of diabetes rates could be used as a method. for detecting community exposure to critical admissions for diabetes varied substantially. among several areas …
Clinical Research and Methods
Vol 37, No 8
589
Association Between Alcohol Consumption and Diabetes Preventive Practices
Lisa D Chew, MD, MPH; Karin M Nelson, MD, MSHS; Bessie A Young, MD, MPH; Katharine A Bradley, MD, MPH
Background and Objectives: Little is known about the effect of alcohol consumption on the quality of care among patients with diabetes We evaluated the association between alcohol consumption and diabetes preventive practices Methods: We analyzed data from the 2001 Behavioral Risk Factor Surveillance System Based on self-reported alcohol consumption in the past 30 days, we categorized participants into the following groups: nondrinkers, moderate drinkers average drinks/day: one for women, two for men, or heavy drinkers average drinks/day: one for women, two for men We then examined the association between alcohol consumption and participants reports of diabetes preventive practices Results: Of 10,980 respondents with a self-reported diagnosis of diabetes, 70 were current nondrinkers, 28 moderate drinkers, and 2 heavy drinkers Heavy drinkers compared with nondrinkers were more likely to report not performing daily glucose self monitoring and not having had
an eye examination in the past year, adjusting for age, gender, race, education, income, marital status, health insurance, diabetes duration, health status, and insulin use Moderate drinkers were more likely than nondrinkers to report not performing daily glucose self monitoring and not having had a provider visit for diabetes in the past year Conclusions: Adults with diabetes who report moderate or heavy alcohol consumption may be at risk for adverse diabetes outcomes due to suboptimal preventive practices Fam Med 2005;378:589-94
Although heavy alcohol consumption is a recognized cause of morbidity and mortality,1-4 few studies have addressed the potential influence of alcohol consumption on self-management of chronic illnesses Diabetes mellitus in particular requires complex self-management skills and close follow-up by health care professionals for successful management Preventive care practices can help reduce both the incidence and the progression of diabetic complications5-8 The association between alcohol and decreased risk of coronary disease9,10 has resulted in recommendations that patients with diabetes can drink within recommended limits11 However, excessive alcohol
intake may alter patient behavior and interfere with chronic disease self-management Alcohol use has been associated with poorer compliance with diabetes preventive care practices in selected populations12-14 For example, increased alcohol use was associated with poorer compliance with insulin injections among patients at a Veterans AdministraFrom the VA Puget Sound Health Care System, Seattle, Wash; and the Department of Medicine, University of Washington
tion diabetes clinic12 Any alcohol use was associated with poorer adherence to exercise recommendations, oral medication regimens, and outpatient visits among patients with diabetes seen at community clinics13 However, these studies used clinic-based, convenience samples with limited generalizability A study of a large managed care organization found that heavy alcohol consumption was a significant predictor of lessfrequent glucose monitoring,14 but other recommended diabetes preventive care practices were not assessed A recent systematic review of the effect of alcohol consumption on diabetes mellitus recommended that further studies on the effect of alcohol use on diabetes self-care behaviors is needed15 This studys purpose
was to evaluate the association between levels of reported alcohol consumption and diabetes quality of care indicators in a large US general population sample of adults with diabetes Specifically, we assessed the level of compliance with diabetes quality of care indicators established by Healthy People 2010 using data from the 2001 Behavioral Risk Factor Surveillance System BRFSS
590
September 2005
Family Medicine and Human Services25 Six components of diabetes preventive practices were assessed: 1 daily glucose self monitoring, 2 at least one hemoglobin A1c test HbA1c during the past year, 3 a foot examination within the past year by a health professional, 4 a dilated eye examination in the past year, 5 formal diabetes education, and 6 at least one visit to a health care provider for diabetes care in the past year Age, gender, race/ethnicity, education, and marital status were used as indicators of sociodemographic status, and health insurance coverage was used as an indicator of access to health care Health insurance coverage was defined from responses to the question, Do you have any kind of health care coverage, including health insurance, prepaid plans such as health
maintenance organizations, or government plans such as Medicare? Because disease severity may influence health care utilization and preventive care practices, duration of diabetes and insulin use were considered indicators of diabetes severity Health status was measured using the single question, Would you say that in general your health is excellent, very good, good, fair, or poor?26-29 Finally, because depression may influence alcohol consumption and diabetes outcomes and we did not have a direct measure of mental illness, we used the number of self-reported mental unhealthy days as a proxy of mental distress from the question, Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?30 All analyses accounted for the complex sampling scheme, the unequal probability of selection, and the oversampling of certain demographic groups The data presented were weighted to the age, gender, and race distributions of the state-specific populations to obtain estimates representative of the adult population in our sample Differences in categorical and continuous variables
were assessed using the chi-square test and the Student t test, respectively The odds of diabetes preventive care across alcohol consumption categories were assessed using logistic regression Multiple logistic regression analyses were used to evaluate the association of alcohol consumption with diabetes quality-of-care indicators, after adjusting for all measured confounders All analyses were performed using STATA-SE 7 Stata Statistical Software: Release Special Edition 70 College Station, Tex: Stata Corporation, 2002 BRFSS data are in the public domain; therefore, the study was exempt from review by the University of Washington Institutional Review Board Results Of 212,510 respondents in the 41 states and the District of Columbia, 10,980 52 reported diabetes and met our inclusion criteria Seventy percent n7,923
Methods The BRFSS is an annual state-specific, cross-sectional telephone survey of the civilian, noninstitutionalized adult population age 18 Funded by the Centers for Disease Control and Prevention, the survey is conducted in all 50 states, the District of Columbia, Puerto Rico, and Guam Detailed descriptions of the purpose, survey methodology, and data analysis have
been described previously16, 17 Briefly, the BRFSS selects state-specific probability samples of households using a multistage cluster-design procedure16, 17 In all states, trained interviewers randomly select one adult from each household and administer a standardized core questionnaire to that individual Each interview lasts approximately 15 minutes In 2001, BRFSS collected data on clinical characteristics and diabetes-specific preventive care practices from respondents with diabetes in 41 states and the District of Columbia BRFSS respondents with self-reported diabetes who lived in states that administered the diabetes-specific module were eligible for this study All respondents were asked, Have you ever been told by a doctor that you have diabetes? If they answered affirmatively, they were considered to have diabetes Individuals with self-reported gestational diabetes were excluded The BRFSS measures used to assess patient sociodemographic characteristics and whether patients have diabetes have good-to-excellent reproducibility and validity18-20 The alcohol section of the 2001 BRFSS questionnaire began with a question about how often those surveyed consumed an
alcohol-containing beverage with one drink defined as one can or bottle of beer, one glass of wine, one can or bottle of wine cooler, one cocktail, or one shot of liquor in the past 30 days Those who reported having alcohol in the past 30 days were then asked how many drinks they typically had on days that they drank The average number of drinks per day was derived by calculating the total number of drinking days over the past month divided by 30, multiplied by the average number of drinks per occasion Based on established convention, we categorized participants into three mutually exclusive groups based on average drinks per day: nondrinkers, moderate drinkers, and heavy drinkers Nondrinkers were those who reported no alcoholic beverages in the past 30 days Moderate drinkers were defined as consuming two alcoholic drinks or less per day on average for a man and one drink per day or less on average for women Heavy drinkers were defined as men who consumed more than two alcohol drinks per day on average and women who consumed more than one drink per day on average21-24 We defined quality of diabetes preventive care based on objectives defined by the Healthy People 2010: Diabetes
report issued by the US Department of Health
Clinical Research and Methods were classified as nondrinkers, 28 n2,859 as moderate drinkers, and 2 n198 as heavy drinkers Heavy drinkers were more likely to be young, male, employed, and smokers P001 Table 1 Nondrinkers were more likely to report their general health as fair or poor compared to moderate and heavy drinkers P001 Self-reported annual HbA1c monitoring, annual foot examination, and formal diabetes education varied little across alcohol consumption categories There was a significant association based on a test for trend P001 between increasing alcohol consumption and not having performed daily glucose self monitoring,
Vol 37, No 8
591
not having received an annual dilated eye examination, and not having had an annual provider visit for diabetes Table 1 To identify other factors associated with nonadherence, we looked at the relationship between patient characteristics and nonadherence to the diabetes preventive practices In bivariate analyses, patient characteristics significantly associated with increased odds of nonadherence to at least one of the six components of diabetes preventive practices were non-white race, no
insulin use, educational attainment less than high school, annual income less than 15,000, less than 5 years with diabetes, and no health care coverage
Table 1 Characteristics of Diabetic Adults by Alcohol Consumption Status
Characteristics Sample, n Population estimates, n Mean age standard error, years Male Race White Black Hispanic Other Education high school Annual income 15,000 15,00035,000 35,000 Missing Married Employed Current smoker Has health care coverage Duration of diabetes 5 years 614 years 15 years Insulin use General health Excellent/very good Good Fair/poor Mental unhealthy days 14 Diabetes care practices No daily glucose self monitoring No annual HbA1c monitoring No annual foot exam No annual dilated eye exam No formal diabetes education No annual provider visit Percentages are weighted estimates P value for diabetes care practices based on test for trend All Diabetics 10,980 11,666,951 588 02 48 67 13 13 6 23 17 31 34 17 61 37 17 90 45 30 25 27 18 34 48 14 45 13 32 27 48 10 Categories of Alcohol Consumption Nondrinkers Moderate Heavy 7,923 2,859 198 7,966,611 601 03 42 65 15 13 7 26 21 33 27 19 60 32 15 90 42 31 27 29 15 32 53 15 42 12 32 26 48
8 2,047,085 566 05 64 71 9 14 5 14 9 28 50 13 66 48 21 89 51 28 21 22 25 40 35 11 50 14 34 29 48 15 152,400 510 19 72 70 10 13 7 17 6 35 50 9 56 52 46 88 51 37 12 21 26 32 42 14 57 15 37 46 52 15 001 001 001 P Value
001 001
001 001 001 63 001
001 001
04 001 14 001 11 71 001
592
September 2005
Family Medicine than nondrinkers to not perform daily glucose self monitoring and to not have had an annual provider visit for diabetes Table 3 Because depression may be associated with both alcohol consumption and diabetes outcomes, we repeated the same multivariate analyses and adjusted the numbers of self-reported mental unhealthy days as a proxy for depression Additional adjustment for the number of mental unhealthy days did not significantly alter our findings Discussion In these national data, adults with diabetes who reported moderate or heavy alcohol consumption were less likely than nondrinkers to report daily glucose self monitoring Heavy drinkers compared to nondrinkers were also more likely to report not receiving annual dilated eye examinations, while moderate drinkers compared to nondrinkers were more likely to report not having had an annual provider visit for diabetes
Our results are consistent with previous studies that examined the association between alcohol use and diabetes preventive care in select clinical populations There are several possible explanations for the observed associations between any alcohol use and diabetes preventive care Excessive alcohol intake may cause patients to be less attentive to their health problems, leading to fewer health care visits and interfering with self-care practices31 Because some diabetic patients may have been advised to abstain, continued drinking may reflect lack of control over alcohol use and potentially alcohol dependence Further, when alcohol problems are identified, clinicians may not have adequate time to address additional diabetes preventive or self-care practices because of the need to discuss alcohol use Finally, increased alcohol consumption may be a marker for other underlying patient factors such as mental health conditions or lack of self-efficacy, which may in turn affect adherence Limitations This study has several limitations The cross-sectional observational design of the study prevented us from being able to assess causality Despite the large, representative sample of the US
population, individuals without telephones were not included, and poorer households were likely underrepresented While we could account for duration of diabetes and insulin use, other measures of diabetes severity and comorbidity that may influence the receipt of preventive care services were not available The determinants of preventive diabetes care are multiple and complex; we were able to roughly account for mental illness such as depression or anxiety that may be associated with both alcohol consumption and diabetes care
Table 2 These results were similar when we evaluated the association between patient characteristics and each diabetes preventive practice separately Compared to nondrinkers, heavy drinkers were at an increased risk of not performing daily glucose self monitoring and not having had an annual dilated eye examination in multivariate analysis that adjusted for age, gender, race, education, marital status, income, health care coverage, duration of diabetes, health status, and insulin use Table 3 After adjusting for similar covariates, moderate drinkers were more likely Table 2 Crude Odds Ratio OR of Nonadherence to At Least One of the Six Diabetes Preventive
Practices
Characteristics Age years 1845 4664 65 Gender Male Female Race White Not white Education High school High school Income 15,000 15,000 Married Yes No Employed No Yes Has health care coverage Yes No Duration of diabetes 5 years 5 years Insulin use Yes No General health Excellent/very good/good Fair/poor Mental unhealthy days 14 No Yes Alcohol consumption Nondrinker Moderate Heavy CI–confidence interval Crude OR 95 CI 10 09 0712 12 0914 10 12 1014 10 13 1115 10 17 1422 10 15 1218 10 11 0913 10 08 0710 10 21 1629 10 18 1522 10 38 3145 10 08 0710 10 08 0611 10 10 0812 20 1132
Clinical Research and Methods
Vol 37, No 8
593
Table 3 Crude and Adjusted OR 95 CI of Nonadherence to Diabetes Preventive Practices Among Participants by Drinking Categories
No daily glucose self monitoring Crude OR Adjusted OR No annual HbA1c monitoring Crude OR Adjusted OR No annual professional foot exam Crude OR Adjusted OR No annual dilated eye exam Crude OR Adjusted OR No formal diabetes education Crude OR Adjusted OR No annual provider visit Crude OR Adjusted OR Referent group: nondrinkers n7,923 Adjusted for age, gender, race, education, marital status, income, health care coverage,
duration of diabetes, health status, and insulin use OR–odds ratio CI–confidence interval Moderate Drinker n2,859 14 1216 13 1115 12 0916 11 0814 11 0913 11 0913 12 1014 12 1014 10 0912 10 0912 21 1727 18 1424 Heavy Drinker n198 18 1129 18 1129 13 0728 13 0627 13 0821 14 0823 24 1538 22 1435 12 0819 13 0820 21 1237 17 0932
Conclusions Heavy alcohol use may be a modifiable behavior affecting diabetes care quality and a barrier to achieving the Healthy People 2010 diabetes objectives Longitudinal studies are needed to evaluate the association of alcohol use with objective quality of care and patient outcomes for diabetes In the meantime, clinicians should be aware that patients with diabetes who report drinking alcohol may be at risk for adverse diabetes outcomes due to suboptimal preventive care practices
Acknowledgments: This study was presented at the 2003 Society of General Internal Medicine National Meeting in Vancouver, British Columbia We thank Dr Edward Boyko for his critical review of the manuscript Views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs, the University of Washington, the
National Institute of Alcohol Abuse and Alcoholism, or the Robert Wood Johnson Foundation Funding source: During the time that this work was completed, Dr Chew was an investigator at the VA Health Services Research and Development of the VA Puget Sound This material is the result of work supported with resources from and the use of facilities at VA Puget Sound, Seattle, Wash Dr Young is supported by the American Diabetes Association Career Development Award and is a Robert Wood Johnson Minority Medical Faculty Development Fellow Dr Bradley is an investigator at the VA Puget Sound Health Care System and is currently supported by the National Institute of Alcohol Abuse and Alcoholism NIAAA K23AA00313 and is a Robert Wood Johnson Foundation Generalist Faculty Scholar Drs Chew, Nelson, Young, and Bradley do not have a conflict of interest with the subject matter or materials discussed in this manuscript Corresponding Author: Address correspondence to Dr Chew, University of Washington, Harborview Medical Center, Department of Medicine, Division of General Internal Medicine, 3259th Avenue, Box 359780, Seattle, WA 98104 206-731-8541 Fax: 206-7316988 lchew@uwashingtonedu REFERENCES 1 Anda
RF, Williamson DF, Remington PL Alcohol and fatal injuries among US adults Findings from the NHANES I Epidemiologic Follow-up Study JAMA 1988;26017:2529-32 2 Kelly JP, Kaufman DW, Koff RS, Laszlo A, Wiholm BE, Shapiro S Alcohol consumption and the risk of major upper gastrointestinal bleeding Am J Gastroenterol 1995;907:1058-64 3 Kiyohara Y, Kato I, Iwamoto H, Nakayama K, Fujishima M The impact of alcohol and hypertension on stroke incidence in a general Japanese population The Hisayama Study Stroke 1995;263:368-72 4 Klatsky AL, Armstrong MA, Friedman GD Alcohol and mortality Ann Intern Med 1992;1178:646-54 5 Allen BT, DeLong ER, Feussner JR Impact of glucose self monitoring on non-insulin-treated patients with type II diabetes mellitus Randomized controlled trial comparing blood and urine testing Diabetes Care 1990;1310:1044-50 6 The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus The Diabetes Control and Complications Trial Research Group N Engl J Med 1993;32914:977-86 7 Litzelman DK, Slemenda CW, Langefeld CD, et al Reduction of lower extremity clinical abnormalities in patients with
non-insulin-dependent diabetes mellitus A randomized, controlled trial Ann Intern Med 1993;1191:36-41 8 Ferris FL, III How effective are treatments for diabetic retinopathy? JAMA 1993;26910:1290-1
Finally, data were based on self-report and may be subject to misclassification bias We relied on the diabetes diagnosis established by self-report, which has been shown to have excellent validity20 We also relied on self-reported alcohol consumption, which has been validated in other epidemiological investigations but may have introduced some error in our analyses32-35 The alcohol consumption measures used in the BRFSS survey underestimated overall alcohol consumption and therefore identified a relatively small number of heavy drinkers In addition, the alcohol consumption measures in BRFSS did not allow us to distinguish between former drinkers and those who never drink who may have comprised our nondrinker category Future studies are needed to evaluate more-sensitive measures of alcohol consumption ie, AUDIT-C32-35 and its relationship to diabetes self-care behaviors
594
September 2005
Family Medicine
23 Bradley KA, Badrinath S, Bush K, Boyd-Wickizer J, Anawalt B Medical risks for
women who drink alcohol J Gen Intern Med 1998;139:62739 24 Sanchez-Craig M, Wilkinson DA, Davila R Empirically based guidelines for moderate drinking: 1-year results from three studies with problem drinkers Am J Public Health 1995;856:823-8 25 Vinicor F, Burton B, Foster B, Eastman R Healthy people 2010: diabetes Diabetes Care 2000;236:853-5 26 Mossey JM, Shapiro E Self-rated health: a predictor of mortality among the elderly Am J Public Health 1982;728:800-8 27 Idler EL, Russell LB, Davis D Survival, functional limitations, and self-rated health in the NHANES I Epidemiologic Follow-up Study, 1992 First National Health and Nutrition Examination Survey Am J Epidemiol 2000;1529:874-83 28 Sundquist J, Johansson SE Self-reported poor health and low educational level predictors for mortality: a population-based follow-up study of 39,156 people in Sweden J Epidemiol Community Health 1997;511:35-40 29 Idler EL, Benyamini Y Self-rated health and mortality: a review of 27 community studies J Health Soc Behav 1997;381:21-37 30 Strine TW, Balluz L, Chapman DP, Moriarty DG, Owens M, Mokdad AH Risk behaviors and health care coverage among adults by frequent mental distress status, 2001 Am J
Prev Med 2004;263:213-6 31 Cherpitel CJ Drinking patterns and problems, drug use and health services utilization: a comparison of two regions in the US general population Drug Alcohol Depend 1999;533:231-7 32 Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA The AUDIT alcohol consumption questions AUDIT-C: an effective brief screening test for problem drinking Ambulatory Care Quality Improvement Project ACQUIP Alcohol Use Disorders Identification Test Arch Intern Med 1998;15816:1789-95 33 Bradley KA, McDonell MB, Bush K, Kivlahan DR, Diehr P, Fihn SD The AUDIT alcohol consumption questions: reliability, validity, and responsiveness to change in older male primary care patients Alcohol Clin Exp Res 1998;228:1842-9 34 Bradley KA, Bush KR, McDonell MB, Malone T, Fihn SD Screening for problem drinking: comparison of CAGE and AUDIT Ambulatory Care Quality Improvement Project ACQUIP Alcohol Use Disorders Identification Test J Gen Intern Med 1998;136:379-88 35 Bradley KA, Bush KR, Epler AJ, et al Two brief alcohol-screening tests from the Alcohol Use Disorders Identification Test AUDIT: validation in a female Veterans Affairs patient population Arch Intern Med 2003;1637:821-9
9 Garg
R, Wagener DK, Madans JH Alcohol consumption and risk of ischemic heart disease in women Arch Intern Med 1993;15310:1211-6 10 Sesso HD, Stampfer MJ, Rosner B, Hennekens CH, Manson JE, Gaziano JM Seven-year changes in alcohol consumption and subsequent risk of cardiovascular disease in men Arch Intern Med 2000;16017:260512 11 American Diabetes Association position statement: evidence-based nutrition principles and recommendations for the treatment and prevention of diabetes and related complications J Am Diet Assoc 2002;1021:109-18 12 Cox WM, Blount JP, Crowe PA, Singh SP Diabetic patients alcohol use and quality of life: relationships with prescribed treatment compliance among older males Alcohol Clin Exp Res 1996;202:327-31 13 Johnson KH, Bazargan M, Bing EG Alcohol consumption and compliance among inner-city minority patients with type 2 diabetes mellitus Arch Fam Med 2000;910:964-70 14 Karter AJ, Ferrara A, Darbinian JA, Ackerson LM, Selby JV Self monitoring of blood glucose: language and financial barriers in a managed care population with diabetes Diabetes Care 2000;234:477-83 15 Howard AA, Arnsten JH, Gourevitch MN Effect of alcohol consumption on diabetes mellitus: a
systematic review Ann Intern Med 2004;1403:211-9 16 Gentry EM, Kalsbeek WD, Hogelin GC, et al The behavioral risk factor surveys: II Design, methods, and estimates from combined state data Am J Prev Med 1985;16:9-14 17 Remington PL, Smith MY, Williamson DF, Anda RF, Gentry EM, Hogelin GC Design, characteristics, and usefulness of state-based behavioral risk factor surveillance: 19811987 Public Health Rep 1988;1034:366-75 18 Stein AD, Courval JM, Lederman RI, Shea S Reproducibility of responses to telephone interviews: demographic predictors of discordance in risk factor status Am J Epidemiol 1995;14111:1097-1105 19 Stein AD, Lederman RI, Shea S The Behavioral Risk Factor Surveillance System questionnaire: its reliability in a statewide sample Am J Public Health 1993;8312:1768-72 20 Midthjell K, Holmen J, Bjorndal A, Lund-Larsen G Is questionnaire information valid in the study of a chronic disease such as diabetes? The Nord-Trondelag Diabetes Study J Epidemiol Community Health 1992;465:537-42 21 The physicians guide to helping patients with alcohol problems National Institute of Alcohol Abuse and Alcoholism Washington, DC: Government Printing Office, 1995 22 Wechsler H, Davenport
A, Dowdall G, Moeykens B, Castillo S Health and behavioral consequences of binge drinking in college A national survey of students at 140 campuses JAMA 1994;27221:1672-7
