Drinking in denial: a cross-sectional analysis of national survey data in Ireland to measure drinkers' awareness of their alcohol use.

OBJECTIVES: Ireland has high per capita alcohol consumption and also has high levels of problematic drinking patterns. While it is accepted that patterns of alcohol consumption in Ireland are a cause for concern, it is not clear if Irish people are actually aware of the extent of their hazardous or harmful pattern of drinking. The aim of this study was to determine awareness of drinking pattern in an Irish population using a representative random sample and to identify characteristics associated with self-awareness of hazardous or harmful drinking.
METHODS: We analysed data from Ireland's 2014/2015 Drug Prevalence Survey which recruited a stratified clustered sample of 7005 individuals aged 15 years and over living in private households. Logistic regression analysis was used to determine characteristics associated with self-awareness of hazardous or harmful drinking.
RESULTS: Almost one half of drinkers had a hazardous or harmful pattern of drinking; 38% engaged in monthly risky single occasion drinking (RSOD) and 10.5% met Diagnostic and Statistical Manual version IV (DSM-IV) criteria for alcohol dependence. Of the 2420 respondents who had a hazardous or harmful pattern of drinking, 67% were unaware of this and misclassified themselves as being either a light or moderate drinker who did not engage in RSOD. An adjusted logistic regression model identified that hazardous and harmful drinkers were more likely to be aware of their drinking pattern if they had completed third level education (OR=1.80, 95% CI: 1.30 to 2.49) while older drinkers (aged 65 and over) were less likely to be aware of their drinking pattern (OR=0.30, 95% CI: 0.14 to 0.65). Subjects who engaged in risk taking behaviours such as illicit drug use and gambling were also significantly more likely to be aware of their drinking pattern.
CONCLUSIONS: The results of this study suggest that patterns of alcohol use in Ireland are problematic. Older respondents and those with lower educational attainment are less likely to be aware of their hazardous or harmful drinking pattern. There is also a population of younger, more-educated drinkers who engage in potentially risk-taking behaviours and these subjects are aware of their harmful drinking pattern. Initiatives to reduce overall alcohol consumption and raise awareness around drinking patterns are required. © Author(s) (or their employer(s)) 2020. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

This was a large national survey which was representative of the Irish population.

The study employed standardised methods for the measurement of hazardous and harmful alcohol consumption.

Although the overall response rate was good, alcohol surveys often fail to recruit the heaviest drinkers, resulting in selection bias.

There were discrepancies between the definitions used to define drinking patterns and the categories that respondents were asked to select from to self-assess their own drinking.

Introduction

Alcohol is the seventh leading risk factor for deaths and is responsible for 10% of global deaths annually among those aged 15–49 years.1 A person’s pattern of drinking is an important determinant of alcohol-related harm. While there has traditionally been a focus on overall volume of drinking, greater attention is now being paid to the impact of drinking pattern on harms over and above the effects from total alcohol consumption. Risky single occasion drinking (RSOD), also referred to as binge drinking or heavy episodic drinking, is associated with a number of negative health, social and economic consequences. Health harms include liver cirrhosis, coronary heart disease and various types of cancer.2–4 RSOD may also impair judgement, increasing the likelihood of driving under the influence of alcohol, intentional self-harm, injury and risky sexual behaviours. It has been described by the WHO as a hazardous pattern of drinking.4 Alcohol dependence may be described as a harmful pattern of drinking; it is a chronic condition and is defined as ‘a cluster of physiological, behavioural, and cognitive phenomena in which the use of alcohol takes on a much higher priority for a given individual than other behaviours that once had greater value’.5

Alcohol use in Ireland is characterised by high per capita consumption and a high level of problematic drinking patterns. While surveys consistently report that 20%–25% of Irish adults abstain from alcohol,6 7 the most recent available figures indicate that Ireland is the sixth heaviest drinking nation among Organisation for Economic Co-operation and Development (OECD) countries in terms of the overall volume of alcohol consumed.8 The WHO reported in 2018 that 41% of all Irish people aged 15 years and over had engaged in heavy episodic drinking or RSOD in the past 30 days, placing Ireland in eighth place among the 194 countries analysed.4 Three-quarters of all alcohol consumed in Ireland is done so as part of an RSOD session.6

While it is accepted that patterns of alcohol consumption in Ireland are a cause for concern, it is not clear if Irish people are actually aware of the extent of their hazardous or harmful pattern of drinking. If it is the case that people are not aware of their drinking pattern this suggests that interventions to increase awareness in Ireland may be required. In Australia, awareness of drinking is low with most people, regardless of their drinking pattern, considering themselves to be an occasional, light or social drinker. Risky drinkers were less likely than low-risk drinkers to be aware of what constituted risky drinking.9 Irish research indicates that those most likely to experience alcohol-related harm are those who are alcohol dependent followed by those who engage in regular RSOD.10 Given the relationship between drinking pattern and alcohol-related harm in Ireland, the aim of this study was to determine awareness of drinking patterns in an Irish population using a representative random sample and to identify characteristics associated with self-awareness of hazardous or harmful drinking.

Methods

Sampling and study population

We analysed data from Ireland’s 2014/2015 Drug Prevalence Survey. This national survey recruited a stratified clustered sample of 7005 individuals aged 15 years and over, living in private households in Ireland. The sampling frame used was the GeoDirectory, which is a list of all addresses in the Ireland, and distinguishes between residential and commercial establishments. A three-stage process was used to construct the sample for this survey. The first stage involved stratifying the population into 10 former health board regions in Ireland. In the second stage of stratification, 421 electoral divisions were selected as the primary sampling units across the 10 former health board regions. Before selection, the primary sampling units were ranked by the following sociodemographic indicators: population density, male unemployment and social class, to ensure that a representative cross-section of areas were included. Finally, in each primary sampling unit, 31 addresses were chosen randomly, and at each address, one person was selected to participate in the survey, using the ‘last birthday’ rule, whereby, the person whose birthday occurred most recently was selected. The achieved sample was weighted by gender, age and former health board region to maximise its representativeness of the general population. A more comprehensive description of the survey’s methodology has been detailed elsewhere.11 The survey involved a face-to-face interview in the participants’ home and a self-completion questionnaire. Respondents also self-completed questions in relation to alcohol dependence and their perception of their own drinking pattern. The home interviews were conducted by trained interviewers using Computer Assisted Personal Interviewing. Interviews were completed between August 2014 and August 2015, and achieved a 61% response rate. No data on non-respondents were collected.

Definitions of drinking patterns

Current drinkers were defined as those who had consumed alcohol at least once in the last 12 months. Non-drinkers, categorised as those who had not consumed alcohol in the past year (n=1608), were excluded from this study.

Hazardous drinking—regular RSOD in the past year

There are no internationally agreed definitions on how much alcohol constitutes an RSOD episode or on what is regular RSOD. RSOD is referred to as ‘heavy episodic drinking’ by the WHO, and is also commonly known as ‘binge drinking’. We defined RSOD as consuming 60 g of pure alcohol on a single drinking occasion similar to the WHO definition.12 Respondents were asked how often they had consumed the equivalent of six standard drinks on a single drinking occasion in the past year. In Ireland, a standard drink contains 10 g of pure alcohol. Frequency of RSOD was measured as follows: daily, 5–6 times a week, 4 times a week, 3 times a week, 2 times a week, once a week, 2–3 times a month, once a month, 6–11 times a year, 2–5 times a year and once a year. The concept of a standard drink and what constitutes 60 g of alcohol was explained in detail to each respondent and visual aids were provided depicting 60 g of alcohol according to beverage type. We defined hazardous drinkers as those who engaged in RSOD at least monthly in the previous 12 months, but who did not meet the criteria for alcohol dependence (box 1).

Low-risk—drinkers who did not meet the criteria for alcohol dependence and who had not engaged in monthly risky single occasion drinking (RSOD) in the past year.

Hazardous—drinkers who had engaged in RSOD at least monthly, but did not meet the criteria for alcohol dependence in the past year.

Harmful—drinkers who met the DSM-IV criteria for dependence in the past year.

Harmful drinking—alcohol dependence

Alcohol dependence was defined according to Diagnostic and Statistical Manual version IV (DSM-IV) criteria, and was measured via self-completed questionnaire using the 10 items that denote alcohol dependence from the Composite International Diagnostic Interview, an instrument that is used in many general population studies.13 Alcohol dependence was established from a positive response in three or more of the seven domains on the DSM-IV diagnostic criteria in the 12 months before the interview.14 Harmful drinkers were defined as those who met the criteria for alcohol dependence, regardless of their RSOD status. Drinkers who met the criteria for both regular RSOD and alcohol dependence were assigned to the alcohol dependence/harmful drinking type. Respondents who did not have complete data on RSOD and a DSM-IV score (n=236) were excluded from the analysis.

Low-risk drinking

For this study, low-risk drinking was defined as drinking that did not fit our criteria of hazardous or harmful drinking that is, those drinkers who were not alcohol dependent and who also did not engage in regular RSOD.

Self-perception of own drinking

Drinkers were asked to describe their own drinking by selecting one of the following six statements: ‘I am a heavy drinker’; ‘I am a heavy drinker and sometimes I binge drink’; ‘I am a moderate drinker’; ‘I am a moderate drinker and sometimes I binge drink’; ‘I am a light drinker’ or ‘I am a light drinker and sometimes I binge drink’. This question was cognitively tested prior to the survey and the wording used reflects the feedback received from the participants following the cognitive testing exercise on their understanding of the terms used. This question was answered by respondents via self-completed questionnaire. No descriptions of the terms ‘light’, ‘moderate’, ‘heavy’ or ‘binge’ were provided to respondents. The terms ‘light’ and ‘moderate’ were used in this question instead of ‘low-risk,’ and ‘binge’ was used instead of RSOD as they are terms typically used by the general public in Ireland. This was also reflected in the cognitive testing of the questionnaire which found that respondents were more familiar with the terms ‘light’, ‘moderate’ and ‘binge’. For ease of analysis and to allow us to make comparisons with the three drinking patterns we measured, that is, low-risk, hazardous and harmful drinking, we collapsed the six statements into three groups. We combined the categories ‘I am a light drinker’ and ‘I am a moderate drinker’; the categories ‘I am a light drinker and sometimes I binge drink’ and ‘I am a moderate drinker and sometimes I binge drink’; and the categories ‘I am a heavy drinker’ and ‘I am a heavy drinker and sometimes I binge drink’. We then compared respondents’ self-perception of their own drinking against their drinking patterns as measured elsewhere in the questionnaire through the RSOD and DSM-IV questions (box 2).

Light/moderate—those who selected one of these statements: ‘I am a light drinker’ or ‘I am a moderate drinker’.

Light/moderate and sometimes binge drink—those who selected one of these statements: ‘I am a light drinker and sometimes I binge drink’ or ‘I am a moderate drinker and sometimes I binge drink’.

Heavy—those who selected one of these statements: ‘I am a heavy drinker’ or ‘I am a heavy drinker and sometimes I binge drink’.

Awareness of hazardous and harmful drinking

For this part of the analysis, we only included those respondents who were classified as hazardous/RSOD or harmful/dependant drinkers (n=2420). Respondents were considered to be unaware of their own hazardous and harmful drinking if they incorrectly underestimated their drinking pattern that is, those regular RSOD drinkers who classified themselves as light or moderate drinkers who do not binge drink and dependent drinkers who classified themselves as light or moderate drinkers who may or may not sometimes binge drink. Respondents were considered to be aware of their own hazardous or harmful drinking pattern if they described themselves as sometimes binge drinking or as a heavy drinker.

Statistical analysis

The distribution of drinking pattern was analysed by sociodemographic and addictive behaviour variables that are associated with alcohol. The sociodemographic variables analysed were age, sex, marital status, education, employment, region, dependent children; and the addictive behaviour variables analysed were smoking status (defined as being a current smoker), last year gambling (excluding lottery) and last year illicit drug use. This was analysed by cross-tabulation and statistical significance was assessed by the Pearson χ2 test. Cross-tabulation was used to compare the drinking pattern of respondents as measured using the RSOD and DSM-IV questions with their self-perceived drinking pattern.

Univariate logistic regression analyses were performed to determine factors associated with self-awareness of drinking pattern. Those variables which were identified as being significant or borderline significant (p<0.1) were then entered into a multivariable logistic regression model which was used to estimate adjusted ORs of being self-aware of hazardous or harmful drinking. This model was adjusted for gender, age, marital status, education, employment, illicit drug use and gambling. The ability of variables identified in multivariable analysis to separate cases from non-cases was evaluated using the c statistic. For all analyses, a p-value of less than 0.05 was considered to indicate statistical significance. Data were analysed using Stata V.15.1. Results are displayed using weighted data.

Results

Drinking patterns of respondents

Of the 7005 survey respondents, 5397 (77.0%, 95% CI: 75.7 to 78.3) had consumed alcohol in the last year; among current, or last year drinkers, a drinking pattern could be assigned to 5144. Just over half (51.6%, 95% CI: 49.9 to 53.2) of drinkers were low-risk drinkers, 38% (95% CI: 36.4 to 39.6) engaged in regular RSOD, and 10.5% (95% CI: 9.4 to 11.6) were dependent drinkers. table 1 presents the characteristics of drinkers by drinking pattern. Men accounted for 51.7% of drinkers, 56.3% of drinkers were employed and 48.8% had completed third level education. Low-risk drinkers were predominantly female, aged over 35 years and married. The characteristics of regular RSOD and dependent drinkers were similar; they were more likely to be male, young and single. Low-risk drinkers were most likely to have dependent children (42.6%). The likelihood of engaging in other addictive behaviours increased as hazardous/harmful drinking pattern increased. Smoking was observed in 18.5% of low-risk drinkers, compared with 31.1% of RSOD drinkers and 49.6% of dependent drinkers. Illicit drug use was observed in 2.8% of low-risk drinkers, 11.6% of RSOD drinkers and 33.9% of dependent drinkers, while the respective figures for gambling were 26.4%, 41.4% and 56.6%. The three drinking pattern categories differed with statistical significance for all variables with the exception of education.

Table 1

Sociodemographic and addictive behaviour characteristics of drinkers by drinking pattern

Weighted count=5144	All drinkers	Low-risk drinkersN (%)	Regular RSOD drinkerN (%)	Dependent drinkersN (%)	P value
	5144 (100)	2652 (51.6)	1953 (38.0)	539 (10.5)
Gender
Male	2659 (51.7)	993 (37.5)	1327 (68.0)	339 (62.8)	<0.001
Female	2485 (48.3)	1659 (62.6)	626 (32.1)	200 (37.2)
Age group
15–24	831 (16.2)	290 (11.0)	344 (17.7)	197 (36.5)	<0.001
25–34	1140 (22.2)	439 (16.6)	515 (26.4)	186 (34.5)
35–64	2576 (50.2)	1503 (56.9)	921 (47.3)	152 (28.1)
65+	582 (11.4)	411 (15.5)	167 (8.6)	5 (0.9)
Marital status
Single/never married	1652 (32.2)	583 (22.0)	739 (37.9)	330 (61.8)	<0.001
Married/cohabiting	3097 (60.3)	1830 (69.1)	1097 (56.2)	170 (31.8)
Divorced/separated/ widowed	386 (7.5)	235 (8.9)	116 (5.9)	35 (6.5)
Education
Primary/none	1099 (21.4)	537 (20.3)	453 (23.3)	109 (20.3)	0.0562
Completed secondary	1531 (29.8)	760 (28.7)	608 (31.2)	163 (30.4)
Completed third level	2502 (48.8)	1351 (51.0)	886 (45.5)	265 (49.4)
Employment
Employed	2896 (56.3)	1410 (53.2)	1205 (61.7)	280 (52.0)	<0.001
Unemployed	500 (9.7)	203 (7.7)	219 (11.2)	78 (14.4)
Student	534 (10.4)	204 (7.7)	203 (10.4)	127 (23.5)
Home duties	533 (10.4)	407 (15.4)	113 (5.8)	13 (2.4)
Retired	521 (10.1)	353 (13.3)	163 (8.3)	6 (1.1)
Other	161 (3.1)	75 (2.8)	51 (2.6)	35 (6.6)
Region
Dublin	1503 (29.2)	724 (27.3)	569 (29.1)	210 (38.9)	<0.001
Outside Dublin	3642 (70.8)	1928 (72.7)	1384 (70.9)	329 (61.1)
Dependent children
Yes	1977 (38.6)	1124 (42.6)	725 (37.2)	128 (24.0)	<0.001
Smoking
Yes	1365 (26.5)	490 (18.5)	607 (31.1)	268 (49.6)	<0.001
Illicit drug use
Yes	483 (9.4)	75 (2.8)	226 (11.6)	182 (33.9)	<0.001
Gambling
Yes	1813 (35.3)	699 (26.4)	809 (41.4)	305 (56.6)	<0.001

Numbers may not add up to the column totals because of missing data.

RSOD, risky single occasion drinking.

Self-perception of own drinking and comparison with own drinking pattern

Information on drinking pattern and self-defined drinking category was available for 5053 respondents. The majority of drinkers (70.9%) classified themselves as light or moderate drinkers who do not binge drink, 26.7% categorised themselves as light or moderate drinkers who sometimes binge drink, and 2.4% classified themselves as heavy drinkers (table 2). Most low-risk drinkers (83.8%) described themselves as light or moderate drinkers. Almost two-thirds of regular RSOD drinkers and one-third (33.8%) of dependent drinkers described themselves as light or moderate drinkers. Just 35.1% of regular RSOD drinkers stated that they sometimes engaged in binge drinking and just 16% of dependent drinkers described themselves as a heavy drinker. A similar trend was observed among males and females. However, dependent female drinkers were less likely than males to describe themselves as a heavy drinker (11.4% vs 18.7%). There were 426 (16.2%) low-risk drinkers and 29 (1.5%) regular RSOD drinkers who over-estimated their drinking pattern.

Table 2

Self-Perceived drinking category by drinking pattern

Weighted count=5053	All drinkers(n=5053)	Low-risk drinkers(n=2634)	Regular RSOD drinkers(n=1890)	Dependent drinkers(n=529)
All drinkers
Light/moderate	3584 (70.9)	2208 (83.8)	1198 (63.4)	179 (33.8)
Light/moderate and sometimes binge	1348 (26.7)	419 (15.9)	663 (35.1)	266 (50.3)
Heavy drinker	121 (2.4)	7 (0.3)	29 (1.5)	85 (16.0)
Male drinkers
Light/moderate	1726 (66.4)	812 (82.2)	802 (62.6)	111 (33.5)
Light/moderate and sometimes binge	783 (30.1)	172 (17.5)	453 (35.4)	159 (47.7)
Heavy drinker	91 (3.5)	3 (0.3)	26 (2.0)	62 (18.7)
Female drinkers
Light/moderate	1859 (75.8)	1395 (84.8)	395 (64.9)	68 (34.2)
Light/moderate and sometimes binge	565 (23.0)	246 (15.0)	210 (34.6)	108 (54.4)
Heavy drinker	30 (1.2)	4 (0.2)	3 (0.5)	23 (11.4)

RSOD, risky single occasion drinking.

Awareness of own drinking among hazardous/harmful drinkers

Of those who had a hazardous or harmful pattern of drinking (n=2420), 67.9% were unaware of this and misclassified themselves as being either a light or moderate drinker. Self-awareness of hazardous or harmful drinking pattern by socio-economic demographics and other addictive behaviours is presented in table 3. In unadjusted analyses, respondents who were younger, who had completed secondary or third level education, and those who had engaged in illicit drug use and gambling in the previous year were significantly more likely to be aware that their drinking pattern was hazardous or harmful. Survey participants who were older, married and who were engaged in home duties or retired were significantly less likely to be aware that their drinking pattern was hazardous or harmful.

Table 3

Unadjusted ORs for factors associated with self-awareness of hazardous or harmful drinking

Weighted count=2420	N	Aware of hazardous/ harmful drinking (n=777)	Unaware of hazardous/ harmful drinking (n=1643)	OR	95% CI	P value
Gender
Female	808	236 (30.4)	572 (34.8)	1	Ref
Male	1612	541 (69.6)	1071 (65.2)	1.22	0.97 to 1.53	0.084
Age group
15–24	529	180 (23.1)	349 (21.3)	1	Ref
25–34	683	282 (36.2)	401 (24.5)	1.36	0.98 to 1.89	0.065
35–64	1041	294 (37.9)	747 (45.6)	0.77	0.57 to 1.02	0.069
65+	162	22 (2.8)	140 (8.6)	0.30	0.19 to 0.48	<0.001
Marital status
Single/never married	1036	364 (47.0)	672 (41.0)	1	Ref
Married/cohabiting	1233	367 (47.4)	866 (52.8)	0.78	0.62 to 0.98	0.037
Divorced/separated/ widowed	145	44 (5.6)	101 (6.2)	0.79	0.54 to 1.16	0.230
Education
Primary/none	533	119 (15.3)	414 (25.3)	1	Ref
Completed secondary	752	231 (29.7)	521 (31.9)	1.54	1.11 to 2.14	0.009
Completed third level	1128	427 (55.0)	701 (42.9)	2.12	1.58 to 2.85	<0.001
Employment
Employed	1448	503 (64.8)	945 (57.5)	1	Ref
Unemployed	352	120 (15.4)	232 (14.1)	0.97	0.72 to 1.31	0.842
Student	324	97 (12.5)	228 (13.9)	0.80	0.55 to 1.16	0.233
Home duties	120	25 (3.2)	95 (5.8)	0.49	0.30 to 0.80	0.004
Retired	163	29 (3.7)	134 (8.2)	0.40	0.27 to 0.59	<0.001
Other	13	4 (0.5)	9 (0.6)	0.73	0.22 to 2.42	0.602
Dependent children
No	1582	518 (66.7)	1065 (65.2)	1	Ref
Yes	827	258 (33.3)	569 (34.8)	0.93	0.74 to 1.17	0.545
Region
Outside Dublin	1652	528 (68.0)	1123 (68.4)	1	Ref
Dublin	768	249 (32.0)	519 (31.6)	1.02	0.79 to 1.31	0.888
Illicit drug use
No	2029	608 (78.3)	1420 (86.5)	1	Ref
Yes	391	169 (21.7)	222 (13.5)	1.78	1.31 to 2.40	<0.001
Smoking
No	1584	493 (63.4)	1091 (66.5)	1	Ref
Yes	836	284 (36.6)	551 (33.6)	1.14	0.91 to 1.44	0.254
Gambling
No	1333	357 (45.9)	976 (59.4)	1	Ref
Yes	1087	420 (54.1)	667 (40.6)	1.72	1.38 to 2.15	<0.001

Numbers may not add up to the column totals because of missing data.

Table results shown in bold are significant (p<0.05).

An adjusted logistic regression model identified that respondents aged 65 years and over were 0.3 times (95% CI: 0.14 to 0.65) as likely to be aware of their hazardous or harmful drinking pattern compared with those aged 15–24 years (table 4). Higher education was also associated with self-awareness of hazardous or harmful drinking in multivariable analysis, with those who had completed third-level education being 1.8 times (95% CI: 1.30 to 4.60) more likely to be aware compared with those who had completed primary education only. Participants who were aware of their hazardous or harmful drinking pattern were also more likely to engage in illicit drug use (OR=1.45, 95% CI: 1.04 to 2.01) or to gamble (OR=1.60, 95% CI: 1.27 to 2.01). The c statistic for a model which included these variables was 0.65 (95% CI: 0.63 to 0.68).

Table 4

Adjusted ORs for factors associated with self-awareness of hazardous or harmful drinking

Variables	OR	95% CI	P value	Wald score
Gender
Female	1	Ref		1.76
Male	1.19	0.92 to 1.54	0.185
Age
15–24	1	Ref		17.26
25–34	1.07	0.71 to 1.62	0.748
35–64	0.66	0.43 to 1.01	0.055
65+	0.30	0.14 to 0.65	0.002
Marital status
Single/never married	1	Ref		3.22
Married	0.98	0.73 to 1.32	0.912
Divorced/separated/ widowed	1.40	0.90 to 2.18	0.135
Education
Primary/none	1	Ref		13.22
Completed secondary	1.36	0.96 to 1.93	0.079
Third level	1.80	1.30 to 2.49	<0.001
Employment
Employed	1	Ref		3.95
Unemployed	1.05	0.75 to 1.49	0.770
Student	0.70	0.44 to 1.13	0.142
Home duties	0.77	0.45 to 1.33	0.354
Retired	1.06	0.57 to 1.95	0.857
Other	0.52	0.13 to 2.16	0.371
Illicit drug use
No	1	Ref		4.96
Yes	1.45	1.04 to 2.01	0.026
Gambling
No	1	Ref		15.75
Yes	1.60	1.27 to 2.01	<0.001

Table results shown in bold are significant (p<0.05).

*ORs are adjusted for all other variables in the table.

Patient and public involvement

No patient involved.

Discussion

Main findings of the study

The results of this nationally representative study of 7005 respondents suggest that patterns of alcohol use in Ireland are problematic. Almost half of all drinkers either engage in frequent RSOD (38%) or score positive for alcohol dependence (10.5%). In addition to hazardous and harmful drinking patterns being commonplace in Ireland, this study finds that a majority of those who engage in such patterns of drinking are unaware of this. Low-risk drinkers were mostly aware of their own pattern of drinking, although 16.2% overestimated their drinking pattern. In comparison, awareness of drinking pattern was low for regular RSOD drinkers and for dependent drinkers. One-third (33.8%) of drinkers with a positive DSM-IV score self-categorised themselves as being either a light or moderate drinker and a further 50.3% described themselves as a light or moderate drinker who sometimes binge drinks. Given that alcohol dependence is a maladaptive pattern of alcohol consumption, manifested by symptoms leading to clinically significant impairment,15 it is particularly concerning that so many Irish people with alcohol dependence believe themselves to be light or moderate drinkers. Our adjusted regression analysis found that the factors independently associated with self-awareness of hazardous or harmful drinking pattern were having a higher educational level and engaging in risk taking behaviours, such as illicit drug use and gambling, while those aged 65 and over were significantly less likely to be aware of their hazardous or harmful drinking pattern. Nevertheless, the c statistic demonstrated that the ability of our model to separate cases from non-cases was poor. This indicates that there are likely to be other factors which we have not identified that are associated with awareness of drinking pattern in this population. It is possible that factors not included in this survey such as personality traits may be associated with awareness of drinking pattern and further research is required to identify these factors.

Strengths and limitations

To the best of our knowledge, this is the first study to attempt to identify factors associated with the public’s self-perception of their own drinking using a general population survey. A further strength is that the survey had a large sample size of 7005, and respondents were selected using a random probability sample that was representative of the Irish population; thus our findings are generalisable to the whole population. We also used valid and reliable measures of hazardous and harmful alcohol consumption, namely the frequency of RSOD and the DSM-IV questionnaire.

However, this study has a number of limitations which need to be considered when interpreting the findings. While our results are nationally representative, response bias may also be considered a limitation; general population surveys such as this often fail to recruit the heaviest drinkers, as they may be difficult to contact and if contacted may be less likely to agree to participate.16 Only a limited number of alcohol questions were included in this survey and they used a 12-month reference period, which may lead to reduced recall for respondents. This survey included the AUDIT-C, but not the full AUDIT. As so many drinkers (73% of men and 41% of women) met the criteria for hazardous drinking using the AUDIT-C, we felt that using measures of RSOD and dependence to denote hazardous and harmful drinking was more appropriate. Self-reporting biases are common to alcohol use surveys and lead to underestimation of alcohol consumption. The usual range of coverage from surveys is in the region of 40%–60%.17 18 In a 2013 Irish population survey, self-reported alcohol consumption based on ‘typical drink questions’ accounted for just 39% of per capita sales, even though the concept of a standard drink was explained in detail to each respondent and visual aids were provided.6 Finally, there were discrepancies between the definitions used to define drinking patterns and the categories that respondents were asked to select from to self-assess their own drinking. However, it was felt that the alcohol terms typically used in clinical and research settings would not be as easily understood by the general public, and this was corroborated by the cognitive testing of the questionnaire that was undertaken prior to the survey.

Comparison with previous work

In relation to drinking patterns, the results of this study suggest that Ireland has a high level of RSOD when compared with other countries.4 The prevalence of alcohol dependence in this study was also high when compared with a study of alcohol dependence in European countries, although it should be noted that a number of different instruments were used to measure dependence in the European report.19

The available evidence suggests that knowledge on standard drinks and drinking guidelines both in Ireland and internationally is limited, which may help explain why so few respondents correctly identified their pattern of drinking. Our findings regarding awareness of hazardous and harmful drinking are similar to a recent study in Australia, which reported that 68% of Australian drinkers who consume 11 or more standard drinks on a ‘typical occasion’ consider themselves a ‘responsible drinker’.20 A Swedish study reported low levels of knowledge of standard drink and hazardous drinking concepts among hazardous drinkers.21 A review of the literature on standard drinks for the European Joint Action on Alcohol found little understanding of what the term ‘standard drink’ actually means and that drinkers are not able to define standard drinks accurately.22 A 2012 Irish survey demonstrated that while 58% had heard of the term ‘standard drink’, just 39% knew how many standard drinks are in a pint of lager and 33% knew how many standard drinks are in a single measure of spirits, which are the typical serving sizes of lager and spirits in Ireland.23 In the UK, knowledge of the previous drinking guidelines was poor, in spite of them having been in place for 20 years. In 2012, only about one-quarter of people were able to provide a correct estimate of how many units it was recommended their gender should not exceed in a day, which corresponded to a lower level of awareness than in 2009. This suggests that previous efforts to raise awareness of recommended drinking limits have not had lasting effect.24 In Australia, 53.5% correctly identified the guideline threshold for women and 20.3% did so for men.25

Knowledge on drinking guidelines in Ireland is also poor. In 2012, just 10% of men and 10% of women knew the gender-specific low-risk limits for alcohol consumption.23 Ireland’s guidelines were last reviewed in 2009.26 The current guidelines recommend that men consume no more than 17 standard drinks and women no more than 11 standard drinks spread over the course of a week, with at least two alcohol free days. No guidance is given in relation to daily low-risk limits. These results suggest that further work on educating the Irish public on low-risk drinking limits is required. Given the high prevalence of frequent RSOD in Ireland, it may also be appropriate to introduce low-risk daily limits. Drinkers in Ireland tend to consume alcohol relatively infrequently but, on the occasions that they do, they are likely to engage in RSOD. In order for individuals to monitor and be aware of their alcohol consumption, knowledge on the standard drink concept and low-risk drinking guidelines is required. It is unrealistic to expect people to stay within low-risk limits and to be able to accurately assess their own hazardous or harmful drinking in the absence of knowledge on what actually constitutes hazardous or harmful drinking.

Policy implications

Public health messaging can be utilised to provide health guidance regarding alcohol use to the general public. A systematic review on the effectiveness of mass media public health campaigns to reduce alcohol consumption and related harms found evidence that such campaigns can be recalled by individuals and can achieve improvements in knowledge about alcohol. There was no evidence that campaigns led to decreased alcohol consumption but the authors concluded that mass media can yield sustained knowledge, which may lay the groundwork for reductions in consumption that are achieved using other public health measures.27 In Denmark, a repeated annual campaign from 1990 to 2000 increased awareness of low-risk drinking guidelines in all subsets of the population throughout the period.28 Hazardous drinkers were more knowledgeable about the guidelines than low-risk drinkers, which shows that this important target group can be reached. There had been limited public health messaging in Ireland on low-risk drinking prior to 2017, when an alcohol campaign ‘Ask About Alcohol’ was commenced to provide clear and authoritative information on alcohol to the public across a number of media platforms. The website for this campaign is the first one dedicated to dealing with alcohol to be created by a State body in Ireland. It provides advice on low-risk drinking limits and contains a drinks calculator so the public can understand exactly how much they are drinking and whether it is within low-risk limits.

This study demonstrates that further initiatives to reduce overall consumption and hazardous and harmful drinking patterns and raise awareness around drinking patterns are required. Based on the existing systematic review evidence on mass media campaigns,27 simply having a public messaging campaign around hazardous and harmful drinking is insufficient to reduce alcohol consumption and problematic patterns of drinking in an alcogenic culture such as Ireland, where proalcohol social norms and alcohol marketing and sponsorship are pervasive. Older people and those with lower educational attainment were less likely to be aware of their hazardous or harmful drinking and efforts should be made to target this group in relation to raising awareness around alcohol use. Our results also suggest that there is a cohort of younger, well-educated drinkers in Ireland who also engage in other potentially risky behaviours and that these subjects are already aware of their hazardous or harmful drinking. Consequently, it is unlikely that public health messaging alone will be sufficient to result in behaviour change for this group in relation to their alcohol use. In 2018, following a protracted process, the Public Health (Alcohol) Act was signed into law. This is the first time that Ireland’s harmful use of alcohol will be addressed coherently in public health legislation. The main provisions of the Act include the introduction of a minimum unit price for alcohol, restrictions on the advertising and sponsorship of alcohol products, the structural separation of alcohol from other non-alcohol products in small shops, convenience stores and supermarkets, and labelling of all alcohol products to provide consumers with information on the number of grams of alcohol per container, calorific content and health warnings. These measures will be enacted over the coming years with the aim of reducing alcohol consumption in Ireland. However, it is important that these initiatives are accompanied by public health messaging. If a comprehensive and sustained public health messaging campaign is implemented alongside the provisions in the Public Health Alcohol Act, the likelihood of both raising awareness and achieving meaningful reductions in alcohol consumption and problematic drinking patterns will be increased. It is also imperative that evaluations on the effectiveness of the legislative measures and the public messaging campaign are undertaken regularly to assess their impacts.

Conclusions

The results of this study indicate that a large proportion of Irish drinkers are not aware that they are consuming alcohol in a way that is potentially damaging to their health. It is likely that sustained public health messaging alongside evidence-based policy measures around pricing, availability and marketing are required to bring about behaviour change among the Irish drinking population.