Original article
Features of excessive alcohol drinking in older adults distinctively captured by behavioral and biological screening instruments: An epidemiological study

https://doi.org/10.1016/S0895-4356(01)00408-5Get rights and content

Abstract

The entire ⩾65-year-old population living in a small Italian town, where alcohol use is almost ubiquitous, was assessed with a frequency–quantity questionnaire for alcohol intake and with two screening instruments for alcohol problems, the CAGE questionnaire and the MCV-γGT test. Aim of the study was to assess whether these instruments identify different subsets of subjects with alcohol problems. Of the 649 participants, 19.1% were at-risk drinkers (average intake >40 g/day in men and >20 g/day in women). Both the screening instruments were positive in only a minority of participants. Of the 377 drinkers, 53 gave ⩾1 affirmative response to the CAGE questionnaire, whereas 24 had a positive MCV-γGT test. The concordance between positive CAGE questionnaire and MCV-γGT test was limited to seven subjects (kappa = 0.10), and these tests identified subjects who differed for several health and psychosocial characteristics. Participants aged ⩾75 years drank less, but had similar prevalence of CAGE and MCV-γGT positive markers as compared to younger participants. In conclusion, excessive drinking is common in the elderly. Screening tests based on behavioral and biological markers identify two different sets of subjects with possible alcohol problems. This might indicate the opportunity to use these instruments in conjunction.

Introduction

In recent years, a growing amount of interest has been devoted to alcohol problems in older persons as a relevant public health issue. The prevalence of formal, major alcohol disorders, such as abuse or dependence, is greater in young- and middle-aged than in older individuals 1, 2, yet the risk of dependence for a given alcohol intake increases with advancing age [3]. Furthermore, alcohol problems include (beyond abuse and dependence) a broad spectrum of physical, psychological, and social disorders, which in older persons are often erroneously attributed to other age-related conditions. Therefore, excessive alcohol consumption frequently remains an unrecognized cause of mortality and morbidity of older persons [4]. It can be argued that alcohol problems are more likely to be overlooked in populations where alcohol drinking is ubiquitous and socially accepted, as in certain regions of Italy 5, 6.

Primary care physicians are well positioned to identify subclinical alcohol disorders in the context of patient contact for other problems, and they are prompted to routinely assess patients about alcohol use and misuse [7]. Nevertheless, many patients with at-risk or harmful drinking habits remain undiagnosed by their physicians [8]. A recent survey reported that only 47% of primary care physicians regularly ask patients about the maximum alcohol intake on a single occasion, and only 13% of them use formal screening tools for alcohol problems [9]. Screening for alcohol problems is usually based on the detection of specific behavioral characteristics with structured interviews, among which the CAGE questionnaire has been extensively used in older adults 10, 11. A different approach is offered by laboratory tests for the metabolic effects of alcohol, among which measurement of mean corpuscular volume (MCV) and gamma-glutamyl-transpeptidase (γGT) is low-cost and usually performed in clinical practice. Both MCV and γGT increase in patients with excessive alcohol intake and their simultaneous elevation has been used as a screening tool in epidemiological settings [12]. To our knowledge, the performance of these two screening approaches has not been compared in older persons living in the community. Due to the multifaceted profile of alcohol problems, screening tools for this condition should be compared not only for their performance versus a common reference standard (criterion validity) 13, 14, but also in terms of specific domains covered by each instrument (construct validity). If instruments that explore different aspects of alcohol problems have a poor concordance, this might suggest the need to use them in conjunction.

This study was carried out to verify the hypothesis that behavioral and biological markers identify distinct subsets of older individuals with alcohol problems. Data used in this analysis were collected in a cross-sectional, epidemiological survey, primarily focused on heart failure, which included all the community-dwelling individuals 65 years of age and over resident in Dicomano, a small rural town near Florence, Italy (“Insufficienza Cardiaca negli Anziani Residenti a Dicomano,” ICARe Dicomano Study) [15]. In this geographic area, production of red wine is extensive and per capita level of alcohol intake is one of the highest in the nation [16].

Section snippets

Study protocol

The ICARe Dicomano Study, whose general design has been detailed elsewhere [15], enrolled the community-dwelling, elderly (⩾65 years) population recorded in the City Registry Office on April 1995. The only exclusion criterion was living in a nursing home. According to the original study protocol [15], multidimensional, geriatric assessment [17] data were collected with home interview, biological testing and clinical examination, after informed consent. Proxy interviews were obtained for

Results

Data on drinking behavior were available in 649 (262 men, 387 women) of 864 subjects originally eligible for the ICARe Dicomano study [15]. Of the 215 subjects not included, 4 died before study onset, 163 refused to participate, 48 had proxy or incomplete interviews. Non-participants tended to be older (age: 74.9 ± 0.5 years) than participants (73.8 ± 0.3 years; P = 0.053) and they were predominantly males (non-participating men: 51.2% vs. non-participating women: 48.8%; P = 0.006). According

Discussion

Depending on the definition considered, the prevalence of at-risk drinkers in Dicomano ranged from 8.9% to 19.1%, higher than in previous studies 2, 25. This result is a further confirmation [26] of the “single population theory” [27], postulating that average alcohol consumption and prevalence of heavy drinkers are closely related. Indeed, among Italian regions, Tuscany ranks third for wine consumption and Dicomano lies within a wine-producing area, in which the highest Italian mean alcohol

Acknowledgements

This project was granted in part by the Italian Ministry of the University and Scientific and Technological Research (National Special Project on Heart Failure) and by the Regional Government of Tuscany. We are indebted to Marco Pahor, MD and Paul E. McGann, MD for their contribution in the preparatory phase of the manuscript. We also thank Miss Tonya Holmes for her secretarial assistance.

References (33)

  • A. Fink et al.

    Alcohol-related problems in older persons. Determinants, consequences, and screening

    Arch Intern Med

    (1996)
  • Gruppo Epidemiologico della Società Italiana di Alcologia

    Bollettino Epidemiologico della Società Italiana di Alcologia

    Alcologia

    (1995)
  • National Institute on Alcohol Abuse and Alcoholism. Eighth Special Report to Congress on Alcohol and Health. Rockville,...
  • R.D. Moore et al.

    Prevalence, detection, and treatment of alcoholism in hospitalized patients

    JAMA

    (1989)
  • P.D. Friedmann et al.

    Screening and intervention for alcohol problems. A national survey of primary care physicians and psychiatrists

    J Gen Intern Med

    (2000)
  • J.A. Ewing

    Detecting alcoholismthe CAGE questionnaire

    JAMA

    (1984)
  • Cited by (24)

    • Geriatric alcoholism: Pathophysiology and dental implications

      2006, Journal of the American Dental Association
      Citation Excerpt :

      These markers include elevation in the level of the liver enzyme gamma-glutamyltransferase, increased mean volume of red blood cells (mean corpuscular volume), elevation in the level of high-density lipoprotein cholesterol and an increase in the level of carbohydrate-deficient transferrin. The sensitivity and specificity of these tests, however, is relatively low.43–45 A sizable number of physicians and some dentists use health questionnaires that are designed specifically to help identify older people who may be using excessive amounts of alcohol.46,47

    • The Bayesian Alcoholism Test had better diagnostic properties for confirming diagnosis of hazardous and harmful alcohol use

      2005, Journal of Clinical Epidemiology
      Citation Excerpt :

      Also BAT might be less accurate in diagnosing HHAU populations with different demographics, e.g. in other age groups, such as seniors or adolescents. Research indicates that the sensitivity of biochemical markers drops dramatically in young alcoholics [59–61] while in older alcoholics they may select a different set of subjects [62]. Second, the external validity of our study must be considered.

    View all citing articles on Scopus
    View full text