Subjective well-being, morbidity and health care consumption by hazardous, harmful and problem alcohol drinkers

Summary

Aim:
The aim of this work was to investigate subjective well-being, morbidity and healthcare needs of persons with hazardous, harmful and problematic alcohol consumption.

Methods:
Data from a questionnaire-based survey performed on a representative sample of 2,221 persons (of these, 51.4% were men) aged 18 to 39 (average age 29.9, s.d. 5.8) were used for the analysis. The level of risk related to alcohol consumption was assessed through the screening questionnaire (the Alcohol Use Disorders Identification Test [AUDIT]) and categorization into 4 groups with critical scores of 8, 16 and 20 was applied to the analysis. The questionnaire focused on the respondents' drinking habits and health and their demographic, social, and psychological background and circumstances.

Results:
The respondents' subjective assessment of their physical and mental health was varied significantly across the different score categories in AUDIT. Over one-quarter of the respondents falling in the category of harmful or problem drinkers rated their physical and/or mental health as poor or very poor. Compared to no-problem alcohol users, problem drinkers sought special help for emotional problems more frequently (3.1% vs. 21.3%; p<0.01); visited their doctors more frequently during the past year (3.1 vs. 4.8; p<0.05); had more episodes of illness (1.4 vs. 6.8; p<0.01), more days of sick leave (10.4 vs. 27.6; p<0.01); and were hospitalized more frequently (6.9% vs. 18.7%; p<0.01).

Discussion and conclusions:
The results support/confirm the link between hazardous, harmful and problematic alcohol consumption on the one hand and the drinkers' perception of their health status and use of medical/healthcare services on the other hand. A modification of inappropriate consumption patterns through a short intervention by a general practitioner can lead to health improvement and reduction of the drinkers' healthcare costs, which are borne by the whole of society.

Key words:
alcohol consumption, harmful drinking, AUDIT, morbidity, young adults

Introduction

Based on recent reviews, studies of the relationship between health and alcohol consumption has focused on mortality depending on alcohol dose. Rehm (1) cites 80 papers of this type in his meta-analysis. A total of 60 diseases/health conditions are reported as being affected by alcohol to a larger or lesser extent (2,3,4). Chronic diseases in which alcohol is involved include, in particular, tumorous diseases (especially cancer of the oral cavity, oesophagus and liver), cardiovascular diseases (hypertension, coronary heart diseases, cerebrovascular diseases), neuropsychiatric disorders (epilepsy, depressions, dependence on alcohol) and gastrointestinal tract diseases (especially alcoholic cirrhosis of the liver). Relative mortality risks depending on gender and alcohol dosage have been identified for the majority of these diseases (4).

Among the acute effects of alcohol, most attention is apparently paid to injuries and health impacts of traffic accidents (5).

Studies devoted to the impact of alcohol drinking on morbidity and on the quality of life are less numerous. The reasons for this include mainly methodological issues and the costs of such studies (6).

Murray and Lopez (6) estimate the overall contribution of alcohol to annual mortality to be 1.5 per cent. In view of the significant impact of alcohol on health and mortality, questions arise as to what strategies could reasonably limit such social and economic losses. Papers by Babor (7) and by Anderson and Baumberg (8) summarise evidence for their conclusions that mere raising of public awareness has little effect and medical treatment of the consequences of alcohol drinking is very costly and not very effective from the public health aspect. The authors suggest that brief intervention in the primary health care setting is an economically feasible and, at the same time, reasonably efficient alternative. The efficiency and effectiveness of brief interventions have been documented by several papers, including the recent review by Bertholet et al. (9). According to the latter, eight studies confirmed the effect of brief interventions, resulting in weekly alcohol consumption reduction by 38 grams in average. Screening and brief interventions are not unknown in the Czech Republic either. Sovinova and Csémy (10) adapted one of the most widespread screening instruments to the Czech setting and tested its properties. As far as brief interventions are concerned, we translated and published the World Health Organization guidelines written by Babor and Higgins (11).

The aim of the present study was to establish the level of risk related to alcohol consumption, measured in a sample of young Czech adults aged 20 to 39 by using The Alcohol Use Disorders Identification Test (AUDIT), and to analyse the relationships between the risk level and the subjects' health during the past year. This is the first study of this type conceived and performed in the Czech Republic. We even found no paper in international literature devoted to the relation between AUDIT alcohol drinking categories and the individuals' health.

Methods

Study group. The study group included 2,221 subjects aged 18 to 39 (mean age 29.9, STD 5.8). The sample was representative of the Czech population with regards to gender, age, level of education and geographic distribution. The parameters of the study group are summarized in Table 1. The subjects were recruited by a two-step selection process. The electoral districts were chosen in the first step. The Complex Samples module was used to perform the stepwise selection, applying a step calculated on the number of registered voters in each electoral district (the number of registered voters is the measure of size in the PPS systematic selection method), whereby the probability of inclusion was identical for each voter. The stepwise selection was performed for each administrative region independently (the administrative region was the stratification variable), whereby appropriate representation of all regions was assured. The required number of electoral districts selected in each administrative region was calculated proportionally with respect to the number of registered voters in the region. In step two, the random walk approach was used; interviewers sought respondents based on a quota system within each electoral district selected. A total of 234 trained field interviewers were involved. Data was collected in October and November 2009. From among the 2663 persons who were invited to participate, 442 (16.3%) refused, mostly stating lack of time as the reason.

Research tool. Data was collected using a questionnaire specifically developed for this survey. The questionnaire consisted of 61 questions, some of which were broken down into sub-questions. Each respondent provided 206 pieces of data. The questionnaire as a whole was divided into four general areas. The first part asked questions regarding the respondent's relationship to alcohol; the remaining three parts were concerned with the respondent's health, psycho-social adaptation, and demographic data including information about the respondent's family and employment.

Data collection method. Information was gathered during a controlled structured interview, mostly conducted in the respondents' homes. From among the 2228 interviews accomplished, seven were finally eliminated because of a relatively large number of unanswered questions.

Electronic data file creation. Information in the questionnaires was rewritten into a program Statistická analýza sociálních dat. The cleaned data set was converted to an Statistical Package for Social Sciences (SPSS) file.

Variables examined. The risk level in relation to alcohol consumption was assessed using the AUDIT screening questionnaire. Mean yearly alcohol consumption (in litres of pure ethanol) was calculated based on the typical frequency and amount for each type of alcoholic beverage. The following health parameters were considered: subjective assessment of physical and mental well-being; seeking of professional help due to physical and mental health problems during the past 12 months; number of medical examinations, episodes of illness, sick days and days of inpatient care, all during the past 12 months.

Statistical analysis. Data was processed by using SPSS ver. 16 software. The chi-square test was used to examine the difference in the frequency distribution, and variance analysis was applied to test the differences between the group averages.

Results

The recommended critical levels in the AUDIT screening questionnaire were used to categorise the risk level as follows: score lower than 8: low risk; score between 8 and 15: hazardous drinking; score between 16 and 19: harmful drinking: score 20 or more: problem drinking or addiction. The distribution of the study sample into those categories is shown in Graph 1; 71% respondents fall in the low risk drinking category, 20.7% fall in the hazardous drinking category, 4.4% fall in the harmful drinking category, and 3.6% fall in the high risk (problem) drinking category. The total average yearly alcohol consumption matches the risk levels. The average consumption is 4.8 litres of 100% alcohol in the low risk drinkers group and increases rapidly and linearly up to 38.8 litres in the problem drinkers group.

Subjective physical and mental well-being depending on the risk level differed statistically significantly (p<0.001) for both genders. It was particularly problem drinkers who reported impaired somatic and/or mental well-being with increased frequency. Graph 2 shows clearly that over one-fifth of the problem drinkers assessed their mental and physical well-being as poor or very poor.

Information regarding how the subjects were actively seeking medical or other professional help is summarised in Table 1. While the categories did not differ in the extent to which the respondents sought medical help for physical problems, the need for professional help for mental problems was considerably different; it was particularly in the problem drinking group that the respondents sought professional help to a several times larger extent than in the lower-risk groups.

Table 2 summarises data regarding the average number of visits to a doctor, average number of diseases and days of illness during the past 12 months, and inpatient hospital care. Statistically significant differences among the different risk groups were identified for all of these variables. The average number of visits to a doctor was three for the whole sample and 4.8 in the problem drinkers group. The latter group also exhibited a larger number of cases of illness during the past 12 months. Problem drinkers were ill 6.8 times on average, as against the 1.3 times in the whole sample. As regards the days-of-illness variable, elevated levels were found even in the harmful drinking group (13.8 days) and were considerably higher in the problem drinkers group (27.6 days). (The critical level F and the statistical significance of the difference are included in Table 2.) The number of hospitalisations was roughly triple in the harmful and problem drinking groups as compared to the lower-risk groups (15.4% and 18.7% respectively, against 6.9% and 5.9%; P<0.001).

Discussion and conclusions

While the vast majority of publications in this area are concerned with specific health impacts (such as mortality for selected diagnoses) particularly depending on the alcohol dosage (4), it was the aim of this study to find whether the risk level measured by a screening test whose score includes, in addition to drinking habits, symptoms of problem drinking and consequences of drinking, is reflected in the general well-being and health of young adults. The results of the study show that harmful and problem drinking does indeed have measurable impacts on the well-being (health) of young adults. Individuals who drink alcohol excessively seek medical help more frequently, are ill more frequently, have a larger number of sick days and are more frequently admitted to hospital as inpatients.

Health is generally also affected by other lifestyle factors, such as lack of physical activity, unsuitable diet, being overweight and, in particular, smoking. Regarding habitual smoking, interaction with alcohol consumption should be considered because smoking correlates with drinking. There were 31% daily smokers in the study sample as a whole and 22% in the low risk group. This suggests that to some extent, smoking will affect the health of the subjects in all drinking-risk groups. Therefore, we also performed analyses controlling for the smoking factor. The effect of alcohol consumption on health remained significant even when the effect of smoking was eliminated.

In our opinion, this study has two important practical implications. Firstly, the results show that the drinking risk levels measured by a simple screening questionnaire relate to health, hence, that the screening instrument can be applied for identification of the risk level. Secondly, brief advice or brief intervention aimed at reducing hazardous or harmful drinking can be implemented in the primary care setting and such advice or interventions are meaningful even in the hazardous drinking category where the health impacts are still minimal and no severe impairment is observed (12, 13, 14). Identification of problem drinking by means of the AUDIT questionnaire can prompt the practitioner to refer the patient to specialised medical care.

Early identification of alcohol problems and brief intervention can result in improved health and cost savings in the specialised treatment of diseases that are contributed to by excessive drinking.

Acknowledgement

This work was supported by Grant # NS 9645-4/2008/ from IGA MZ CR.

Affiliation of the senior author:
Hana Sovinová, MD
Státní zdravotní ústav
Šrobárova 48
100 42 Praha 10
sovinova@szu.cz
Phone: 267082328