S3 Fig. Unadjusted mortality risk for individual items of the Healthy Beverage Score (HBS) when comparing extreme categories (quartile 4 vs. quartile 1) in the ENRICA Study from baseline (2008-2010) to January 2022 (N = 12,161).
HBS, Healthy Beverage Score; HR, hazard ratio; CI, confidence interval.
https://doi.org/10.1371/journal.pmed.1004337.s007
(TIF)
References
- 1. Afshin A, Sur PJ, Fay KA, Cornaby L, Ferrara G, Salama JS, et al. Health effects of dietary risks in 195 countries, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2019 May 11;393(10184):1958–72. pmid:30954305
- 2. Shan Z, Guo Y, Hu FB, Liu L, Qi Q. Association of Low-Carbohydrate and Low-Fat Diets With Mortality Among US Adults. JAMA Intern Med. 2020 Apr 1;180(4):513–23. pmid:31961383
- 3. Becerra-Tomás N, Blanco Mejía S, Viguiliouk E, Khan T, Kendall CWC, Kahleova H, et al. Mediterranean diet, cardiovascular disease and mortality in diabetes: A systematic review and meta-analysis of prospective cohort studies and randomized clinical trials. Crit Rev Food Sci Nutr. 2020 Apr 11;60(7):1207–27. pmid:30676058
- 4. Boushey C, Ard J, Bazzano L, Heymsfield S, Mayer-Davis E, Sabaté J, et al. Dietary Patterns and All-Cause Mortality: A Systematic Review. USDA Nutrition Evidence Systematic Review. 2020.
- 5. Kruseman M, Chatelan A, Farina E, Carrard I, Cela J, Guessous I, et al. Assessing Overall Diet Quality: Development and Evaluation of the Performance of a Short Self-Administrated Questionnaire SCASA. Nutrients. 2021 Feb 1;13(2):1–17.
- 6. García-Conesa MT, Philippou E, Pafilas C, Massaro M, Quarta S, Andrade V, et al. Exploring the Validity of the 14-Item Mediterranean Diet Adherence Screener (MEDAS): A Cross-National Study in Seven European Countries around the Mediterranean Region. Nutrients. 2020 Oct 1;12(10):1–18. pmid:32992649
- 7. Wagner S, Merkling T, Girerd N, Bozec E, Van den Berghe L, Hoge A, et al. Quality of Beverage Intake and Cardiometabolic and Kidney Outcomes: Insights From the STANISLAS Cohort. Front Nutr. 2022 Jan 7;7(8):1120. pmid:35071290
- 8. Leiper JB. Fate of ingested fluids: factors affecting gastric emptying and intestinal absorption of beverages in humans. Nutr Rev. 2015 Sep 1;73(suppl_2):57–72. pmid:26290292
- 9. Malik VS, Li Y, Pan A, De Koning L, Schernhammer E, Willett WC, et al. Long-term Consumption of Sugar-Sweetened and Artificially Sweetened Beverages and Risk of Mortality in US adults. Circulation. 2019 Apr 30;139(18):2113. pmid:30882235
- 10. Giuliani A, Zuccarini M, Cichelli A, Khan H, Reale M. Critical Review on the Presence of Phthalates in Food and Evidence of Their Biological Impact. Int J Environ Res Public Health. 2020 Aug 2;17(16):1–43.
- 11. Duffey KJ, Davy BM. The Healthy Beverage Index Is Associated with Reduced Cardiometabolic Risk in US Adults: A Preliminary Analysis. J Acad Nutr Diet. 2015 Oct 1;115(10):1682–1689.e2.
- 12. Hu EA, Anderson CAM, Crews DC, Mills KT, He J, Shou H, et al. A Healthy Beverage Score and Risk of Chronic Kidney Disease Progression, Incident Cardiovascular Disease, and All-Cause Mortality in the Chronic Renal Insufficiency Cohort. Curr Dev Nutr. 2020;4(6):nzaa088. pmid:32551412
- 13. Rodríguez-Artalejo F, Graciani A, Guallar-Castillón P, León-Muñoz LM, Zuluaga MC, López-García E, et al. Rationale and Methods of the Study on Nutrition and Cardiovascular Risk in Spain (ENRICA). Rev Esp Cardiol. 2011 Oct;64(10):876–82.
- 14.
Farrán A, Zamora R, Cervera P. Tablas de composición de alimentos del CESNID. Edicions Universitat de Barcelona, editor. Barcelona; 2003.
- 15. Guallar-Castillón P, Sagardui-Villamor J, Balboa-Castillo T, Sala-Vila A, Astolfi MJA, Pelous MDS, et al. Validity and reproducibility of a Spanish dietary history. PLoS ONE. 2014 Jan 20;9(1). pmid:24465878
- 16. Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett DR, Tudor-Locke C, et al. Compendium of physical activities: A second update of codes and MET values. Vol. 43, Medicine and Science in Sports and Exercise. Med Sci Sports Exerc. 2011;2011:1575–1581.
- 17. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean Diet and Survival in a Greek Population. N Engl J Med. 2003 Jun 26;348(26):2599–608. pmid:12826634
- 18.
Enders C. Applied Missing Data Analysis. The Guildford Press. 2010:46–48.
- 19. Dominguez LJ, Donat-Vargas C, Banegas JR, Barbagallo M, Rodríguez-Artalejo F, Guallar-Castillón P. Adherence to a Healthy Beverage Score Is Associated with Lower Frailty Risk in Older Adults. Nutrients. 2022 Sep 1;14(18):3861. pmid:36145237
- 20. Torres-Collado L, Compañ-Gabucio LM, González-Palacios S, Notario-Barandiaran L, Oncina-Cánovas A, Vioque J, et al. Coffee Consumption and All-Cause, Cardiovascular, and Cancer Mortality in an Adult Mediterranean Population. Nutrients. 2021 Apr 1;13(4):1241. pmid:33918797
- 21. Navarro AM, Martinez-Gonzalez M, Gea A, Grosso G, Martín-Moreno JM, Lopez-Garcia E, et al. Coffee consumption and total mortality in a Mediterranean prospective cohort. Am J Clin Nutr. 2018 Nov 1;108(5):1113–20. pmid:30475964
- 22. Gunter MJ, Murphy N, Cross AJ, Dossus L, Dartois L, Fagherazzi G, et al. Coffee drinking and mortality in 10 European countries: A multinational cohort study. Ann Intern Med. 2017 Aug 15;167(4):236–47. pmid:28693038
- 23. Simon J, Fung K, Raisi-Estabragh Z, Aung N, Khanji MY, Kolossváry M, et al. Light to moderate coffee consumption is associated with lower risk of death: a UK Biobank study. Eur J Prev Cardiol. 2022 May 6;29(6):982–91. pmid:35048949
- 24. 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;22(360):k194. pmid:29167102
- 25. Kim Y, Je Y, Giovannucci E. Coffee consumption and all-cause and cause-specific mortality: a meta-analysis by potential modifiers. Eur J Epidemiol. 2019 Aug 15;34(8):731–52. pmid:31055709
- 26. Li Q, Liu Y, Sun X, Yin Z, Li H, Cheng C, et al. Caffeinated and decaffeinated coffee consumption and risk of all-cause mortality: a dose–response meta-analysis of cohort studies. J Hum Nutr Diet. 2019 Jun 1;32(3):279–87. pmid:30786114
- 27. Ludwig IA, Clifford MN, Lean MEJ, Ashihara H, Crozier A. Coffee: Biochemistry and Potential Impact on Health. R Soc Chem. 2014;5:1695–1717. pmid:24671262
- 28. Lara-Guzmán OJ, Medina S, Álvarez R, Oger C, Durand T, Galano JM, et al. Oxylipin regulation by phenolic compounds from coffee beverage: Positive outcomes from a randomized controlled trial in healthy adults and macrophage derived foam cells. Free Radic Biol Med. 2020;20(160):604–617. pmid:32745768
- 29. Higashi Y. Coffee and Endothelial Function: A Coffee Paradox? Nutrients. 2019 Sep 1;11(9):2104. pmid:31487926
- 30. De Marco LM, Fischer S, Henle T. High molecular weight coffee melanoidins are inhibitors for matrix metalloproteases. J Agric Food Chem. 2011 Nov 9;59(21):11417–23. pmid:21961901
- 31. Acheson KJ, Zahorska-Markiewicz B, Pittet P, Anantharaman K, Jéquier E. Caffeine and coffee: their influence on metabolic rate and substrate utilization in normal weight and obese individuals. Am J Clin Nutr. 1980;33(5):989–997. pmid:7369170
- 32. Reis CEG, Dórea JG, da Costa THM. Effects of coffee consumption on glucose metabolism: A systematic review of clinical trials. J Tradit Complement Med. 2019 Jul 1;9(3):184. pmid:31193893
- 33. Rodríguez-Artalejo F, López-García E. Coffee Consumption and Cardiovascular Disease: A Condensed Review of Epidemiological Evidence and Mechanisms. J Agric Food Chem. 2018 May 30;66(21):5257–63. pmid:29276945
- 34. Loftfield E, Cornelis MC, Caporaso N, Yu K, Sinha R, Freedman N. Association of Coffee Drinking With Mortality by Genetic Variation in Caffeine Metabolism: Findings From the UK Biobank. JAMA Intern Med. 2018 Aug 1;178(8):1086–97. pmid:29971434
- 35. Liu D, Li Z-H, Shen D, Zhang P-D, Song W-Q, Zhang W-T, et al. Association of Sugar-Sweetened, Artificially Sweetened, and Unsweetened Coffee Consumption With All-Cause and Cause-Specific Mortality. Ann Intern Med. 2022 May 31;175(7):909–17.
- 36. Du Y, Lv Y, Zha W, Hong X, Luo Q. Effect of coffee consumption on dyslipidemia: A meta-analysis of randomized controlled trials. Nutr Metab Cardiovasc Dis. 2020 Nov 27;30(12):2159–70. pmid:33239163
- 37. Landais E, Moskal A, Mullee A, Nicolas G, Gunter MJ, Huybrechts I, et al. Coffee and Tea Consumption and the Contribution of Their Added Ingredients to Total Energy and Nutrient Intakes in 10 European Countries: Benchmark Data from the Late 1990s. Nutrients. 2018 Jun 5;10(6):725. pmid:29874819
- 38. Inoue-Choi M, Ramirez Y, Cornelis MC, de González AB, Freedman ND, Loftfield E. Tea Consumption and All-Cause and Cause-Specific Mortality in the UK Biobank A Prospective Cohort Study. Ann Intern Med. 2022 Sep 1;175(9):1201–11. pmid:36037472
- 39. Hanus O, Samkova E, Křížova L, Hasoňova L, Kala R. Role of Fatty Acids in Milk Fat and the Influence of Selected Factors on Their Variability—A Review. Mol A J Synth Chem Nat Prod Chem. 2018;23(7):1636. pmid:29973572
- 40. Wang S, Liu Y, Cai H, Li Y, Zhang X, Liu J, et al. Decreased risk of all-cause and heart-specific mortality is associated with low-fat or skimmed milk consumption compared with whole milk intake: A cohort study. Clin Nutr. 2021 Nov 1;40(11):5568–75. pmid:34656953
- 41. Xu X, Kabir A, Barr ML, Schutte AE. Different Types of Long-Term Milk Consumption and Mortality in Adults with Cardiovascular Disease: A Population-Based Study in 7236 Australian Adults over 8.4 Years. Nutrients. 2022 Feb 1;14(3):704. pmid:35277068
- 42. Jakobsen MU, Trolle E, Outzen M, Mejborn H, Grønberg MG, Lyndgaard CB, et al. Intake of dairy products and associations with major atherosclerotic cardiovascular diseases: a systematic review and meta-analysis of cohort studies. Sci Rep. 2021;11(1):1–28.
- 43. Engel S, Elhauge M, Tholstrup T. Effect of whole milk compared with skimmed milk on fasting blood lipids in healthy adults: a 3-week randomized crossover study. Eur J Clin Nutr. 2018;72(2):249–254. pmid:29229955
- 44. Sánchez-Moya T, Planes-Muñoz D, Frontela-Saseta C, Ros-Berruezo G, López-Nicolás R. Milk whey from different animal species stimulates the in vitro release of CCK and GLP-1 through a whole simulated intestinal digestion. Food Funct. 2020 Aug 19;11(8):7208–16. pmid:32756716
- 45. Khan IT, Bule M, Ullah R, Nadeem M, Asif S, Niaz K. The antioxidant components of milk and their role in processing, ripening, and storage: Functional food. Vet World. 2019;12(1):12. pmid:30936650
- 46. Ahvanooei MRR, Norouzian MA, Vahmani P. Beneficial Effects of Vitamins, Minerals, and Bioactive Peptides on Strengthening the Immune System Against COVID-19 and the Role of Cow’s Milk in the Supply of These Nutrients. Biol Trace Elem Res. 2021 Nov 27;200(11):4664–77. pmid:34837602
- 47. Naghshi S, Sadeghi O, Larijani B, Esmaillzadeh A. High vs. low-fat dairy and milk differently affects the risk of all-cause, CVD, and cancer death: A systematic review and dose-response meta-analysis of prospective cohort studies. Crit Rev Food Sci Nutr. 2022;62(13):3598–3612. pmid:33397132
- 48. Zhang Z, Zeng X, Li M, Zhang T, Li H, Yang H, et al. A Prospective Study of Fruit Juice Consumption and the Risk of Overall and Cardiovascular Disease Mortality. Nutrients. 2022 May 1;14(10):2127. pmid:35631268
- 49. Fardet A, Richonnet C, Mazur A. Association between consumption of fruit or processed fruit and chronic diseases and their risk factors: a systematic review of meta-analyses. Nutr Rev. 2019 Jun 1;77(6):376–87. pmid:30995309
- 50. Collin LJ, Judd S, Safford M, Vaccarino V, Welsh JA. Association of Sugary Beverage Consumption With Mortality Risk in US Adults: A Secondary Analysis of Data From the REGARDS Study. JAMA Netw Open. 2019 May 1;2(5):1–11. pmid:31099861
- 51. Anderson JJ, Gray SR, Welsh P, Mackay DF, Celis-Morales CA, Lyall DM, et al. The associations of sugar-sweetened, artificially sweetened and naturally sweet juices with all-cause mortality in 198,285 UK Biobank participants: A prospective cohort study. BMC Med. 2020 Apr 24;18(1):1–12.
- 52. Pan B, Ge L, Lai H, Wang Q, Zhang Q, Yin M, et al. Association of soft drink and 100% fruit juice consumption with all-cause mortality, cardiovascular diseases mortality, and cancer mortality: A systematic review and dose-response meta-analysis of prospective cohort studies. Crit Rev Food Sci Nutr. 2021;62(32):8908–8919. pmid:34121531
- 53. Pepin A, Stanhope KL, Imbeault P. Are Fruit Juices Healthier Than Sugar-Sweetened Beverages? A Review. Nutrients. 2019;11(5):1006. pmid:31052523
- 54. Alkutbe R, Redfern K, Jarvis M, Rees G. Nutrient Extraction Lowers Postprandial Glucose Response of Fruit in Adults with Obesity as well as Healthy Weight Adults. Nutrients. 2020 Mar 1;12(3):1–13. pmid:32183321
- 55. Semnani-Azad Z, Khan TA, Blanco Mejia S, De Souza RJ, Leiter LA, Kendall CWC, et al. Association of Major Food Sources of Fructose-Containing Sugars With Incident Metabolic Syndrome: A Systematic Review and Meta-analysis. JAMA Netw Open. 2020 Jul 1;3(7):e209993. pmid:32644139
- 56. European Commission. Health Promotion and Disease Prevention. Food-Based Dietary Guidelines in Europe—table 1 | Knowledge for policy [Internet]. European Comission. 2023 [cited 2023 Nov 21]. Available from: https://knowledge4policy.ec.europa.eu/health-promotion-knowledge-gateway/food-based-dietary-guidelines-europe-table-1_en.
- 57. Martínez Hernández JA, Cámara Hurtado M, Giner Pons RM, González Fandos E, López García E, Mañes Vinuesa J, et al. Report of the Scientific Committee of the Spanish Agency for Food Safety and Nutrition (AESAN) on the review and update of Dietary Recommendations for the Spanish population. Vol. 32. Madrid; 2020.
- 58. Ebbeling CB, Feldman HA, Steltz SK, Quinn NL, Robinson LM, Ludwig DS. Effects of sugar-sweetened, artificially sweetened, and unsweetened beverages on cardiometabolic risk factors, body composition, and sweet taste preference: a randomized controlled trial. J Am Heart Assoc. 2020 Aug 4;9(15):e015668. pmid:32696704
- 59. Qin P, Li Q, Zhao Y, Chen Q, Sun X, Liu Y, et al. Sugar and artificially sweetened beverages and risk of obesity, type 2 diabetes mellitus, hypertension, and all-cause mortality: a dose–response meta-analysis of prospective cohort studies. Eur J Epidemiol. 2020 Jul 1;35(7):655–71. pmid:32529512
- 60. Meng Y, Li S, Khan J, Dai Z, Li C, Hu X, et al. Sugar-and artificially sweetened beverages consumption linked to type 2 diabetes, cardiovascular diseases, and all-cause mortality: A systematic review and dose-response meta-analysis of prospective cohort studies. Nutrients. 2021 Aug 1;13(8):2636. pmid:34444794
- 61.
Rios-Leyvraz M, Montez J. Health effects of the use of non-sugar sweeteners: a systematic review and meta-analysis. [Internet]. 1st ed. World Health Organization. Geneva: World Health Organization; 2022 [cited 2023 Nov 21]. Available from: https://www.who.int/publications/i/item/9789240046429.
- 62. Borges MC, Louzada ML, de Sá TH, Laverty AA, Parra DC, Garzillo JMF, et al. Artificially Sweetened Beverages and the Response to the Global Obesity Crisis. PLoS Med. 2017 Jan 1;14(1):1–9.
- 63. Deo P, Chern C, Peake B, Tan SY. Non-nutritive sweeteners are in concomitant with the formation of endogenous and exogenous advanced glycation end-products. Int J Food Sci Nutr. 2020 Aug 17;71(6):706–14. pmid:31918589
- 64. Lima MTNS, Howsam M, Anton PM, Delayre-orthez C, Tessier FJ. Effect of Advanced Glycation End-Products and Excessive Calorie Intake on Diet-Induced Chronic Low-Grade Inflammation Biomarkers in Murine Models. Nutrients. 2021 Sep 2;13(9):3091. pmid:34578967
- 65. Ruiz-Ojeda FJ, Plaza-Díaz J, Sáez-Lara MJ, Gil A. Effects of Sweeteners on the Gut Microbiota: A Review of Experimental Studies and Clinical Trials. Adv Nutr. 2019 Jan 1;10(Suppl 1):S31. pmid:30721958
- 66. Liu BN, Liu XT, Liang ZH, Wang JH. Gut microbiota in obesity. World J Gastroenterol. 2021 Jul 7;27(25):3837. pmid:34321848
- 67. Xi B, Veeranki SP, Zhao M, Ma C, Yan Y, Mi J. Relationship of Alcohol Consumption to All-Cause, Cardiovascular, and Cancer-Related Mortality in U.S. Adults J Am Coll Cardiol. 2017 Aug 22;70(8):913–22.
- 68. Krenz M, Korthuis RJ. Moderate ethanol ingestion and cardiovascular protection: From epidemiologic associations to cellular mechanisms. J Mol Cell Cardiol. 2012 Jan 1;52(1):93–104. pmid:22041278
- 69. Chiva-Blanch G, Badimon L. Benefits and Risks of Moderate Alcohol Consumption on Cardiovascular Disease: Current Findings and Controversies. Nutrients. 2020 Jan 1;12(1):108.
- 70. Rumgay H, Murphy N, Ferrari P, Soerjomataram I. Alcohol and Cancer: Epidemiology and Biological Mechanisms. Nutrients. 2021 Sep 11;13(9):3173. pmid:34579050
- 71. Peng B, Yang Q, Joshi RB, Liu Y, Akbar M, Song BJ, et al. Role of Alcohol Drinking in Alzheimer’s Disease, Parkinson’s Disease, and Amyotrophic Lateral Sclerosis. Int J Mol Sci. 2020 Apr 1;21(7):1–21. pmid:32230811
- 72. Nicoletti A, Ponziani FR, Biolato M, Valenza V, Marrone G, Sganga G, et al. Intestinal permeability in the pathogenesis of liver damage: From non-alcoholic fatty liver disease to liver transplantation. World J Gastroenterol. 2019 Sep 9;25(33):4814. pmid:31543676
- 73. You M, Arteel GE. Effect of ethanol on lipid metabolism. J Hepatol. 2019 Feb 1;70(2):237. pmid:30658725
- 74. Gardner JD, Mouton AJ. Alcohol Effects on Cardiac Function. Compr Physiol. 2015 Apr 1;5(2):791–802.
- 75. Fuchs FD, Fuchs SC. The Effect of Alcohol on Blood Pressure and Hypertension. Curr Hypertens Rep. 2021 Nov 11;23(10):1–6.
- 76. Liu YT, Lee JH, Tsai MK, Wei JCC, Wen CP. The effects of modest drinking on life expectancy and mortality risks: a population-based cohort study. Sci Rep. 2022 May 6;12(1):1–10.
- 77. Ortolá R, García-Esquinas E, López-García E, León-Muñoz LM, Banegas JR, Rodríguez-Artalejo F. Alcohol consumption and all-cause mortality in older adults in Spain: an analysis accounting for the main methodological issues. Addiction. 2019 Jan 1;114(1):59–68. pmid:30063272
- 78. Guan SP, Kumar SN, Fann DY, Kennedy BK. A mechanistic perspective on the health promoting effects of alcohol–a focus on epigenetics modification. Alcohol. 2022 Aug 18;107:91–6. pmid:35987314
- 79. Bryazka D, Reitsma MB, Griswold MG, Abate KH, Abbafati C, Abbasi-Kangevari M, et al. Population-level risks of alcohol consumption by amount, geography, age, sex, and year: a systematic analysis for the Global Burden of Disease Study 2020. Lancet (London, England). 2022 Jul 7;400(10347):185. pmid:35843246