1Division of Biosciences, Department of Life Sciences, College of Health, Medicine and Life Sciences, Brunel University London, London UB8 3PH, UK; moc.oohay@gnaijnuyix (X.J.); [email protected] (F.D.); [email protected] (A.I.B.)Find articles by
1Division of Biosciences, Department of Life Sciences, College of Health, Medicine and Life Sciences, Brunel University London, London UB8 3PH, UK; moc.oohay@gnaijnuyix (X.J.); [email protected] (F.D.); [email protected] (A.I.B.)Find articles by
1Division of Biosciences, Department of Life Sciences, College of Health, Medicine and Life Sciences, Brunel University London, London UB8 3PH, UK; moc.oohay@gnaijnuyix (X.J.); [email protected] (F.D.); [email protected] (A.I.B.)
3Department of Metabolism, Digestion and Reproduction, Faculty of Medicine, Imperial College London, London W12 7TA, UKFind articles by
1Division of Biosciences, Department of Life Sciences, College of Health, Medicine and Life Sciences, Brunel University London, London UB8 3PH, UK; moc.oohay@gnaijnuyix (X.J.); [email protected] (F.D.); [email protected] (A.I.B.)
4Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London W12 7TA, UKFind articles by
Alcohol consumption is linked to urinary sodium excretion and both of these traits are linked to hypertension and cardiovascular diseases (CVDs). The interplay between alcohol consumption and sodium on hypertension, and cardiovascular diseases (CVDs) is not well-described. Here, we used genetically predicted alcohol consumption and explored the relationships between alcohol consumption, urinary sodium, hypertension, and CVDs. Methods: We performed a comparative analysis among 295,189 participants from the prospective cohort of the UK Biobank (baseline data collected between 2006 and 2010). We created a genetic risk score (GRS) using 105 published genetic variants in Europeans that were associated with alcohol consumption. We explored the relationships between GRS, alcohol consumption, urinary sodium, blood pressure traits, and incident CVD. We used linear and logistic regression and Cox proportional hazards (PH) models and Mendelian randomization in our analysis. Results: The median follow-up time for composite CVD and stroke were 6.1 years and 7.1 years respectively. Our analyses showed that high alcohol consumption is linked to low urinary sodium excretion. Our results showed that high alcohol GRS was associated with high blood pressure and higher risk of stroke and supported an interaction effect between alcohol GRS and urinary sodium on stage 2 hypertension (Pinteraction = 0.03) and CVD (Pinteraction = 0.03), i.e., in the presence of high urinary sodium excretion, the effect of alcohol GRS on blood pressure may be enhanced. Conclusions: Our results show that urinary sodium excretion may offset the risk posed by genetic risk of alcohol consumption.
Cardiovascular disease (CVD) is a global public health problem, killing 17 million people annually [1]. Alcohol consumption plays a role in the development of hypertension and CVD [2,3,4] and reducing alcohol consumption has been shown to lower blood pressure [5].
The sodium balance plays an important role in blood pressure regulation [6,7]. Urinary sodium excretion is associated with blood pressure [8]. We have previously shown that higher genetic risk of urinary sodium is associated with higher systolic blood pressure (SBP) and diastolic blood pressure (DBP) [9]. Alcohol consumption has been reported to decrease urinary sodium excretion [10]. A recent cross-sectional study [11] used data from older adults in northern China and found that the combination of alcohol consumption and sodium intake imposed a greater risk of hypertension. This implies that there might be a synergic effect between alcohol consumption and sodium on the risk of hypertension.
Advances in genetic data acquisition and analysis have improved our understanding of complex relationships and the biological mechanisms underpinning complex diseases. Recent genome-wide association studies (GWASs) among Europeans identified genetic variants in the form of single nucleotide polymorphisms (SNPs) associated with alcohol consumption [12,13] and urinary sodium [9].
To better understand the relationships between alcohol consumption, urinary sodium, hypertension, and CVDs, we constructed a genetic risk score (GRS) for alcohol consumption based on 105 SNPs associated with alcohol consumption in Europeans [12,13]. We explored the relationship between GRS and alcohol consumption and the effect of GRS on blood pressure, risk of hypertension, and CVDs in relation with urinary sodium excretion. We performed our study using individual-level data from 295,189 UK Biobank (UKB) participants.
7. GRS for Alcohol Consumption
We calculated a GRS for alcohol consumption based on 105 published SNPs (Supplementary Table S3) associated with alcohol consumption in Europeans [12,13]. The SNP selection process is illustrated in . To summarize, we obtained SNPs (N = 145) identified from two large-scale GWAS meta-analyses on alcohol consumption [12,13]. After removing duplicates, we assessed the linkage disequilibrium (LD) among the remaining SNPs using LDlink [22] and PLINK version 1.9 (Shaun Purcell, Christopher Chang, Boston, MA, USA, URL:www.cog-genomics.org/plink/1.9/ accessed on 12 July 2022) [23]. We defined SNP pairs with R2 > 0.1 as correlated SNPs. Within each SNP pair, we removed the SNP with the weaker association with alcohol consumption as indicated by a larger p value. As a result of these exclusions, 105 SNPs (Supplementary Table S3) remained for the alcohol GRS calculation.
To calculate the GRS for each UKB participant, we sought the effect estimates for the 105 alcohol SNPs from the alcohol GWAS meta-analysis study by Liu et al. [13]. To avoid sample overlap with the UKB data, we specifically extracted the effect estimates from the summary statistics excluding UKB and 23andMe provided by Liu et al. [13]. We also checked that this summary statistics dataset used the same genome build assembly (GRCh37) as the UKB genotype data [14]. To calculate the GRS, we multiplied the effect estimates by the number of risk alleles each UKB participant carries on the alcohol SNPs. The products were then summed across all SNPs to produce an overall weighted GRS for each participant. We standardized the weighted GRS for further analysis.
To better understand the etiology behind our findings, we performed a series of secondary analyses, including regression analyses, between urinary sodium and various alcoholic beverages and various outcomes such as diabetes, myocardial infarction, stroke, and CVD. We additionally performed Mendelian randomization (MR) analysis between urinary sodium and alcohol consumption. MR uses genetic variants (SNPs) that are robustly associated with an exposure of interest as instrumental variables to assess the causal effect of the exposure on an outcome [24]. In our analysis, the frequency of alcohol consumption was considered as the exposure and urinary sodium was considered as the outcome. The alcohol SNP selection process is illustrated in . To further process these SNPs for MR analysis, we clumped these SNPs at a distance = 10,000 kb and R2 = 0.001 [25] to ensure the independence of SNPs according to the MR guidelines. This removed 27 additional SNPs. Consequently, to identify any weak alcohol instruments that could lead to weak instrument bias in the MR analysis [26], a parameter called F-statistics was used, where if <10 indicates a weak instrument [27,28]. We calculated F-statistics using a published formula [29]:F=R2(N−2)/(1−R2)
where R2 is the variance in alcohol consumption explained by each SNP and was calculated using a published formula [30]; N is the size of the sample in which SNP–alcohol consumption association test statistics were calculated. After removal of the weak instruments, we were eventually left with 33 SNPs for the two-sample MR analysis. We obtained the corresponding test statistics for the association of these SNPs with urinary sodium within the UKB from Pazoki et al. [9]. We performed SNP harmonization across the alcohol and urinary sodium datasets and checked the strand orientation prior to analysis. To derive the MR estimates, we used the inverse variance weighted (IVW) method, which calculates the MR causal estimate with the highest precision [31]. However, these methods assume no horizontal pleiotropy. Horizontal pleiotropy occurs when the SNPs affect urinary sodium through traits other than alcohol consumption and these traits are not in the causal pathway from alcohol consumption to urinary sodium. When horizontal pleiotropy is left unbalanced, it can bias the MR causal estimate [32]. To detect the presence of unbalanced horizontal pleiotropy and correct the MR estimate for any bias due to this, we performed additional sensitivity tests such as the MR-Egger [32], weighted median [33], and the weighted mode tests [34]. An MR-Egger intercept p value < 0.05 suggests overall unbalanced horizontal pleiotropy [32]. To claim significance on MR analyses, we used a p value threshold of 0.05.
We assessed the variance in alcohol consumption explained by GRS using the adjusted R2 estimate from a linear regression model regressing alcohol consumption on GRS. To assess the predictability of the GRS at different alcohol consumption levels, we compared the percentage variation in alcohol consumption explained by the GRS across alcohol consumption quintiles comprising 5 equal groups.
We investigated the association of the GRS with SBP, DBP, Stage 1 hypertension, Stage 2 hypertension, and incident CVDs. We used linear regression for SBP and DBP, logistic regression for Stage 1 and Stage 2 hypertension, and performed survival analysis using Cox proportional hazards (PH) regression that takes the follow-up time into account for incident CVD traits. In our survival analyses, individuals who were lost to follow-up, died of diseases not under study, or did not develop diseases of interest at the end of the follow-up were censored. We assessed the PH assumption for every Cox model using statistical tests that used Schoenfeld residuals [35] against the follow-up time. When the p value for the PH assumption global test was <0.05, i.e., the overall PH assumption was violated, we examined which specific covariate violated the PH assumption in the model. For time-varying continuous covariate(s), we added interaction terms, with the follow-up time split into groups. The interaction term was modeled on a multiplicative scale in a linear regression model for blood pressure, logistic model for hypertension, and Cox model for CVDs.
We tested two statistical models in all analyses. We adjusted model 1 for age, age2, and sex. In model 2, we additionally adjusted for major known cardiovascular and genetic confounders, including smoking status, DASH diet, Townsend deprivation score, sedentary lifestyle, and genetic principal components. We adjusted the analyses for the interaction between alcohol GRS and urinary sodium, where the interaction term was statistically significant.
We calculated the statistical power for the associations of alcohol GRS with hypertension and CVDs using Quanto (version 1.2.4, Los Angeles, CA, USA) [36,37]. Using a two-sided significance threshold of 0.05 and a range of a number of cases and effect estimates from 0.9–1.3, we obtained an estimation of the statistical power for our analyses.
Study Population and Exclusion Criteria
UKB is a large prospective cohort set up in 22 centers across the United Kingdom. It consists of over 500,000 participants aged 40 to 69 recruited between 2006 and 2010 [14]. Genetic data on 487,409 participants were available for analysis. We applied several exclusion criteria ( ). We excluded participants who withdrew consent (N = 109), participants of non-European ancestry (N = 28,547), first- or second-degree relatives (N = 34,876), pregnant women or women unsure of pregnancy status at baseline (N = 232), prevalent CVD cases (N = 25,561), sex mismatch (N = 140), participants with health-related change in drinking habits (N = 65,579), non-alcoholic drinkers or participants with missing alcohol consumption data (N = 36,023), and participants with missing data regarding the main study variables (N = 1153). A total of 295,189 participants remained for analysis. Participants with health-related change in drinking habits included participants who self-reported to reduce/stop their drinking for one of the following reasons on a touch screen question: (1) illness or ill health, (2) doctor’s advice, or (3) health precaution.
New study shows long-term effects of alcohol on brains
FAQ
Does drinking alcohol lower your sodium?
Can you drink wine on a low sodium diet?
What’s worse for blood pressure salt or alcohol?
How long does it take to recover from low sodium levels?
Does red wine have a lot of salt?
This adds some salt to the wine but not much. Four ounces of domestic red wine typically contains about 12 milligrams of sodium, while domestic whites have about 19 milligrams in the same serving. Imported wines have less, at 6 milligrams for reds and just 2 milligrams for whites.
What are the effects of drinking one cup of red wine almost every day?
Wines rich in polyphenols, particularly resveratrol, anthocyanins, and catechins, are the best antioxidants for wine. By scavenging free oxygen radicals and reactive nitrogenous radicals, resveratrol protects the brain and nerve cells by permeating the blood-brain barrier. It also plays a role in the prevention of cardiovascular diseases. Additionally, it lessens platelet aggregation, which prevents blood clots or thrombi from forming. If we can also be careful with the balanced diet, this potential benefit would be good.
How does wine affect alcohol consumption?
Thus, by increasing the time wine stays in the stomach, food consumption effectively slows alcohol uptake. Consequently, the liver has more time to metabolize ethanol and, the maximum concentration reached in the blood is more likely to remain acceptably low.
Does drinking wine affect potassium levels?
Nevertheless, excessive alcohol consumption can disturb the uptake of calcium, magnesium, selenium, and zinc; and increases the excretion of zinc by the kidneys. The low sodium and high potassium content of wine makes it one of the more effective sources of potassium for individuals using diuretics.