Abstract
Background
Alcohol use is associated with increased blood pressure among adults with hypertension, but it is unknown whether some of the observed relationship is explained by mediating behaviors related to alcohol use.
Objective
We assess the potential indirect role of smoking, physical inactivity, unhealthy diet, and poor medication adherence on the association between alcohol use and blood pressure among Black and White men and women with hypertension.
Design
Adjusted repeated-measures analyses using generalized estimating equations and mediation analyses using inverse odds ratio weighting.
Participants
1835 participants with hypertension based on ACC/AHA 2017 guidelines in three most recent follow-up exams of the longitudinal Coronary Artery Risk Development in Young Adults cohort study (2005–2016).
Main Measures
Alcohol use was assessed using both self-reported average ethanol intake (drinks/day) and engagement in heavy episodic drinking (HED) in the past 30 days. Systolic and diastolic blood pressure (SBP, DBP) were measured by trained technicians (mmHg). Smoking, physical inactivity, and diet were self-reported and categorized according to American Heart Association criteria, and medication adherence was assessed using self-reported typical adherence to antihypertensive medications.
Key Results
At baseline (2005–2006), 57.9% of participants were Black and 51.4% were women. Mean age (standard deviation) was 45.5 (3.6) years, mean SBP was 128.7 (15.5) mmHg, and mean DBP was 83.2 (10.1) mmHg. Each additional drink per day was significantly associated with higher SBP (β = 0.713 mmHg, 95% confidence interval (CI): 0.398, 1.028) and DBP (β = 0.398 mmHg, 95% CI: 0.160, 0.555), but there was no evidence of mediation by any of the behaviors. HED was not associated with blood pressure independent of average consumption.
Conclusions
These findings support the direct nature of the association of alcohol use with blood pressure and the utility of advising patients with hypertension to limit consumption in addition to other behavioral and pharmacological interventions.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11606-021-07375-3.
KEY WORDS: alcohol use, blood pressure, health behaviors, mediation
BACKGROUND AND OBJECTIVE
Hypertension is highly prevalent in the USA, experienced by approximately 45.6% of adults.1 However, blood pressure control is achieved by a limited proportion of adults with hypertension2,3 and behavioral interventions are recommended for optimal control even among those taking antihypertensive medications. Along with maintaining a healthy weight, engaging in regular physical activity, and following a healthy diet, adults with hypertension are counseled to limit alcohol intake.1,4
Reduced alcohol consumption is associated with lower blood pressure and risk of hypertension in both intervention trials and Mendelian randomization studies.5–10 It is unclear, however, if some of the observed relationship may be explained by mediating behaviors that are detrimental to blood pressure control and, simultaneously, made more likely by alcohol use. Alcohol use has been associated with unhealthy diet and smoking,11 particularly among adults with hypertension.12,13 Similarly, individuals who drink may forget to take their medications as a result of intoxication or may choose to skip a medication to avoid toxic interactions.14 Recent work has shown alcohol use to be associated with antihypertensive medication non-adherence.14–17
In this study, we assess the mediating role of suboptimal cardiovascular health behaviors, including smoking, limited physical activity, unhealthy diet, and poor medication adherence in the association between alcohol use and blood pressure among Black and White men and women with hypertension. We hypothesize that these behaviors will partially, but not fully, mediate the relationship between alcohol and blood pressure.
METHODS
Participants
We used data from the Coronary Artery Risk Development in Young Adults (CARDIA) study, a multi-center prospective cohort study designed to examine the development and determinants of cardiovascular disease and their risk factors.18 The study began in 1985 and enrolled 5115 Black and White men and women between ages 18 and 30 at four study centers in the following cities: Birmingham, AB; Chicago, IL; Minneapolis, MN; and Oakland, CA. Recruitment was conducted to achieve balance in sex, race, age, and education within each center. Follow-up exams occurred 2, 5, 7, 10, 15, 20, 25, and 30 years after the initial exam. Written informed consent was provided at each exam, and the study was approved by institutional review boards at each of the study centers.
This analysis used data from the three most recent follow-up exams (years 20, 25, and 30, administered in 2005–2006, 2010–2011, and 2015–2016 respectively). Exposures, outcomes, mediators, and time-dependent covariates were all assessed at each exam, which occur over several hours on a single day. The analysis was restricted to participants with hypertension, defined by current guidelines as having systolic blood pressure (SBP) greater than or equal to 130 mmHg and diastolic blood pressure (DBP) greater than or equal to 80 mmHg, or currently taking antihypertensive medications.1 The sample was also restricted to participants who reported drinking within the last year.19 Participants were included in the analytic sample for as many and any combination of the three exams in which they participated and met these criteria. As a result, each included participant contributed between one and three exam observations. After eliminating observations with incomplete covariate data (n = 243 participants), the resultant sample contained data on 1835 unique participants with a total of 3350 observations. Specifically, 929 participants were observed at the year 20 exam, 1207 participants at year 25, and 1214 participants at year 30.
Main Measures
The main exposure was past-year average alcohol consumption, measured using self-reported typical intake (in total ethanol) and expressed as drinks per day (calculation detail in Appendix). We additionally assessed engagement in heavy episodic drinking (HED), as different patterns of drinking may differentially influence blood pressure. HED was measured using a dichotomous variable for which participants received a value of one if they reported having five or more drinks on the same occasion one or more times in the past 30 days and a value of zero otherwise.
The outcomes, SBP and DBP, were assessed by trained technicians at each exam. After resting for 5 min, participants’ blood pressure was measured on the right arm three times at 1-min intervals. Measurements were taken with an automated oscillometric monitor (OmROn HEM907XL) then calibrated and standardized to sphygmomanometric measures to eliminate any machine bias.20,21 The averages of the second and third measurements were used for analysis.
As mediators, we considered three health behaviors that have been negatively associated with hypertension management in prior studies (smoking, limited physical activity, unhealthy diet), and medication adherence among participants currently taking antihypertensive medications. All were assessed by questionnaire. Smoking, physical activity, and diet were categorized based on the American Heart Association (AHA) criteria for poor, intermediate, and ideal cardiovascular health.22,23 Smoking status was considered poor if the respondent currently smoked, intermediate if they quit smoking within the last 12 months, or ideal if they never smoked or quit over 12 months ago. For physical activity, under 100 exercise units of moderate and vigorous physical activity per week was considered poor, 100–299 units intermediate, and over 300 ideal (corresponding to approximately 30 min of exercise five times per week). At the year 20 exam, participants completed a detailed diet history and diet was classified according to the number of components of the AHA-recommended diet met by the reported diet. Meeting one or fewer of these components was classified as poor, two to three components was classified as intermediate, and four or more components was classified as ideal. During the year 25 and 30 exams, diet history was not assessed, so classification was based on the frequency with which respondents consumed fast food and sugar-sweetened beverages, with more than two times per week considered poor, twice per week or less intermediate, and none ideal. Further detail on questionnaires and the calculation of diet measures is available in the Appendix.
Medication adherence was assessed using a question that asked respondents who reported taking antihypertensive medications “on average, how often do you take your medication as prescribed?” Those who answered “all or almost all of the time” were considered adherent, and those who answered “most of the time,” “some of the time”, “rarely,” or “never or almost never” were considered non-adherent.
Time-invariant covariates included study center, sex, and race. Time-dependent covariates included age, educational attainment (less than high school degree, high school and some college, college degree or higher), average household income (less than $50,000, $50,000–100,000, over $100,000), marital status, depression (Center for Epidemiologic Studies Depression score above 16, suggesting the presence of depressive symptoms24), health insurance status (insured versus uninsured), and number of coexisting chronic medical conditions (including diabetes, heart disease, kidney disease, cancer, and lung disease).25–30
Statistical Analysis
First-Stage Analyses
We first used repeated-measures analyses to examine associations between alcohol use and blood pressure outcomes. We estimated one series of models for average alcohol consumption and another for both average consumption and HED to ascertain if this pattern of drinking was associated with blood pressure independent of the amount consumed. We used generalized estimating equations (GEE) with identity link functions and exchangeable covariance structure to account for within-individual correlation. Models included the covariates listed above and estimated Huber-White robust standard errors.31,32 We additionally used locally weighted regression33 of average consumption and blood pressure to visually evaluate whether associations were consistent across all values of consumption, as some previous work suggests that there may be a threshold of drinks per day below which reductions in alcohol use do not affect blood pressure.5
We then examined associations between alcohol use and each of the mediators, as mediation by nature requires the existence of these associations. These models used GEE with multinomial logistic regression (for categorical behaviors) and modified Poisson regression (for medication adherence) and were adjusted for covariates. The model for medication adherence was restricted to participants currently taking antihypertensive medications.
Mediation Analysis
Following these analyses, we conducted mediation analyses for any combination of exposure, outcome, and mediators for which we observed a statistically significant association (p < 0.05) between the exposure and the outcome and at least one mediator. Mediation was assessed using inverse odds ratio weighting (IORW),34,35 a semi-parametric counterfactual-based approach that decomposes the total effect of the exposure on the outcome into the natural direct effect (independent of the mediator) and the natural indirect effect (via the mediator). This method involves estimating a series of regressions, coefficients from which represent the total effect and the direct effect. The difference between the direct effect coefficient and the total effect coefficient represents the indirect effect. Standard errors and confidence intervals are obtained using non-parametric bootstrapping. Further detail on this method and its implementation in this analysis is presented in the appendix. Mediation was considered present when indirect effects were statistically significant at the 0.05 level with 1000 iterations. If mediation was present for multiple mediators, we estimated their cumulative effect by including all mediators in the inverse odds ratio weight.
Sensitivity Analyses
First, we sought evidence of mediation using the traditional Baron and Kenny method (1986).36 This method is not used as the primary method as it cannot account for exposure-mediator interactions nor for multiple mediators at once. Nonetheless, it may still be useful in detecting mediation for confirmatory purposes. Second, for consistency, we estimated models measuring diet with only fast food and sugar-sweetened beverage consumption for all exams. Third, we looked for differential effects by sex, race, and the use of antihypertensive medications by including interaction terms in the first-stage analyses. Fourth, to avoid inadvertently excluding data from participants who may be borderline hypertensive and fluctuating in and out of hypertension, we repeated analyses on a sample including all three exam observations for participants who met the definition of hypertension at least once during the three exams. Finally, we defined hypertension using the previous guidelines of SBP greater than or equal to 140 mmHg or DBP greater than or equal to 90 mmHg.37 Our main analyses used the current clinical guidelines in an effort to most accurately capture the population of interest, but these guidelines were not in place at the time of data collection.
Key Results
First-Stage Analyses
The mean age of participants at the year 20 exam was 45.5 years old (standard deviation (SD) = 3.6). At each exam, the sample was over 51% female and over 55% Black. Participants consumed on average 0.9 drinks per day (year 20 SD = 1.7, year 25 SD = 1.6, year 30 SD = 1.3), and approximately 30% reported engaging in HED. Mean SBP increased slightly over the years, from 128.7 mmHg (SD = 15.5) in year 20 to 129.6 mmHg (SD = 15.6) in year 25 to 130.3 mmHg (SD = 16.5) in year 30. Mean DBP decreased from 83.2 mmHg (SD = 10.1) to 82.6 mmHg (SD = 10.2) to 80.5 mmHg (SD = 10.6). The percent of participants taking antihypertensive medications also increased from 42.5 to 51.0 to 57.3% in years 20, 25, and 30 respectively. These variables as well as demographic and socioeconomic characteristics of participants at each exam are presented in Table 1.
Table 1.
Participant Characteristics at Each CARDIA Follow-Up Exam, 2005–2015 (n = 1835)
| Observations per exam year | Year 20 (2005) (n = 929) |
Year 25 (2010) (n = 1207) |
Year 30 (2015) (n = 1214) |
|---|---|---|---|
| Mean ± SD or n (%) | Mean ± SD or n (%) | Mean ± SD or n (%) | |
| Age (years) | 45.5 ± 3.6 | 50.3 ± 3.6 | 55.2 ± 2.6 |
| Female | 477 (51.4) | 621 (51.5) | 652 (52.7) |
| Race | |||
| Black | 538 (57.9) | 687 (56.9) | 673 (55.4) |
| White | 391 (42.1) | 520 (43.1) | 541 (44.6) |
| Education | |||
| Less than high school degree | 41 (4.4) | 57 (4.7) | 44 (3.6) |
| High school and some college | 507 (54.6) | 603 (50.0) | 590 (48.6) |
| College degree or higher | 381 (41.0) | 547 (45.3) | 580 (47.8) |
| Average household income | |||
| < $50,000 | 336 (36.2) | 430 (35.6) | 381 (31.4) |
| $50,000–99,999 | 325 (35.0) | 398 (33.0) | 348 (28.7) |
| $100,000+ | 268 (28.9) | 379 (31.4) | 485 (39.9) |
| Insurance status | |||
| Employer sponsored | 634 (68.3) | 812 (67.3) | 846 (69.7) |
| Medicare/Medicaid | 42 (4.5) | 83 (.9) | 121 (10.0) |
| Veterans Health Administration | 12 (1.3) | 18 (1.5) | 19 (1.6) |
| Self-insured | 143 (15.4) | 134 (11.1) | 139 (11.5) |
| Uninsured | 98 (10.6) | 160 (13.3) | 89 (7.3) |
| Married/living with partner | 564 (60.7) | 707 (58.6) | 721 (59.4) |
| Experiencing depressive symptoms | 186 (20.0) | 208 (17.2) | 201 (16.6) |
| Number of coexisting chronic conditions* | 0.5 ± 0.7 | 0.5 ± 0.8 | 0.7 ± 0.9 |
| Alcohol use | |||
| Average daily consumption, drinks/day | |||
| 0 | 324 (34.9) | 380 (31.5) | 366 (30.2) |
| 0.1–1.0 | 359 (38.6) | 472 (39.1) | 507 (41.8) |
| 1.1–2.0 | 135 (14.5) | 199 (16.5) | 178 (14.7) |
| 2.1–3.0 | 55 (5.9) | 74 (6.1) | 83 (6.8) |
| 3+ | 56 (6.1) | 82 (6.8) | 80 (6.6) |
| Heavy episodic drinking in past 30 days | 278 (29.9) | 372 (30.8) | 355 (29.2) |
| Blood pressure | |||
| Systolic blood pressure, mmHg | 128.7 ± 15.5 | 129.6 ± 15.6 | 130.3 ± 16.5 |
| Diastolic blood pressure, mmHg | 83.2 ± 10.1 | 82.6 ± 10.2 | 80.5 ± 10.6 |
| Taking antihypertensive medications | 395 (42.5) | 615 (51.0) | 696 (57.3) |
| Behavior and medication regimens | |||
| Smoking (n = 1823) | |||
| Poor | 222 (24.9) | 253 (21.4) | 205 (17.1) |
| Intermediate | 28 (3.1) | 27 (2.3) | 30 (2.5) |
| Ideal | 668 (72.8) | 904 (76.4) | 967 (80.5) |
| Physical activity (n = 1835) | |||
| Poor | 203 (21.9) | 248 (20.6) | 277 (22.8) |
| Intermediate | 320 (34.5) | 424 (35.1) | 456 (37.6) |
| Ideal | 405 (43.6) | 535 (44.3) | 480 (39.6) |
| Diet (n = 1802) | |||
| Poor | 390 (48.1) | 941 (78.2) | 840 (70.5) |
| Intermediate | 403 (49.7) | 188 (15.6) | 243 (20.4) |
| Ideal | 18 (2.2) | 75 (6.2) | 108 (9.1) |
| Medication adherence (n = 991) | |||
| Adherent | 292 (75.5) | 497 (83.7) | 573 (83.2) |
| Non-adherent | 95 (24.6) | 97 (16.3) | 116 (16.8) |
Included participants are those with hypertension (SBP ≥ 130 mmHg, DBP ≥ 80 mmHg, or currently taking antihypertensive medications) and who have consumed alcohol within the past year
*Coexisting chronic conditions include diabetes, heart disease, kidney disease, cancer, and lung disease
Data on mediators were complete for 98% of participants. In all three exams, most respondents had ideal smoking behaviors (never smoked or had quit at least 1 year prior), but most also had diets considered poor by AHA criteria. Physical activity was more evenly distributed. Of the participants taking antihypertensive medications at each exam, over 75% were adherent (see Table 1).
In adjusted repeated-measures analyses, average daily alcohol consumption was significantly associated with higher blood pressure (Table 2). One additional drink per day was, on average, associated with 0.71 mmHg higher SBP (95% confidence interval (CI): 0.40, 1.03) and 0.36 mmHg higher DBP (95% CI: 0.16, 0.56), before additionally adjusting for HED. These associations are represented in Figures 1 and 2 along with the results of locally weighted regression analyses, which suggest roughly linear relationships without any substantial thresholds. HED was not significantly associated with SBP or DBP independent of average consumption.
Table 2.
Adjusted β-Coefficients for Association Between Alcohol Use and Blood Pressure Among Adults with Hypertension, CARDIA, 2005–2016
|
Systolic blood pressure (mmHg) (n = 1835) |
Diastolic blood pressure (mmHg) (n = 1835) |
|||
|---|---|---|---|---|
|
Estimate (95% CI) |
Estimate (95% CI) |
|||
| Average alcohol intake (drinks/day) |
0.71*** (0.40, 1.03) |
0.68*** (0.35, 1.01) |
0.36*** (0.16, 0.56) |
0.32** (0.11, 0.53) |
| Heavy episodic drinking |
0.29 (− 1.00, 1.57) |
0.33 (− 0.51, 1.17) |
||
Estimates are from generalized estimating equations with exchangeable covariance structure and identity link function. All models are adjusted for center, age, sex, race, level of educational attainment, income, marital status, depressive symptoms, health insurance status, and number of coexisting chronic medical conditions. Each vertical column represents one model. Boldface indicates statistical significance
**p < 0.01; ***p < 0.001
Figure 1.
Linear trends and locally weighted regression of systolic blood pressure and average daily alcohol consumption, n = 1835
Figure 2.
Linear trends and locally weighted regression of diastolic blood pressure and average daily alcohol consumption, n = 1835
Regarding mediators, higher average alcohol consumption and HED were independently associated with lower odds of ideal smoking status. Conversely, alcohol consumption was associated with higher odds of ideal diet, and there was no additional independent association of HED. Neither average alcohol consumption nor HED was significantly associated with physical activity or medication adherence (Table 3).
Table 3.
Adjusted Odds or Risk Ratios for Potential Behavioral Mediators and Alcohol Use Among Adults with Hypertension, CARDIA, 2005–2016
|
Smoking (n = 1823†) |
Physical activity (n = 1835) |
Diet (n = 1802†) |
Medication adherence (n = 991†) |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Odds ratio (95% CI) |
Odds ratio (95% CI) |
Odds ratio (95% CI) |
Risk ratio (95% CI) |
|||||||||||
| Intermediate | Ideal | Intermediate | Ideal | Intermediate | Ideal | Adherent | ||||||||
| Average alcohol intake (drinks/day) |
0.99 (0.89, 1.09) |
0.96 (0.85, 1.08) |
0.82*** (0.77, 0.87) |
0.84*** (0.78, 0.89) |
1.06 (0.96, 1.16) |
1.03 (0.94, 1.14) |
1.10 (1.00, 1.21) |
1.07 (0.97, 1.18) |
1.03 (0.97, 1.10) |
1.03 (0.96, 1.10) |
1.15*** (1.06, 1.23) |
1.14** (1.04, 1.24) |
0.997 (0.98, 1.02) |
0.996 (0.98, 1.02) |
| Heavy episodic drinking |
1.22 (0.70, 2.12) |
0.76** (0.64, 0.91) |
1.17 (0.91, 1.50) |
1.21 (0.95, 1.55) |
1.03 (0.83, 1.38) |
1.10 (0.76, 1.59) |
1.01 (0.95, 1.07) |
|||||||
Estimates are from generalized estimating equations using multinomial logistic regression (smoking, physical activity, and diet) and modified Poisson regression (medication adherence). The reference for smoking, physical activity, and diet is poor, and the reference for adherence is non-adherent. The adherence model was estimated only among participants currently taking antihypertensive medications. All models are adjusted for center, age, sex, race, level of educational attainment, income, marital status, depressive symptoms, health insurance status, and number of coexisting chronic medical conditions. Each vertical column represents one model. Boldface indicates statistical significance
**p < 0.01; ***p < 0.001
†Samples are less than full because some participants were missing data on mediators
Mediation Analyses
Based on the associations noted in the first stage of analysis, smoking and diet were evaluated as mediators of the association between average alcohol consumption and blood pressure. There were no statistically significant indirect effects of either behavior in relation to SBP or DBP (Table 4). As a result, no cumulative indirect effects were estimated.
Table 4.
IORW Mediation of Average Daily Alcohol Consumption and Blood Pressure Associations by Health Behaviors, CARDIA, 2005–2016
| Total effect | Direct effect | Indirect effect | |
|---|---|---|---|
| Smoking (n = 1823†) | |||
| SBP (mmHg) |
0.72*** (0.37, 1.07) |
0.93* (0.10, 1.75) |
− 0.21 (− 1.06, 0.64) |
| DBP (mmHg) |
0.33** (0.12, 0.55) |
0.25 (− 0.13, 0.62) |
0.09 (− 0.30, 0.48) |
| Diet (n = 1802†) | |||
| SBP (mmHg) |
0.78** (0.43, 1.13) |
0.80*** (0.42, 1.18) |
− 0.02 (− 0.13, 0.10) |
| DBP (mmHg) |
0.34** (0.11, 0.57) |
0.40** (0.14, 0.66) |
− 0.06 (− 0.14, 0.03) |
Table presents adjusted linear regression coefficients and 95% confidence intervals (in parentheses) estimated using nonparametric bootstrapping (1000 iterations) clustered at the participant level. Boldface indicates statistical significance
IORW inverse odds ratio weighting
*p < 0.05, **p < 0.01, ***p < 0.001
†Samples are less than full because some participants were missing data on behaviors
Sensitivity Analyses
The findings of no detectable indirect effects were robust to the method of evaluating mediation (see Appendix Table 1). Similarly, findings were consistent using an alternative diet measure. There were no significant interactions between alcohol use and sex, race, or medication status with regard to outcomes. Neither using the larger sample nor the previous hypertension definition (140/90 mmHg) led to appreciably different results.
DISCUSSION
Greater average daily alcohol consumption was associated with slightly but significantly higher SBP and DBP among a cohort of Black and White men and women with hypertension, and these relationships were not demonstrably mediated by other behaviors associated with hypertension management. These results are consistent with findings from intervention trials and Mendelian randomization studies that suggest a causal relationship between alcohol consumption and increased blood pressure.5–10 Previous research has demonstrated that there are multiple biological mechanisms through which this relationship may manifest. For example, alcohol can compromise arterial-vascular function via diminished baroreceptor sensitivity and can disrupt the renin-angiotensin-aldosterone system that controls the body’s fluid regulation.38–41 It can also stimulate the sympathetic nervous system, leading to increased noradrenaline, and interfere with endothelial cell function and decrease nitric oxide availability, causing buildup of arterial plaque.38,40–42
However, this study is unique in that it considers whether alcohol consumption relates to blood pressure in additional indirect ways beyond these direct biological connections. We found no evidence of such indirect relationships. Such findings have clinical significance, as they reinforce the benefit of screening adults with hypertension for alcohol use. Multiple recent studies have suggested that most adults with hypertension receive advice to limit dietary sodium and to increase exercise (66–76%), but no more than one third receive advice to reduce alcohol intake.43,44 Perhaps, as a result, adults with diagnosed hypertension have been found to be no different in terms of their engagement in moderate and excessive alcohol consumption than adults with undiagnosed hypertension.13 Had the association between alcohol use and hypertension been considerably mediated by behaviors such as diet, exercise, and smoking habits, there might be diminishing returns to counseling patients with hypertension about alcohol use in addition to these other behavioral modifications. However, given the lack of mediation, this limited counseling may represent a missed opportunity for intervention. While the associations observed in this study were small, they are similar to those observed between age and blood pressure for adults over 40 (SBP increases ~ 0.7 mmHg per year of age45). Further, intervention studies have identified more substantial effects of alcohol reduction on blood pressure—over 3 mmHg for SBP and over 2 mmHg for DBP. These effect sizes are similar in magnitude to effects of other behavioral interventions, including self-monitoring and some diet modifications.5,7,46,47 Still, with meta-analysis suggesting that, among this age group, a 2-mmHg-lower usual SBP is associated with approximately 10%-lower risk of death from stroke and 7%-lower risk of death from ischemic heart disease and other vascular causes,48 even small improvements can be meaningful.
Our remaining findings are generally in line with previous research, which supports the observed associations of alcohol use with smoking and diet quality and the lack of association observed with physical activity.49–51 Other studies have identified relationships between average consumption and poor medication adherence, but the discrepancy in our study may be due to different study populations or measures of adherence; many of these studies assessed adherence using pharmacy fills or validated scales,14–17 not available in CARDIA. The absence of an association between HED and blood pressure was also anticipated, as this association is inconsistent in the literature.52–57
Limitations
This study has several limitations. First, as the analysis is restricted to Black and White men and women from their late thirties to early sixties, our findings may not be generalizable to adults who are older or younger, or from other racial or ethnic backgrounds. Second, as noted, medication adherence was measured using a single question that has not been validated and may not be as sensitive as other measurements of adherence. Similarly, in years 25 and 30, diet was assessed in what is likely a less robust manner than through the complete diet history, as in year 20. While this method has been used similarly in the past when full diet histories were unavailable,23 it may not as comprehensively capture diet quality. Future studies may consider using more robust measures of these behaviors as well as additional behaviors linked to alcohol and blood pressure, such as sleep, on which we lacked data. Third, all of the mediators are subject to measurement error which may bias indirect effects away from the null and direct effects toward the null.58 However, corrections for measurement error in IORW require having an additional measure for validation of the mediator (i.e., a gold standard),59 which was not available. Fourth, temporality of alcohol use and the mediators cannot be established. Finally, IORW, while more appropriate for this analysis than traditional mediation methods, produces more variable estimates and thus may not be able to identify small effects. While traditional methods still failed to identify indirect effects, it remains possible that indirect effects exist that we were underpowered to detect.
CONCLUSION
This study suggests that average daily alcohol use is directly associated with systolic and diastolic blood pressure among middle-aged Black and White men and women with hypertension and provides no evidence that these relationships are mediated by other behaviors. Advising patients with hypertension to limit alcohol consumption in addition to other behavior regimens and pharmacological intervention may provide added benefit for lowering blood pressure.
Supplementary Information
(DOCX 23.5 kb)
Funding
The Coronary Artery Risk Development in Young Adults (CARDIA) study is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with the University of Alabama at Birmingham (HHSN268201800005I & HHSN268201800007I), Northwestern University (HHSN268201800003I), University of Minnesota (HHSN268201800006I), and Kaiser Foundation Research Institute (HHSN268201800004I). This manuscript has been reviewed by CARDIA for scientific content. The authors thank the investigators, the staff, and the participants of the CARDIA study for their valuable contributions.
Dr. Phillips’ effort was supported by a training award (T32HL069771) from NHLBI. Dr. Tajeu’s effort was supported by a career development award (1K01HL151974-01).
The content of this report is the responsibility of the authors and does not represent the official view of NHLBI or the U.S. National Institutes of Health.
Declarations
Conflict of Interest
The authors declare that they do not have a conflict of interest.
The data used in this analysis is subject to a data use agreement and cannot be widely shared, but readers interested in data access or replication should contact the corresponding author or the CARDIA Coordinating Center.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension. 2018;71(6). 10.1161/HYP.0000000000000065 [DOI] [PubMed]
- 2.Muntner P, Hardy ST, Fine LJ, et al. Trends in Blood Pressure Control Among US Adults With Hypertension, 1999-2000 to 2017-2018. JAMA. 2020;324(12):1190. doi: 10.1001/jama.2020.14545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ritchey MD, Gillespie C, Wozniak G, et al. Potential need for expanded pharmacologic treatment and lifestyle modification services under the 2017 ACC/AHA Hypertension Guideline. J Clin Hypertens. 2018;20(10):1377–1391. doi: 10.1111/jch.13364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dickinson HO, Mason JM, Nicolson DJ, et al. Lifestyle interventions to reduce raised blood pressure: a systematic review of randomized controlled trials. J Hypertens. 2006;24(2):215–233. doi: 10.1097/01.hjh.0000199800.72563.26. [DOI] [PubMed] [Google Scholar]
- 5.Roerecke M, Kaczorowski J, Tobe SW, Gmel G, Hasan OSM, Rehm J. The effect of a reduction in alcohol consumption on blood pressure: a systematic review and meta-analysis. Lancet Public Health. 2017;2(2):e108–e120. doi: 10.1016/S2468-2667(17)30003-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Taylor B, Irving HM, Baliunas D, et al. Alcohol and hypertension: gender differences in dose-response relationships determined through systematic review and meta-analysis: Alcohol and hypertension: a meta-analysis. Addiction. 2009;104(12):1981–1990. doi: 10.1111/j.1360-0443.2009.02694.x. [DOI] [PubMed] [Google Scholar]
- 7.Xin X, He J, Frontini MG, Ogden LG, Motsamai OI, Whelton PK. Effects of Alcohol Reduction on Blood Pressure: A Meta-Analysis of Randomized Controlled Trials. Hypertension. 2001;38(5):1112–1117. doi: 10.1161/hy1101.093424. [DOI] [PubMed] [Google Scholar]
- 8.Holmes MV, Dale CE, Zuccolo L, et al. Association between alcohol and cardiovascular disease: Mendelian randomisation analysis based on individual participant data. BMJ. 2014;349:g4164. doi: 10.1136/bmj.g4164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chen L, Smith GD, Harbord RM, Lewis SJ. Alcohol intake and blood pressure: a systematic review implementing a Mendelian randomization approach. PLoS Med. 2008;5(3):e52. doi: 10.1371/journal.pmed.0050052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Millwood IY, Walters RG, Mei XW, et al. Conventional and genetic evidence on alcohol and vascular disease aetiology: a prospective study of 500 000 men and women in China. The Lancet. 2019;393(10183):1831–1842. doi: 10.1016/S0140-6736(18)31772-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ma J, Betts NM, Hampl JS. Clustering of Lifestyle Behaviors: The Relationship between Cigarette Smoking, Alcohol Consumption, and Dietary Intake. Am J Health Promot. 2000;15(2):107–117. doi: 10.4278/0890-1171-15.2.107. [DOI] [PubMed] [Google Scholar]
- 12.Rittmueller SE, Frey MS, Williams EC, Sun H, Bryson CL, Bradley KA. Association Between Alcohol Use and Cardiovascular Self-Care Behaviors Among Male Hypertensive Veterans Affairs Outpatients: A Cross-Sectional Study. Substance Abuse. 2015;36(1):6–12. doi: 10.1080/08897077.2014.932318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kim MT, Han HR, Hill MN, Rose L, Roary M. Depression, substance use, adherence behaviors, and blood pressure in urban hypertensive black men. Ann Behav Med. 2003;26(1):24–31. doi: 10.1207/S15324796ABM2601_04. [DOI] [PubMed] [Google Scholar]
- 14.Bryson CL. Alcohol Screening Scores and Medication Nonadherence. Ann Intern Med. 2008;149(11):795. doi: 10.7326/0003-4819-149-11-200812020-00004. [DOI] [PubMed] [Google Scholar]
- 15.Cené CW, Dennison CR, Powell Hammond W, Levine D, Bone LR, Hill MN. Antihypertensive Medication Nonadherence in Black Men: Direct and Mediating Effects of Depressive Symptoms, Psychosocial Stressors, and Substance Use: Medication Nonadherence in Black Men. J Clin Hypertens. 2013;15(3):201–209. doi: 10.1111/jch.12056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Steiner JF, Ho PM, Beaty BL, et al. Sociodemographic and Clinical Characteristics Are Not Clinically Useful Predictors of Refill Adherence in Patients With Hypertension. Circ Cardiovasc Qual Outcomes. 2009;2(5):451–457. doi: 10.1161/CIRCOUTCOMES.108.841635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bandi P, Goldmann E, Parikh NS, Farsi P, Boden-Albala B. Age-Related Differences in Antihypertensive Medication Adherence in Hispanics: A Cross-Sectional Community-Based Survey in New York City, 2011–2012. Prev Chronic Dis. 2017;14. 10.5888/pcd14.160512 [DOI] [PMC free article] [PubMed]
- 18.Friedman GD, Cutter GR, Donahue RP, et al. Cardia: study design, recruitment, and some characteristics of the examined subjects. JClin Epidemiol. 1988;41(11):1105–1116. doi: 10.1016/0895-4356(88)90080-7. [DOI] [PubMed] [Google Scholar]
- 19.Shaper AG, Wannamethee G, Walker M. ALCOHOL AND MORTALITY IN BRITISH MEN: EXPLAINING THE U-SHAPED CURVE. The Lancet. 1988;332(8623):1267–1273. doi: 10.1016/S0140-6736(88)92890-5. [DOI] [PubMed] [Google Scholar]
- 20.Gunderson EP, Chiang V, Lewis CE, et al. Long-term blood pressure changes measured from before to after pregnancy relative to nonparous women. Obstet Gynecol. 2008;112(6):1294–1302. doi: 10.1097/AOG.0b013e31818da09b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jacobs DR, Yatsuya H, Hearst MO, et al. Rate of Decline of Forced Vital Capacity Predicts Future Arterial Hypertension: The Coronary Artery Risk Development in Young Adults Study. Hypertension. 2012;59(2):219–225. doi: 10.1161/HYPERTENSIONAHA.111.184101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lloyd-Jones DM, Hong Y, Labarthe D, et al. Defining and Setting National Goals for Cardiovascular Health Promotion and Disease Reduction: The American Heart Association’s Strategic Impact Goal Through 2020 and Beyond. Circulation. 2010;121(4):586–613. doi: 10.1161/CIRCULATIONAHA.109.192703. [DOI] [PubMed] [Google Scholar]
- 23.Whitaker KM, Jacobs DR, Kershaw KN, et al. Racial Disparities in Cardiovascular Health Behaviors: The Coronary Artery Risk Development in Young Adults Study. Am J Prev Med. 2018;55(1):63–71. doi: 10.1016/j.amepre.2018.03.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Beekman ATF, Deeg DJH, Van Limbeek J, Braam AW, De Vries MZ, Van Tilburg W. BRIEF COMMUNICATION: Criterion validity of the Center for Epidemiologic Studies Depression scale (CES-D): results from a community-based sample of older subjects in the Netherlands. Psychol Med. 1997;27(1):231–235. doi: 10.1017/S0033291796003510. [DOI] [PubMed] [Google Scholar]
- 25.Oates GR, Juarez LD, Hansen B, Kiefe CI, Shikany JM. Social Risk Factors for Medication Nonadherence: Findings from the CARDIA Study. Am J Health Behav. 2020;44(2):232–243. doi: 10.5993/AJHB.44.2.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Briesacher BA, Gurwitz JH, Soumerai SB. Patients At-Risk for Cost-Related Medication Nonadherence: A Review of the Literature. J Gen Intern Med. 2007;22(6):864–871. doi: 10.1007/s11606-007-0180-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kang H, Lobo JM, Kim S, Sohn MW. Cost-related medication non-adherence among U.S. adults with diabetes. Diabetes Res Clin Pract. 2018;143:24–33. doi: 10.1016/j.diabres.2018.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mentz RJ, Greiner MA, Muntner P, et al. Intentional and unintentional medication non-adherence in African Americans: Insights from the Jackson Heart Study. Am Heart J. 2018;200:51–59. doi: 10.1016/j.ahj.2018.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Vawter L, Tong X, Gemilyan M, Yoon PW. Barriers to Antihypertensive Medication Adherence Among Adults- United States, 2005. J Clin Hypertens. 2008;10(12):922–929. doi: 10.1111/j.1751-7176.2008.00049.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Mochari H, Ferris A, Adigopula S, Henry G, Mosca L. Cardiovascular Disease Knowledge, Medication Adherence, and Barriers to Preventive Action in a Minority Population. Prev Cardiol. 2007;10(4):190–195. doi: 10.1111/j.1520-037X.2007.06619.x. [DOI] [PubMed] [Google Scholar]
- 31.Huber PJ. The behavior of maximum likelihood estimates under nonstandard conditions. In: Lucien Lecam & Jersey Neyman, editors. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. Vol 1. University of California Press; 1967:221-233.
- 32.White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica: J Econometric Soc. Published online 1980:817-838.
- 33.Cleveland WS. Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc. 1979;74(368):829–836. doi: 10.1080/01621459.1979.10481038. [DOI] [Google Scholar]
- 34.Tchetgen Tchetgen EJ. Inverse odds ratio-weighted estimation for causal mediation analysis. Stat Med. 2013;32(26):4567–4580. doi: 10.1002/sim.5864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Nguyen QC, Osypuk TL, Schmidt NM, Glymour MM, Tchetgen Tchetgen EJ. Practical Guidance for Conducting Mediation Analysis With Multiple Mediators Using Inverse Odds Ratio Weighting. Am J Epidemiol. 2015;181(5):349–356. doi: 10.1093/aje/kwu278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Baron RM, Kenny DA. The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173. doi: 10.1037/0022-3514.51.6.1173. [DOI] [PubMed] [Google Scholar]
- 37.Chobanian AV. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. The JNC 7 Report. JAMA. 2003;289(19):2560. doi: 10.1001/jama.289.19.2560. [DOI] [PubMed] [Google Scholar]
- 38.Piano MR. Alcohol’s Effects on the Cardiovascular System. Alcohol Res. 2017;38(2):219–241. [PMC free article] [PubMed] [Google Scholar]
- 39.Puddey IB, Vandongen R, Beilin LJ, Rouse IL. Alcohol Stimulation of Renin Release in Man: Its Relation to the Hemodynamic, Electrolyte, and Sympatho-Adrenal Responses to Drinking*. J Clin Endocrinol Metab. 1985;61(1):37–42. doi: 10.1210/jcem-61-1-37. [DOI] [PubMed] [Google Scholar]
- 40.Kodavali L, Townsend RR. Alcohol and its relationship to blood pressure. Curr Sci Inc. 2006;8(4):338–344. doi: 10.1007/s11906-006-0074-z. [DOI] [PubMed] [Google Scholar]
- 41.Tasnim S, Tang C, Musini VM, Wright JM. Effect of alcohol on blood pressure. Cochrane Hypertension Group, ed. Cochrane Database Syst Rev. 2020;2020(7). 10.1002/14651858.CD012787.pub2 [DOI] [PMC free article] [PubMed]
- 42.Grassi GM, Somers VK, Renk WS, Abboud FM, Mark AL. Effects of alcohol intake on blood pressure and sympathetic nerve activity in normotensive humans: a preliminary report. J Hypertens. 1989;7:S20–21. doi: 10.1097/00004872-198900076-00007. [DOI] [PubMed] [Google Scholar]
- 43.Liu X, Byrd JB, Rodriguez CJ. Use of physician-recommended non-pharmacological strategies for hypertension control among hypertensive patients. J Clin Hypertens. 2018;20(3):518–527. doi: 10.1111/jch.13203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Sripipatana A, Pourat N, Chen X, Zhou W, Lu C. Exploring racial/ethnic disparities in hypertension care among patients served by health centers in the United States. J Clin Hypertens. 2019;21(4):489–498. doi: 10.1111/jch.13504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Wolf-Maier K. Hypertension Prevalence and Blood Pressure Levels in 6 European Countries, Canada, and the United States. JAMA. 2003;289(18):2363. doi: 10.1001/jama.289.18.2363. [DOI] [PubMed] [Google Scholar]
- 46.Gay HC, Rao SG, Vaccarino V, Ali MK. Effects of Different Dietary Interventions on Blood Pressure: Systematic Review and Meta-Analysis of Randomized Controlled Trials. Hypertension. 2016;67(4):733–739. doi: 10.1161/HYPERTENSIONAHA.115.06853. [DOI] [PubMed] [Google Scholar]
- 47.Tucker KL, Sheppard JP, Stevens R, et al. Self-monitoring of blood pressure in hypertension: A systematic review and individual patient data meta-analysis. Rahimi K, ed. PLoS Med. 2017;14(9):e1002389. doi: 10.1371/journal.pmed.1002389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002;360(9349):1903–1913. doi: 10.1016/s0140-6736(02)11911-8. [DOI] [PubMed] [Google Scholar]
- 49.Noble N, Paul C, Turon H, Oldmeadow C. Which modifiable health risk behaviours are related? A systematic review of the clustering of Smoking, Nutrition, Alcohol and Physical activity (‘SNAP’) health risk factors. Prev Med. 2015;81:16–41. doi: 10.1016/j.ypmed.2015.07.003. [DOI] [PubMed] [Google Scholar]
- 50.Cummings JR, Gearhardt AN, Ray LA, Choi AK, Tomiyama AJ. Experimental and observational studies on alcohol use and dietary intake: a systematic review. Obes Rev. 2020;21(2). 10.1111/obr.12950 [DOI] [PMC free article] [PubMed]
- 51.Dodge T, Clarke P, Dwan R. The Relationship Between Physical Activity and Alcohol Use Among Adults in the United States: A Systematic Review of the Literature. Am J Health Promot. 2017;31(2):97–108. doi: 10.1177/0890117116664710. [DOI] [PubMed] [Google Scholar]
- 52.Wellman RJ, Vaughn JA, Sylvestre MP, O’Loughlin EK, Dugas EN, O’Loughlin JL. Relationships Between Current and Past Binge Drinking and Systolic Blood Pressure in Young Adults. J Adolesc Health. 2016;58(3):352–357. doi: 10.1016/j.jadohealth.2015.10.251. [DOI] [PubMed] [Google Scholar]
- 53.Hayibor LA, Zhang J, Duncan A. Association of binge drinking in adolescence and early adulthood with high blood pressure: findings from the National Longitudinal Study of Adolescent to Adult Health (1994–2008) J Epidemiol Community Health. 2019;73(7):652–659. doi: 10.1136/jech-2018-211594. [DOI] [PubMed] [Google Scholar]
- 54.Fan A. Drinking pattern and blood pressure among non-hypertensive current drinkers: findings from 1999–2004 National Health and Nutrition Examination Survey. Clin Epidemiol. Published online January 2013:21. 10.2147/CLEP.S12152 [DOI] [PMC free article] [PubMed]
- 55.Ng Fat L, Bell S, Britton A. A life-time of hazardous drinking and harm to health among older adults: findings from the Whitehall II prospective cohort study. Addiction. 2020;115(10):1855–1866. doi: 10.1111/add.15013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Piano MR, Burke L, Kang M, Phillips SA. Effects of Repeated Binge Drinking on Blood Pressure Levels and Other Cardiovascular Health Metrics in Young Adults: National Health and Nutrition Examination Survey, 2011-2014. J Am Heart Assoc. 2018;7(13). 10.1161/JAHA.118.008733 [DOI] [PMC free article] [PubMed]
- 57.Pajak A, Szafraniec K, Kubinova R, et al. Binge Drinking and Blood Pressure: Cross-Sectional Results of the HAPIEE Study. Barengo NC, ed. PLoS ONE. 2013;8(6):e65856. doi: 10.1371/journal.pone.0065856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.VanderWeele TJ. Mediation Analysis: A Practitioner’s Guide. Annu Rev Public Health. 2016;37(1):17–32. doi: 10.1146/annurev-publhealth-032315-021402. [DOI] [PubMed] [Google Scholar]
- 59.Nguyen TT, Tchetgen Tchetgen EJ, Kawachi I, Gilman SE, Walter S, Glymour MM. Comparing Alternative Effect Decomposition Methods: The Role of Literacy in Mediating Educational Effects on Mortality. Epidemiology. 2016;27(5):670–676. doi: 10.1097/EDE.0000000000000517. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX 23.5 kb)