Critique 213: A meta-analysis on lifestyle factors, cardiovascular disease, and total mortality for a very large number of middle-aged and elderly women – 27 March 2018

Colpani V, Baena CP, Jaspers L, van Dijk GM, Farajzadegan Z, Dhana K, Tielemans MJ, Voortman T, Freak-Poli R, Veloso GGV, Chowdhury R, Kavousi M, Muka T. Franco OH. Lifestyle factors, cardiovascular disease and all-cause mortality in middle-aged and elderly women: a systematic review and metaanalysis.   European Journal of Epidemiology 2018; pre-publication. https://doi.org/10.1007/s10654-018-0374-z

Authors’ Abstract

Cardiovascular disease (CVD) risk factors, incidence and death increases from around the time of menopause comparing to women in reproductive age. A healthy lifestyle can prevent CVD, but it is unclear which lifestyle factors may help maintain and improve cardiovascular health for women after menopausal transition. We conducted a systematic review and meta-analysis of prospective cohort studies to evaluate the association between modifiable lifestyle factors (specifically smoking, physical activity, alcohol intake, and obesity), with CVD and mortality in middle-aged and elderly women.

Pubmed, Embase, among other databases and reference lists were searched until February 29th, 2016. Study specific relative risks (RR) were meta-analyzed using random effect models. We included 59 studies involving 5,358,902 women. Comparing current versus never smokers, pooled RR were 3.12 (95% CI 2.15–4.52) for CHD incidence, 2.09 (95% CI 1.51–2.89) for stroke incidence, 2.76 (95% CI 1.62–4.71) for CVD mortality and 2.22 (95% CI 1.92–2.57) for all-cause mortality. Physical activity was associated with a decreased risk of 0.74 (95% CI 0.67–0.80) for overall CVD, 0.71 (95% CI 0.67–0.75) for CHD, 0.77 (95% CI 0.70–0.85) for stroke, 0.70 (95% CI 0.58–0.84) for CVD mortality and 0.71 (95% CI 0.65–0.78) for all-cause mortality. Comparing moderate drinkers versus non-drinkers, the RR was 0.72 (95% CI 0.56–0.91) for CHD, 0.63 (95% CI 0.57–0.71) for CVD mortality and 0.80 (95% CI 0.76–0.84) for all-cause mortality. For women with BMI 30–35 kg/m2 the risk was 1.67 (95% CI 1.24–2.25) for CHD and 2.3 (95% CI 1.56–3.40) for CVD mortality, compared to normal weight. Each 5 kg/m2 increase in BMI was associated with 24% (95% CI 16–33%) higher risk for all-cause mortality.

This meta-analysis suggests that physical activity and moderate alcohol intake were associated with a reduced risk for CVD and mortality. Smoking and higher BMI were associated with an increased risk of these endpoints. Adherence to a healthy lifestyle may substantially lower the burden of CVD and reduce the risk of mortality among middle-aged and elderly women. However, this review highlights important gaps, as lack of standardized methods in assessing lifestyle factors and lack of accurate information on menopause status, which should be addressed by future studies in order to understand the role of menopause on the association between lifestyle factors and cardiovascular events.

Forum Comments

Most epidemiologic studies have been consistent in demonstrating that subjects who do not smoke, are not obese, are physically active, and consume small to moderate amounts of alcohol have significantly less cardiovascular disease (CVD) and have a lower risk of total mortality. The beneficial effect of each is usually highly significant, and the net health effect of all such factors is remarkable.  The majority of studies have included younger subjects, and more often the studies have been based on males.  There are much less data on the effects of lifestyle on elderly women.

The present paper reports a well-done meta-analysis on a very large number of middle-aged or elderly women, based on studies that total more than five million subjects. It concludes that significant and large lowering of risk occurs among women who do not smoke, are not obese, are physically active, and drink alcohol moderately. It emphasizes how lifestyle factors can play a major role in terms of the risk of CVD and death for middle-aged/elderly women.

Comments by individual Forum members: Reviewer Ellison pointed out some strengths of the analysis: “It is based only on cohort studies, essentially all of which were classified as high-quality studies; it provides a large amount of data for women following menopause, when their risk of CVD increases markedly; and there was a very large number of cases of CVD within the cohort: more than 10,000 subjects developed CVD, almost 28,000 had coronary disease, more than 15,000 had a stroke, and total mortality was recorded for almost 200,000 women.

“To help assure that the effects were applicable for post-menopausal women, the authors carried out sub-analyses limited to women 50 or more years of age; the results in more than 200,000 such women who were moderate drinkers showed very similar results as those reported for the total group of subjects.” Reviewer Thelle stated: “I support these comments.  I would add that the main purpose of the paper was not just about alcohol, women and coronary heart disease (CHD), but a more general attempt to elucidate CVD risk factors and behaviours regarding women and cardiovascular risk.”  Reviewer Finkel added:  “This paper reports again on the benefits and liabilities of common lifestyle factors, with the expected conclusions.  The focus on women in their second half of life is helpful, and politically comforting, though I don’t know that it adds to our basic understanding of the association.”

As for the effects of alcohol on CVD and mortality, Forum members noted that in their comparisons with non-drinkers, the authors used < 98 g/week (about one drink/day) as the upper limit of “moderate,” and > 98 g/week of alcohol as “high intake;” they later report that “high intake” was 98-322 g/week of alcohol in their studies. Thus, their definition of “moderate” falls within the US Dietary Guidelines of no more than an average of one drink/day for women.  The authors report that, within the total studies reviewed, 17 articles evaluated alcohol intake, comprising 880,834 women.  Compared with non-drinkers, moderate drinkers had a reduced risk of 0.72 for fatal and non-fatal CHD, 0.63 for CVD mortality, and 0.80 for all-cause mortality; all relations were statistically significant.  For about 90,000 women who were reported to have “high alcohol intake,” the authors state that two studies showed 30% higher CVD mortality (not significant) and one study showed 7% lower all-cause mortality (also not significant).

Forum member Skovenborg commented: “The authors do not mention the possible effect of alcohol consumption on median age at menopause, which has been described by Taneri et al and Kinney et al.  The latter noted: ‘The estimated median age at menopause was 2.2 (95% CI 0.5, 3.9) years later for women who drank alcohol 5–7 days/week (13% of the sample) than for women who did not drink alcohol (54%).  For women who drank at least 1 day/week, the estimated shift was 1.3 (95% CI 0.2, 2.3) years.’”  It is unclear to what extent the differences in age at menopause may have affected the results in the present study.

Other research on the effects of a “healthy lifestyle”: Forum member Stockley described a number of specific previous studies on a healthy lifestyle that support the findings of the present analysis. She states: “For example, Ford et al’s study of 16,958 US individuals followed for 18 years by the US Centers for Disease Control and Prevention (CDC) examined the relationship between four low-risk behaviors and mortality where ‘moderate consumption of alcohol’ (≤ 2 drinks/day for men and ≤ 1 drink/day for women) was considered as one of ‘four healthy lifestyle behaviors that exert a powerful and beneficial effect on mortality.’  The other low-risk behaviors were non-smoking, eating a healthy diet, and physical activity.  Ford et al stated that: ‘The number of low-risk behaviors was inversely related to the risk for mortality.’  Compared with participants who had no low-risk behaviors, which included abstinence from alcohol as well as excessive alcohol consumption, those who had all four experienced significantly reduced all-cause mortality and had an average lifespan of 11.1 years longer!

“An earlier study by Chiuve et al also included light-to-moderate alcohol consumption (5 to 30 g/day) as one of five low-risk behaviors associated with a reduced risk of coronary heart disease irrespective of concurrent medication for hypertension or hypercholesterolemia. These behaviours was based on the Healthy Eating Index (HEI), created by the US Department of Agriculture; the HEI defined moderate alcohol consumption of 1.5 to 2.5 drinks/day as ideal servings for men and 0.5 to 1.5 drinks/day as ideal for women on the basis of the lower risk of cardiovascular disease associated with moderate alcohol consumption (McCullough et al).

“A similar, little publicised Australian study of 7989 individuals aged 65-83 years followed for five years showed consistent results with this CDC study (Spencer et al). The eight selected low-risk behaviors included having no more than two alcoholic (total 20 g alcohol) drinks/day. Individuals with five or more of the selected low-risk behaviors had a lower risk of death from any cause within five years compared with those having less than five. In addition, Lee et al showed that although light-to-moderate drinkers may have better risk factor profiles than non-drinkers, including higher socioeconomic status and fewer functional limitations (such as activities of daily living, instrumental activities of daily living and mobility), which explain some of the survival advantage associated with alcohol consumption, light-to-moderate drinkers still maintain their survival advantage even after adjustment for these factors.

“Further, Sun et al showed that in addition to lower mortality, women who were moderate alcohol consumers surviving to age 70 years and older generally had less disability and disease, and more signs of ‘successful ageing.’ For ‘regular’ light-to-moderate alcohol consumers (on 5-7 days/week), there was an approximately 50% greater chance of such successful ageing compared with non-drinkers.”

Do changes in alcohol consumption result in increase/decrease in risk of CVD or other diseases?  Forum member Waterhouse noted: “This study confirms the general effect of alcohol on CVD observed in numerous other reports, but is interesting because of the focus on post-menopausal women.  However, the authors state that they undertook the study in order to provide this population with data on the value of healthy lifestyle decisions, presumably to encourage change towards beneficial practices.  Unfortunately, these recommendations are compromised because they did not look at how changing drinking, smoking, or exercise would affect outcomes.  I understand that such data would be much more sparse or not available, but my concern stands.  I think it would be appropriate to suggest that future studies look at how changes in lifestyle factors alter health outcomes after some time has passed.”

Other Forum members agreed that it would be important to evaluate changes in alcohol during middle-age or later for their effect on health; however, change is more difficult to assess in epidemiological studies, and the reason that certain people increase or decrease their alcohol intake is rarely known.

Ellison noted that many animal experiments demonstrate that the administration of alcohol, wine, or wine polyphenolics modify CVD risk factors in the expected direction. As for short-term clinical trials in humans, an early paper by Rimm et al in 1999 summarized trials of alcohol administration, and showed that alcohol resulted in increased HDL-cholesterol, apolipoprotein A, and showed a tendency for improvement in coagulation factors.  A later, large meta-analysis of intervention studies by Brien et al had similar conclusions: “Moderate alcohol consumption had favourable effects on levels of high density lipoprotein cholesterol, apolipoprotein A1, adiponectin, and fibrinogen.”  Both studies concluded that such trials support the observational findings that alcohol consumption is a causative factor in reducing the risk of CVD.

As for effects of changes in alcohol intake on blood pressure, in a meta-analysis of clinical trials Xin et al reported: “Overall, alcohol reduction was associated with a significant reduction in mean systolic and diastolic blood pressures of −3.31 mmHg and −2.04 mmHg, respectively. Effects of intervention were enhanced in those with higher baseline blood pressure.”  In fact, these authors also state: “The participants in the 15 trials we studied tended to be fairly heavy alcohol drinkers (≥ 3 drinks/d). Therefore, we were not able to examine the effect of moderate alcohol consumption on BP.”

For observational data from cohort studies, changes in intake among healthy people who do not abuse alcohol are limited. In the Health Professionals’ Study, Joosten et al reported from repeated assessments for more than 400,000 subject-years of follow up, stating: “A 7.5 g/day (approximately half a glass) increase in alcohol consumption over 4 years was associated with lower diabetes risk among initial nondrinkers (HR 0.78) and drinkers initially consuming < 15 g/day (HR 0.89), but not among men initially drinking ≥ 15 g/day (HR 0.99).”

In the ARIC study, King et al reported: “Of 7,697 participants who had no history of cardiovascular disease and were non-drinkers at baseline, within a 6 year follow-up period 6.0% began moderate alcohol consumption (2 drinks per day or fewer for men, 1 drink per day or fewer for women) and 0.4% began heavier drinking. After 4 years of follow up, new moderate drinkers had a 38% lower chance of developing cardiovascular disease than did their persistently non-drinking counterparts. This difference persisted after adjustment for demographic and cardiovascular risk factors (OR 0.62, 95% CI 0.40-0.95).”  For total mortality in this study, there was a similar decrease in risk between the new drinkers and persistent non-drinkers (OR 0.71, CI 0.31-1.64), but the difference was not statistically significant.  In another study (Roerecke et al) alcohol abusers who reported a reduction in alcohol consumption showed a lower risk of death, as would be expected.

Confirmation of observational studies by clinical trials: Skovenborg also noted that the authors do not mention one of the most commonly cited examples of why some believe that results from observational studies should not be trusted, as their results may not be confirmed by intervention trials.  He referred to a 2018 publication in the British Medical Journal, in which Mozaffarian and Forouhi state: “One of the most commonly cited examples of why observational studies should not be trusted comes outside nutrition. Observational studies found lower rates of heart disease in women taking hormone replacement therapy, but the Women’s Health Initiative trial found the opposite.  This was widely thought to show insurmountable limitations of observational research.  Yet, additional follow-up in the trial showed benefit in younger women (who were most representative of the observational cohorts) and not in older women who had been enrolled to increase statistical power because of their higher risk of heart disease. This concordance between observational and interventional findings, which was also striking for other clinical endpoints, has been largely overlooked.  Systematic comparisons, including up to 1,583 meta-analyses of 228 conditions, find similar close concordance between randomised trials and observational studies (Concato et al, Anglemyer et al).”

References from Forum critique

Anglemyer A, Horvath HT, Bero L. Healthcare outcomes assessed with observational study designs compared with those assessed in randomized trials. Cochrane Database Syst Rev 2014;(4):MR000034.24782322.

Brien SE, Ronksley PE, Turner BJ, Mukamal KJ, Ghali WA. Effect of alcohol consumption on biological markers associated with risk of coronary heart disease: systematic review and meta-analysis of interventional studies. BMJ 2011;342:d636.  doi:10.1136/bmj.d636

Chiuve SE, McCullough ML, Sacks FM, Rimm EB. Healthy lifestyle factors in the primary prevention of coronary heart disease among men: benefits among users and nonusers of lipid-lowering and antihypertensive medications. Circulation 2006;114:160–167.

Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med 2000;342:1887-1892.  10.1056/NEJM200006223422507 10861325.

Ford ES, Zhao G, Tsai J, Li C. Low-Risk Lifestyle Behaviors and All-Cause Mortality: Findings From the National Health and Nutrition Examination Survey III Mortality Study.  Am J Public Health 2011;101:1922–1929. doi: 10.2105/AJPH.2011.300167

Joosten MJ, Chiuve SE, Mukamal KJ, Hu FB, Hendriks HFJ, Rimm EB. Changes in Alcohol Consumption and Subsequent Risk of Type 2 Diabetes in Men. Diabetes 2011;60:74–79.

King DE, Mainous III, AG, Geesey ME. Adopting Moderate Alcohol Consumption in Middle-age: Subsequent Cardiovascular Events.  Am J Med 2008;121:201–206.

Kinney A et al. Alcohol, caffeine and smoking in relation to age at menopause. Maturitas 2006;54:27-38.

Lee SJ, Sudore RL, Williams BA, Lindquist K, Chen HL, Covinsky KE. Functional Limitations, Socioeconomic Status, and All‐Cause Mortality in Moderate Alcohol Drinkers. J Am Geriatr Soc 2009;57:955-962.

McCullough ML, Feskanich D, Stampfer MJ, Giovannucci EL, Rimm EB, Hu FB, Spiegelman D, Hunter DJ, Colditz GA, Willett WC. Diet quality and major chronic disease risk in men and women: moving toward improved dietary guidance.  Am J Clin Nutr 2002;76:1261-1271.

Mozaffarian D, Forouhi NG. Dietary guidelines and health — is nutrition science up to the task? BMJ 2018;360:k822 doi: 10.1136/bmj.k822

Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ. Moderate alcohol intake and lower risk of coronary heart disease: meta­analysis of effects on lipids and haemostatic factors.  BMJ 1999;319:1523–1528.

Roerecke M, Gual AO, Rehm J. Reduction of alcohol consumption and subsequent mortality in alcohol use disorders: systematic review and meta-analyses.  The Journal of Clinical Psychiatry 2013;74:e1181-1189.

Spencer CA, Jamrozik K, Norman PE, Lawrence-Brown M. A simple lifestyle score predicts survival in healthy elderly men. Prev Med 2005;40:712–717.

Sun Q, Townsend MK, Okereke OI, Rimm EB, Hu FB, Stampfer MJ, Grodstein F. Alcohol Consumption at Midlife and Successful Ageing in Women: A Prospective Cohort Analysis in the Nurses’ Health Study.  PLoS Med 2011;8:e1001090.  doi.org/10.1371/journal.pmed.1001090

Taneri PE, Kiefte-de Jong JC, Bramer WM, Daan NM, Franco OH, Muka T. Association of alcohol consumption with the onset of natural menopause: a systematic review and meta-analysis.  Lifestyle factors, cardiovascular disease and all-cause mortality in middle-aged and elderly Hum Reprod Update. 2016;22(4):516–28. https://doi.org/10. 1093/humupd/dmw013.

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;38:1112-1117. doi.org/10.1161/hy1101.093424.

Forum Summary

Overall, Forum members considered this to be a well-done paper that provides additional information on alcohol’s effects on CVD and mortality among a very large number of middle-aged or elderly women. Although women markedly increase their risk of CVD following menopause, this is a group of women for whom data are sparse.  Weaknesses of the analysis include a lack of information on the pattern of drinking (binge versus regular), type of beverage, and drinking alcohol with or without food.  It is of note that in some of the studies included in the meta-analysis, factors such as HDL-cholesterol and fibrinogen were adjusted for in the study; given that these are mechanisms of alcohol’s effects, overall results will surely underestimate the effects.  Further, as the authors point out, this study provides little information on the physiologic mechanisms of alcohol’s effects on health.

Forum members praised the authors for this meta-analysis of a very large dataset. It strengthens previous analyses that show that each of the lifestyle factors considered – not smoking, not becoming obese, being physically active, and regularly consuming small to moderate amounts of alcohol – benefit health.  The more healthy lifestyle factors a subject has, the greater the benefit.

It will be especially important that future research evaluates what happens if subjects change their lifestyle habits, as such results would help target specific behavioral changes.  There are very limited data available on this at present, but such studies that have been done suggest that the onset of moderate drinking among non-drinkers, or a slight increase in alcohol consumption among light drinkers, have beneficial effects on health.  Still, it is appreciated that moderate alcohol consumption is only one of a number of factors making up a “healthy lifestyle,” which is clearly associated with less cardiovascular and other diseases and a much lower risk of mortality.

*                  *                      *                      *                      *                      *                      *

Comments on this critique by the International Scientific Forum on Alcohol Research were provided by the following members:

Andrew L. Waterhouse, PhD, Department of Viticulture and Enology, University of California, Davis, USA

Susan J van Rensburg, PhD, Department of Pathology, Stellenbosch University, Tygerberg, South Africa

Dag S. Thelle, MD, PhD, Department of Biostatistics, Institute of Basic Medical Sciences, University of Oslo, Norway; Section for Epidemiology and Social Medicine, Sahlgrenska Academy, University of Gothenburg, Sweden

Creina Stockley, PhD, MSc Clinical Pharmacology, MBA; Health and Regulatory Information Manager, Australian Wine Research Institute, Glen Osmond, South Australia, Australia

Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark

Harvey Finkel, MD, Hematology/Oncology, Retired (Formerly, Clinical Professor of Medicine, Boston University Medical Center, Boston, MA, USA)

R. Curtis Ellison, MD, Professor of Medicine, Section of Preventive Medicine & Epidemiology, Boston University School of Medicine, Boston, MA, USA