The Scottish Government

« Previous | Contents | Next »

Listen

SUMMARY OF REVISIONS

• The average strength by volume of some alcohol products has been increasing in recent years. As a result there has been some concern that the methods used by surveys to convert the amounts of alcohol adults report having consumed into units have been underreporting consumption levels.
• This report uses the new alcohol unit conversion factors developed by the Office for National Statistics to present an update of the 2003 Scottish Health Survey alcohol consumption chapter published in 2005 (Volume 2: Chapter 1).
• Using the previous unit conversion factors 27% of men reported their usual alcohol consumption as being more than the recommended limit of 21 units per week. Using the new conversion factors shows this to be 34%. The impact of using the updated method was greatest for men in the age groups 25-34 to 45-54, the proportion in these groups exceeding the weekly limit increased by 7-9 percentage points. In contrast, the corresponding increase for all other age groups was 5 or 6 percentage points.
• The proportion of women who reported drinking more than the recommended limit of 14 units per week was 14% using the previous factors and 23% with the updated factors. The impact of using the updated method was greatest for women in all age groups between 16-24 and 54-65, the proportion in these age groups exceeding the weekly limit increased by 9-12 percentage points. In contrast, the proportion of women aged 75 and over exceeding weekly limits increased by just 2 points, from 3 to
  5 percent.
• On average, men drank 20.3 units of alcohol per week, an increase of 18% on the 17.2 units estimated using the previous factors.
• Women drank on average 9.1 units per week, an increase of 40% on the 6.5 units per week previously estimated.
• Boys aged 13-15 reported drinking 1.0 units per week, an increase of 25% on the 0.8 units previously estimated. The increase was larger for girls aged 13-15, whose reported consumption increased by 30% from 1.0 to 1.3 units.
• The estimates based on the previous conversion factors showed that levels of consumption were highest among women in managerial and professional households and in the highest income quintile; consumption decreased along with household income, and was lowest for women in semi-routine and routine occupations. The impact of the updated method is greatest for women in managerial and professional households, women in the highest income quintile, and in the least deprived 20% of areas. As a consequence the socio-demographic differences in women's weekly drinking previously reported are now even more pronounced.
• Using the previous calculation method there was no consistent pattern in men's levels of consumption for either socio-economic classification or household income. As found for women - though to a slightly lesser degree - the updated method had the greatest impact on estimates for men in managerial and professional households, men in the highest income quintile, and in the least deprived 20% of areas. The updated method of calculation therefore reveals a pattern for weekly drinking and socio-demographic factors for men that was not previously evident. The proportion of men exceeding weekly limits decreased as household income declined and as deprivation increased. Men in professional, managerial and intermediate households were more likely to exceed weekly limits than men in other types of household.
• Turning to consumption on the heaviest drinking day in the past week, two changes to the way the estimates have been calculated need to be highlighted. The methods for calculating daily unit consumption have been updated using the same conversion factors used for the weekly drinking estimates. However, the thresholds for reporting daily unit consumption have also changed.
  The 1998 and 2003 chapters defined daily drinking above recommended limits as 3 or more units for women and 4 or more for men, and defined 'binge' drinking as consuming 6 or more units for women and 8 or more for men. These thresholds have now been adjusted to use the following thresholds to enable comparability with surveys of alcohol consumption across Britain of:
• Drinking above 'safe' daily drinking limits - women: more than 3 units; men: more than 4 units
• 'binge' drinking - women: more than 6 units; men: more than 8 units.
• Using the revised conversion factors shows that similar proportions of men (63%) and women (64%) who had drunk in the previous week had exceeded the daily benchmarks for their sex, although men were still more likely to binge drink than women (40% and 33% respectively).
• Among men who had drunk alcohol during the preceding week, using the previous unit conversion factors 54% said they drank more than 4 units of alcohol and 30% said they drank more than 8 units on their heaviest drinking day. The corresponding figures using the updated unit conversion factors are 63% and 40%.
• The impact of the updated conversion factors for men's daily drinking followed no consistent pattern; the percentage point increase was greatest for men aged 35-44 and those aged 75 and over. The impact on binge drinking was a little clearer: the percentage point increase in the proportion of men drinking more than 8 units a day was greatest for those in the age groups from 25-34 to 45-54 (up 12 points), while the impact on men aged 75 and over was much less (up just 4 points).
• Among women who had drunk alcohol during the preceding week, using the previous unit conversion factors 44% reported drinking more than 3 units on their heaviest drinking day, and 19% drank more than the binge drinking definition of 6 units. Using the updated unit conversion factors these figures were 64% and 33% respectively.
• The impact of using the revised conversion factors was greater for women's daily drinking estimates than for men's. The impact on daily drinking was greatest for women in the age groups from 25-34 to 55-64 (an increase of between 21 and 26 points) and was much less pronounced for women in the youngest and oldest age groups. The impact on binge drinking levels was greatest among those in the 25-34 to 45-54 year old age groups, the proportion of these groups who reported drinking more than 6 units a day increased by 17 points using the updated revised factors. The corresponding increase for those in the 16-24 and 55-64 age groups was 12 points. The updated method made no difference to estimates of binge drinking among women aged 75 and over, the proportion was just 1% with both methods.
• The impact of using the revised conversion factors was similar across all NS-SEC groups for men.
  In contrast, the impact of using the revised factors was greater for women in professional/managerial and intermediate households than it was for other groups. As a consequence of using the revised factors a consistent pattern between daily drinking and NS-SEC emerges for women.
• The pattern between household income and daily drinking was not consistent before the revised factors were applied. The impact of using the revised factors was greatest for men and women in the highest income households which resulted in a clearer pattern emerging of consumption being greatest among those in the highest income households.
• The impact of using the revised conversion factors was greatest for men and women in the least deprived 20% of areas. As a consequence the pattern between men's daily drinking and deprivation was less pronounced than it had appeared using the original factors. The impact was even greater for women, whereas the original factors suggested a relationship between deprivation and consumption, as a result of using the revised factors there was no longer a consistent pattern.

1.1 INTRODUCTION

Misuse of alcohol contributes to a wide range of health and social problems, including high blood pressure, liver cirrhosis, pancreatitis, cancer, suicide, accidents, and antisocial behaviours including crime and domestic violence. The cost of alcohol misuse to Scottish society in the year preceding the 2003 Scottish Health Survey (2002-03) was estimated as £1.13 billion. ¹ More recent estimates, for the year 2006-07, now put this figure at around £2.25 billion. ² This figure includes costs to NHS Scotland (£405 million), social work services (£170 million), criminal justice and emergency services (£385 million), and wider economic costs, such as absenteeism, health problems and deaths among the working age population (£820 million). The human costs of deaths among the non-working age population were estimated at £470 million. No estimates for the value of the reduced quality of life associated with alcohol-related conditions were included.

Some of these adverse outcomes are associated most with sustained heavy drinking, whereas some may be linked with excessive drinking on one or more occasion. Thus public health advice about alcohol consumption is framed in relation to recommended weekly limits, and safe daily drinking patterns.

In terms of weekly limits, men are advised to drink no more than 21 units of alcohol per week, and women no more than 14. The 1999 White Paper "Towards a Healthier Scotland" set out the Executive's aim to reduce the incidence of adults exceeding the weekly limits to 31% of men and 12% of women by 2005, and 29% of men and 11% of women by 2010. ³

In terms of daily consumption, regular drinking of more than 3 or 4 units a day for men, or more than 2 or 3 for women, is likely to result in increasing health risk and is not advised. ⁴ There is no significant health risk for adults who regularly consume less than these amounts, though two alcohol free days a week is also advised. For men aged over 40 and post-menopausal women, drinking up to 2 units a day can have a moderate protective effect against heart disease. ⁵

Although there is no standard definition of 'binge drinking' in the UK, Alcohol Concern recommends using the Office of National Statistics' definition of heavy drinking: 6 or more units for women and 8 or more for men. To enable comparisons to be drawn with other major surveys of alcohol consumption in Great Britain, this report has adopted the binge drinking definitions employed in the General Household Survey and the Health Survey for England of more than 8 units for men and more than 6 for women. Concerns about excessive drinking of this type were evident in the Scottish Executive's 2002 "Plan for Action on Alcohol". ⁶ There continues to be considerable concern about the health and social harm associated with a range of excessive drinking patterns, including 'binge drinking' and regular drinking above the daily benchmarks and weekly limits.

The consumption targets introduced in 1999 have now been superseded by the national indicator to reduce alcohol-related hospital admissions introduced as part of the Scottish Government 2007 Spending Review ⁷ and a new NHS performance monitoring ( HEAT) target to deliver alcohol "brief interventions" during the period 2008/09-2010/11. ⁸
In addition to this, the Scottish Government will be publishing a discussion paper on its long-term strategic approach to tackling misuse in summer 2008 and will be working with NHS Health Scotland to develop a range of indicators (including consumption measures) as part of their on-going monitoring and evaluation of the strategy.

When the Scottish Executive published its "Plan for Action on Alcohol Problems: Update" in February 2007, concerns were raised about the Scottish Health Survey's suitability for measuring trends in alcohol consumption due to its intermittent reporting. ⁹ However, the survey will be running continuously between 2008 and 2011 so concerns about the regularity of its data collection can now be allayed somewhat. A more fundamental concern relates to the ability of surveys to capture alcohol consumption levels accurately. These concerns extend beyond the weaknesses typically associated with surveys such as social desirability biases and recall errors on the part of respondents. Instead, the methods used by surveys to convert the amounts of alcohol informants report having consumed into units have been questioned. ¹⁰ Recent increases in the typical strengths and serving size of alcohol products mean that some of the unit assumptions that were devised for use in the General Household Survey ( GHS), the Health Survey for England ( HSE) and the Scottish Health Survey ( SHeS) are no longer valid and are, as a consequence, contributing to the underreporting of alcohol consumption. In December 2007, the Office for National Statistics published updated conversion factors for converting drink volumes to alcohol units ¹¹ and these are now being used to estimate alcohol consumption in these three surveys. To enable comparability with estimates from the GHS and HSE, and to better reflect the current position relating to alcohol consumption, this new method has been applied to the 2003 Scottish Health Survey data and the results are reported in this chapter. These updated estimates for 2003 replace those originally published in 2005. ¹²

1.2 METHODS

1.2.1 Data collection

Alcohol consumption was measured in relation to four key aspects: usual weekly consumption, daily consumption on the heaviest drinking day, indicators of problem drinking (including physical dependence) and the social context of drinking.

The module of drinking questions in 2003 followed that used in 1998. This differed in a number of ways from the 1995 Scottish Health Survey questions (see 1998 report for more details).

To estimate weekly consumption, informants aged 16 and over, after preliminary questions on whether they drank alcohol at all, were asked how often during the past 12 months they had drunk each of six different types of alcoholic drink:

• normal beer, lager, cider and shandy
• strong beer, lager and cider
• sherry and martini
• spirits and liqueurs
• wine
• alcoholic soft drinks ("alcopops").

From this question, the average number of days a week the informant had drunk each type of drink was estimated. A follow-up question asked how much of each drink type they had usually drunk on each occasion. This data was converted into units of alcohol and multiplied by the amount they said they usually drank on any one day (see below for discussion of this process). ¹³

It is well known that surveys tend to underestimate adults' levels of alcohol consumption for a number of reasons, including problems of memory, social desirability, and the difficulties involved in assigning an average estimate to an activity that varies from day to day. ¹⁴ Also, as the questions ask about 'usual' behaviour, responses are unlikely to reflect occasions of heavier drinking. Nevertheless, survey estimates provide useful comparisons of the consumption of different population groups and enable change over time to be monitored.

Daily consumption was measured by asking about drinking in the week preceding the interview, and looked at actual consumption on the heaviest drinking day in that week. Informants aged 16 and over were asked whether they had drunk alcohol in the past seven days. If they had, they were asked on how many days and, if on more than one, whether they had drunk the same amount on each day or more on one day than others. If they had drunk more on one day than others, they were asked how much they drank on that day.
If they had drunk the same on several days, they were asked how much they drank on the most recent of those days. If they had drunk on only one day, they were asked how much they had drunk on that day. In each case, the questions asked for details of the amounts consumed of each of the six types of drink (similar to those used for establishing average weekly consumption), rather than asking informants to give a direct estimate of units consumed.

The CAGE questionnaire was asked of informants aged 16 and over, and highlights up to six indicators of problem drinking, including three indicators of physical dependency on alcohol. This questionnaire was administered in self-completion format due to the sensitive nature of the questions. Questions were also included, for the first time, about the social context of drinking, that is, where informants usually are when they drink alcohol and who they are usually with.

1.2.2 Unit calculations and new conversion factors

In the UK, a standard unit of alcohol is 10 millilitres or around 8 grams of ethanol; it is used to indicate the potency of an alcohol product. As described above, the majority of advice given in relation to safe alcohol consumption refers to units. The need for accurate estimates of units consumed is therefore paramount. However, there are numerous difficulties associated with calculating units at a population level, not least of which is the variability of alcohol strengths and the fact that these change over time.

The strength of an alcohol product is usually expressed as a percentage "alcohol by volume" ( ABV) figure based on the number of units in a litre. A litre of spirits with an ABV of 40% will therefore contain 40 units, while a litre bottle of wine with an ABV of 10% will contain 10. However, the ABV of similar types of alcohol products can also vary quite markedly, for example wine is typically in the range of 11.5%-14.5% ABV. ¹⁰ When the original factors for converting the volumes reported by survey informants into units were devised the ABV assumed for wine was 8% meaning that a 750ml bottle of wine of this strength would have contained 6 units. In contrast, a typical bottle of wine nowadays has an ABV of around 12% and contains 9 units.

To take account of changes in the strength of some alcohol products the Office for National Statistics undertook a review of the existing methodology for converting volumes into units in the General Household Survey and the Health Survey for England (on whose methodology the Scottish Health Survey is based). This work resulted in the publication of new unit conversion factors based on average strengths of alcohol in 2006; these factors have also been applied retrospectively to the estimates from the 2005 GHS and updated figures have been published. ¹¹ These factors have been adopted for use in the HSE from 2006 onwards and the Scottish Health Survey from 2008 onwards. The most recent SHeS data from the 2003 survey, published in 2005, has also been reanalysed using the new conversion factors and this report presents the results of this exercise.

As described above, information was collected about the volumes of alcohol informants had drunk in a typical week and also on their heaviest drinking day in the week preceding the survey. The following table outlines how the volumes of alcohol reported in the 2003 SHeS were originally converted into units and how the new conversion factors have changed this. The first point to note is that the new conversion factors only apply to the following products/volumes: small cans or bottles of normal beer, half pints or small cans/ bottles of strong beer, wine and alcopops.

Source: adapted from Table 1 in Goddard, 2007 and Table presented in HSE 2006 alcohol chapter ^{11, 15}

a The GHS used a conversion factor of 2.3 for large cans/bottles of strong beer whereas SHeS and HSE have always used 3.

^b From 2008 onwards SHeS will record glass sizes for wine consumption (125ml, 175ml and 250ml).
In 2003 wine was not recorded in bottle volumes but interviewers were instructed that a bottle contained 6 units.

1.2.3 Daily drinking threshold changes

The previous section outlined the process of updating the unit conversion factors for the volumes of alcohol informants reported consuming. In addition to this change the thresholds used to define daily unit consumption in excess of recommendations have also changed. Official guidance advises women not to regularly consume 3 or more units a day and men not to regularly consume 4 or more. ⁴ The previously published figures defined daily drinking above these recommended amounts as 3 or more units for women and 4 or more units for men. Heavy drinking, used as a proxy for 'binge' drinking, was defined as consuming 6 or more units for women and 8 or more units for men (i.e. twice the levels used to report daily drinking). To enable comparability with the Health Survey for England and the General Household Survey these thresholds have now been adjusted, as set out in the table below. The daily drinking estimates in this report are therefore based on both the updated unit factors and these new thresholds.

The survey did not ask informants how regularly they drank the amounts reported for their heaviest drinking day in the previous week so it is not entirely accurate to infer that someone who drank more than these amounts on one day in the previous week is necessarily in contravention of the official guidance about safe daily drinking patterns. However, it is probably appropriate to consider someone reporting daily consumption levels greatly in excess of these amounts as being at heightened risk of unsafe or harmful drinking behaviour.

The rest of this report updates and replaces the original figures published in 2005 and, where possible, follows the format of the original chapter. The impact of using the new conversion factors on the various estimates and sub-group analyses is also discussed. Summary tables illustrating the impact of the changes in percentage terms are also presented. More detailed tables of the changes are available at: http://www.scotland.gov.uk/Topics/Statistics/Browse/Health/scottish-health-survey/Publications . It is important to note from the outset that all the new estimates of the amount of alcohol consumed in Scotland presented here are based solely on changes to the strength of the alcohol consumed; no changes have been made to the estimates of the overall volume of alcohol products consumed.

1.3 ESTIMATED WEEKLY CONSUMPTION

1.3.1 Usual consumption by age and sex

Table 1.1 shows self-reported usual weekly drinking habits of men and women by age group. Men were more likely to drink, and to report drinking greater amounts of alcohol, than women. Among women, 13% said they did not drink at all compared with 8% of men. Women were also more likely than men to report drinking less than one unit per week (16% of women and 7% of men).

In terms of age, the proportion of non-drinkers was relatively low (10% or less) until the 75+ age category among men, when it increased to 20%. It was also under 10% among women until the 55-64 age category when it increased to 14%, and it then continued to increase to 21% among 65-74 year olds and to 32% among women aged 75 and over.

The majority of men and women reported drinking less than the recommended limits. About one in three men (34%) and one in four women (23%) reported drinking more than the recommended weekly limit of 21 units for men and 14 for women. On average, men drank 20.3 units of alcohol per week, and women drank 9.1 units.

The proportion of men and women exceeding the weekly limits decreased with age; the pattern for both genders was broadly similar, albeit with different overall consumption levels. Among men, the proportion who said they exceeded 21 units per week was fairly uniform (at around 36%) up to the 55-64 age category, after which the proportion decreased to 27% of those aged 65-74, and 21% of those aged 75 and over. The pattern for reported mean consumption was similar.

Among women, it was the youngest four age groups (16-24 to 45-54) who were the most likely to report drinking more than the recommended 14 units per week (ranging between 27% and 32%). This proportion then decreased to 21% at age 55-64, 13% at 65-74, and then to only 5% for women aged 75 and over. Reported mean consumption also decreased with age, from between 10.3 and 12.2 units for the 16-24 to 45-54 age groups to 2.7 units among women aged 75 and over.
Table 1.1, Figure 1A

Figure 1A Proportion exceeding government guidelines on weekly alcohol consumption (21 units for men, 14 units for women), by age and sex

1.3.2 Trends over time in usual consumption

It is as yet unclear at what point in the time series the previous unit conversion factors started to depart from the assumptions on which they were based but it is likely to have been a gradual process over time rather than a sudden jump. While it was felt appropriate to revise the 2003 data in line with the new conversion factors, it is questionable whether applying factors based on strengths of alcohol products in 2006 to data collected in 1998 and 1995 would be a valid exercise. Without further information to identify when survey assumptions of drink size and strength started to depart from what was available to people, it is not possible to be definitive about recent drinking trends in Scotland. However, comparing the revised estimates for 2003 with the original 1995 and 1998 figures, although problematic, provides a better picture of drinking in Scotland than the previous data allowed.

Table 1.2 shows usual weekly alcohol consumption levels in the 1995, 1998 and 2003 Scottish Health Surveys. The upper age limit for adults has changed over the three survey years. The sample in 1995 included adults aged 16-64; in 1998 the upper age limit was extended to 74, and in 2003 it was removed. Comparisons between the three survey years are therefore limited to the age range common to all of them (16-64).

It is difficult to interpret trends from a table which has only three data points, as the estimate for a particular year may be an outlier at the extreme of the sampling error distribution. For example, looking at men, the proportion of 16-24 year olds who reported exceeding 21 units a week showed a large increase between 1995 and 1998, and a similar decrease between 1998 and 2003. However, it is not possible to say whether the 1998 estimate is unusually high or the 1995 and 2003 estimates are particularly low. Looking at the consumption of all men aged 16-64 suggests a pattern of increasing consumption over time (albeit marginal), from 33% in 1995, to 34% in 1998, and 36% in 2003.

The pattern for women aged 16-64 is more consistent, showing a continual increase between 1995 and 2003, with the increase between 1998 and 2003 being the most notable: the proportion exceeding 14 units per week increased from 13% in 1995 to 15% in 1998, and again to 28% in 2003. The size of the increase between 1998 and 2003, relative to that between 1995 and 1998, will in part be explained by the use of revised conversion factors in 2003. This is discussed in more detail in Section 1.3.6 below.

For men, the survey results are not encouraging in terms of meeting the target set by the Scottish Executive in 1999 of no more than 31% of men aged 16-64 drinking more than 21 units a week by 2005: the proportion of men who reported exceeding this level was 36% in the 2003 survey. Women, the survey suggests, were also moving away from the target figure of a maximum of 12% drinking more than 14 units a week: in 2003, 28% of women aged 16-64 exceeded the recommended level, compared with 15% five years previously.
Table 1.2

1.3.3 Usual alcohol consumption by socio-demographic factors

To ensure that the comparisons presented in this section are not confounded by the different age profiles of the sub-groups, the data have been age-standardised. On the whole the differences between observed and age-standardised percentages are small. Therefore, the percentages presented in the text below are observed (un-standardised). The tables report both the observed and the age-standardised figures. The data presented in these sections in the version of this chapter published in 2005 were age-standardised.

Socio-economic classification ( NS-SEC)

National Statistics-Socio-Economic Classification ( NS-SEC) is a classification of social position that was introduced in the 2001 population census. It has similarities to Registrar General's Social Class. Informants are assigned to an NS-SEC category based on the current or former occupation of the household reference person. The five categories presented in this report include: managerial and professional, intermediate (e.g. clerical, administrative, sales), small employers and own account workers, lower supervisory and technical, and semi-routine and routine occupations.

Among women, there was a clear relationship in self-reported levels of alcohol consumption and NS-SEC with consumption levels higher among women in managerial and professional households than those in lower supervisory and technical and semi-routine and routine households. Women in managerial and professional households were the most likely to exceed 14 units per week (33%) and had the highest mean consumption levels (12.0 units). Mean unit consumption levels were lowest among women in lower supervisory and technical households (7.4 units) and semi-routine and routine households (6.6 units). The proportions exceeding 14 units in these two groups were 18% and 15% respectively.

Among men, there were some differences reported in mean consumption (e.g. mean units ranged from 17.5 for those in small employer/own account worker households to 21.7 for those in managerial/professional households) but the pattern was not as clear as for women. However, men in managerial and professional households (39%) and intermediate households (40%) were more likely to exceed 21 units per week than men in other types of households (ranging from 30% to 32%).
Table 1.3

Equivalised household income

Equivalised annual household income is a measure of household income that takes account of the number of persons in the household. The income data presented here are based on quintiles.

Among women, there was a clear relationship between equivalised household income and reported weekly consumption levels. The proportion of women who said they drank more than 14 units a week decreased with income, from 40% of those in the highest quintile, to 15% and 16% of those in the fourth and fifth lowest quintiles, respectively.

A similar pattern was seen for men: the proportion reporting consumption of more than 21 units a week decreased from 45% in the highest quintile to 25% in the lowest. Although mean consumption levels did not show as obvious a pattern with equivalised income quintiles, they were nevertheless highest among men in the highest quintile (23.3 units) and lowest amongst men in the fourth (17.7 units) and fifth (19.0 units) quintiles.
Table 1.4, Figure 1B

Figure 1B Proportion exceeding government guidelines on weekly alcohol consumption (observed), by equivalised household income quintile and sex

Scottish Index of Multiple Deprivation ( SIMD)

The Scottish Index of Multiple Deprivation ( SIMD) is the Scottish Government's official measure of area-based multiple deprivation. It is based on 31 indicators in six individual domains of current income, employment, housing, health, education, skills and training and geographic access to services and telecommunications. SIMD is calculated at data zone level, enabling small pockets of deprivation to be identified (data zones have between 500 and 1,000 people living in them). ¹⁶ The data zones are ranked from most deprived (1) to least deprived (6505) on the overall SIMD index. The result is a comprehensive picture of relative area deprivation across Scotland. SIMD is reported here using quintiles.

The relationship between SIMD and drinking behaviour was similar to that found for
NS-SEC and household income. The proportion of women who reported drinking more than 14 units per week was highest among those in the least deprived quintile (32%) and declined to 18% in the most deprived, quintile. Mean consumption reported by women in the two least deprived quintiles was higher than in the three most deprived quintiles, though the differences were on the whole small (ranging from 10.9 units in the second quintile to 7.6 units in the fifth).

For men there was a little less variation between deprivation quintiles, especially in the mean units consumed per week. The proportion exceeding 21 units a week varied more, ranging from 38% in the least deprived quintile to 30% in the most deprived quintile. The corresponding figure for the other three quintiles was 34%.
Table 1.5

1.3.4 Frequency of consumption by age and sex

Men and women were asked a separate question about how frequently they usually drank alcohol. Men were more likely than women to drink almost every day (14% compared with 8%), and at the other extreme, women were more likely than men to never drink or to do so at most a couple of times a year (24% compared with 13%). About one third (35%) of men and one fifth (22%) of women drank on three or more days a week.

The frequency of drinking varied by age for both sexes. Young men were most likely to drink once or twice a week (44% of those aged 16-24), while very few (3%) drank almost every day. The proportion drinking once or twice a week decreased, while drinking every day increased, with age. Drinking patterns were more polarised among older men: among the oldest category of men (75 and over), 29% drank almost every day, while a similar proportion (30%) never drank or did so only once or twice a year.

The pattern for women was similar, albeit at a lower frequency. The likelihood of drinking every day increased along with age, from 1% for women aged 16-24 to 11% among those aged 75 and over. However, women aged 75 and over were three times as likely as those aged 16-24 to never drink or drink only a couple of times a year (51% versus 16%).
Table 1.6

1.3.5 Types of drink

Men's reported usual weekly consumption of alcohol was made up chiefly of normal strength beer (9 units), followed by wine (5 units) and spirits (4 units).

Preferences for different types of drink varied considerably by age. Normal strength beer accounted for just over half of the reported weekly alcohol consumption by men in the youngest age group (11 out of 20 units), but this decreased to less than one quarter of the alcohol consumed by men aged 75 and over (3 out of 13 units). The pattern for wine consumption was rather different, it constituted just one of the total units consumed by men aged 16-24 whereas it ranged between 4 and 7 units for those in the age groups 35-44 and up. Although the average number of units of spirits consumed by men in the different age categories did not vary much (between 3 and 6 units), spirits made up a fifth of the
20 units consumed by young men aged 16-24 compared with just under a third of the 13 units consumed by men aged 75 and over.

Women's total weekly alcohol consumption (9 units) is largely comprised of wine (5 units), followed by spirits (2 units). The amount of alcohol reported by women decreased with age for all types of drink except wine which showed an increase until the age group 45-54
(7 units per week) and then declined thereafter. Among young women (aged 16-24), spirits was the most common drink (4 units per week) followed by wine and normal strength beer (3 units of each per week). Wine was the most common drink for all other age groups, except those aged 75 and over who drank wine and spirits in equal amounts.
Table 1.7

1.3.6 Impact of the unit recalculations on weekly drinking estimates

Although the tables discussed in this section present the changes that result from using the revised factors in percentage terms, the text below uses the percentage point changes between the two methods to help interpret the overall impact of using these factors on proportions (means are described using percentage changes). Where overall proportions are small in the first place, percentage point changes are a better indicator of overall impact than raw percentage increases which can exaggerate the scale of a difference between two figures (e.g. an increase of just 2 percentage points from 2 to 4 is a 100% increase).

Usual consumption by age and sex

Table 1.1A shows that the previous method to calculate units estimated that 27% of men reported their usual alcohol consumption as being more than the recommended limit of 21 units per week. The updated method shows this to be 34%. The impact of using the revised factors is greatest for men in the age groups 25-34, 35-44 and 45-54, the proportion in these groups who exceed the weekly limits increases by between 7 and 9 percentage points. In contrast the corresponding increase was 5 or 6 percentage points for all other age groups.

The proportion of women who reported drinking more than the recommended limit of 14 units per week increased from 14% using the previous method to 23% with the revised factors. The impact of using the updated method is greatest for women aged under 65, the proportion of women in this age group who exceed the weekly limit increased by between 9 and 12 percentage points. In contrast, the corresponding increases among women aged 65-74 and 75 and over were just 5 and 2 percentage points respectively (from 8 to 13 percent and from 3 to 5 percent).

Figure 1C Proportion exceeding government guidelines on weekly alcohol consumption (21 units for men, 14 units for women), by age and sex - original and revised units

On average, men were previously estimated to drink 17.2 units of alcohol per week; using the revised conversion factors increased this figure to 20.3 units, an increase of 18%. Previous estimates suggested that women drank on average 6.5 units per week; this increased to 9.1 units based on the new conversion factors, an increase of 40%.
Table 1.1A, Figure 1C

Although men consume more than twice as many units per week as women on average (20.3 and 9.1 respectively) the overall impact of the new conversion factors is much greater for women than for men. This is accounted for by the fact that the biggest change made to the conversion factors is for wine, which accounts for more of women's total alcohol consumption than it does for men's. As noted above in section 1.3.5, just over half of women's total weekly alcohol consumption (9 units) is comprised of wine (5 units). In contrast, at 5 units, wine represents just a quarter of men's total usual weekly alcohol consumption.
Table 1.7, Table 1.7A

Trends over time in usual consumption

As noted at the start of Section 1.3.2, only the 2003 estimates have been revised using the new conversion factors. The 1995 and 1998 figures are unchanged. Comparing the revised estimates for 2003 with the original 1995 and 1998 figures is problematic but it provides a better picture of drinking in Scotland than the previous data allowed.

The above comparisons result in a number of different conclusions from those based on the original 2003 estimates. First, the previous estimates reported 29% of men aged 16-64 to be consuming more than 21 units per week and that this represented a decrease over time. Secondly, on the basis of this figure of 29%, the target set in 1999 of 31% of men exceeding the weekly limit by 2005 appeared to have been met (on the assumption that major reversals of this trend would not occur before 2005). The revised figure for men aged 16-64 is 36% therefore both of the preceding conclusions are no longer valid. A slightly different picture emerges for women aged 16-64; the proportion estimated to exceed 14 units of alcohol a week was originally estimated to be 17% while the revised estimate is 28%. Although the conclusions based on the previous estimates still stand - that women's drinking has increased over time and has moved in the opposite direction of the target of 12% of women exceeding 14 units per week by 2005 - the revised estimates suggest that the scale of this increase is much greater than previously thought. Although these targets have now been superseded by those set out in the Scottish Government's new performance framework, ⁷ it is still useful to review the conclusions that were previously drawn from the 2003 data.
Table 1.2

Usual alcohol consumption by socio-demographic factors

The estimates based on the previous calculation method showed that levels of consumption were highest among women in managerial and professional households and in the highest income quintile; consumption decreased along with household income, and was lowest for women in semi-routine and routine occupations. The impact of the updated method is greatest amongst women in managerial and professional households, women in the highest income quintile, and in the least deprived 20% of areas. As a consequence the socio-demographic differences in women's weekly drinking previously reported are now even more pronounced. Figure 1D illustrates the impact of the changes in relation to equivalised household income that are presented in Table 1.4A. It shows that using the new conversion factors results in the proportion of women in the highest income quintile drinking more than 14 units almost doubling, from 22% to 40%.

Using the previous calculation method there was no consistent pattern in men's levels of consumption in terms of the mean units consumed for either socio-economic classification or household income, though for the latter there was a decrease in the proportion of men exceeding 21 units per week as income decreased (from 32% in the highest quintile to 23% in the lowest). As found for women - though to a slightly lesser degree - the updated method has the greatest impact on estimates for men in managerial and professional households, men in the highest income quintile, and in the least deprived 20% of areas. The revised conversion factors therefore reveal a pattern for weekly drinking and socio-demographic factors for men that was not previously evident, while the finding in relation to household income and the likelihood of exceeding 21 units is now even more pronounced (as shown in Figure 1D and discussed above in Section 1.3.6).
Tables 1.3A, 1.4A, 1.5A, Figure 1D

Figure 1D Proportion exceeding government guidelines on weekly alcohol consumption (observed), by equivalised household income quintile and sex - original and revised units

1.4 ESTIMATED DAILY CONSUMPTION

1.4.1 Alcohol consumption on the heaviest drinking day in the last week by age and sex

The proportions of men and women who said they had drunk alcohol in the past week were 72% and 58% respectively. Among men who drank in the past week, 24% had drunk on only one day and 23% on two days, with decreasing proportions for three (15%), four (11%), five (6%) and six days (5%) and then increasing again to 16% having drunk on all seven days. The mean number of drinking days was 2.4 out of seven days (based on all men). Among women, the mean was lower at 1.6 days in the last seven. Among those who drank in the past week, women were more likely than men to drink on only one day (36%), and only 10% drank on all seven days. (Table not shown.)

As in the 1998 survey, among informants who drank in the past week, the proportion drinking every day increased with age: for men, from 3% at age 16-24 to 45% at age 75 and over; for women, from 2% to 28% respectively. However, those in the oldest age group were also the most likely to say they had not had a drink at all in the past week: 44% of men and 67% of women. (Table not shown.)

Informants who had an alcoholic drink in the past week were asked about how much they had drunk on their heaviest drinking day; their responses were converted to units of alcohol, as described in Section 1.2. Table 1.8 shows amounts consumed on the heaviest (or only) drinking day, for informants who had a drink in the past week. The survey data are not sufficient to determine whether current advice on sensible drinking is being followed (as described in Sections 1.1 and 1.2), since the advice refers to 'regular' drinking rather than consumption on the heaviest drinking day.

'Binge' drinking, that is drinking an excessive amount on a single occasion, is thought to be less healthy than drinking moderate amounts more regularly. Although there is no medically-specified criterion to define binge drinking, the GHS and HSE use more than 8 units for men and more than 6 for women as proxies. These levels are reported on in this section, along with the more than 4 units (men)/more than 3 units (women) per day which are used by the same surveys to provide an indicator of the proportions exceeding the daily benchmarks. See Section 1.2.3 for a discussion of how these thresholds have changed since the 2003 results were originally published.

Among men who drank in the past week, 63% drank more than 4 units on their heaviest drinking day, and 40% drank more than 8 units. (Based on all men, not just those who drank, the proportions are 45% and 29% respectively.) Mean consumption was 9.0 units for those who drank (or 6.4 units based on all men). There was a clear inverse relationship with age among men: the youngest age group had the highest mean consumption levels (12.7 units) and were the most likely to drink more than 4 units (80%) and to binge drink (60%). Mean consumption, rates of binge drinking and the proportion exceeding 4 units decreased for each successive age group, so that by age 75 and over, mean consumption was 4.0 units, 30% drank more than 4 units and binge drinking was found among only 9% of men.

Looking at women who drank in the past week, 64% drank more than 3 units on their heaviest drinking day, and 33% drank more than 6 units. (Based on all women, not just those who drank, the proportions are 37% and 19% respectively.) Mean consumption was 6.2 units (3.5 units based on all women). The age pattern was similar to that for men, but at lower levels. The proportion drinking more than 3 units was highest amongst those aged 16-24 and 25-34, at 79%, and was markedly lower, 17%, among those aged 75 and over. Similarly, binge drinking decreased from 60% among women aged 16-24, to just 1% for women aged 75 and over, and mean consumption fell from 9.7 units to 2.2 units respectively.
Table 1.8, Figure 1E, Figure 1F

Figure 1E Proportion of men who had drunk in the past week who drank more than 4 units, and more than 8 units, on their heaviest drinking day, by age

Figure 1F Proportion of women who had drunk in the past week who drank more than 3 units, and more than 6 units, on their heaviest drinking day, by age

1.4.2 Trends in daily consumption since 1998

As there has been a change to both the unit conversion factors and the thresholds used to report daily drinking, no revisions have been made to the previously published trend data for 1998 and 2003. These are presented in Table 1.9 of the original alcohol consumption chapter published in 2005. ¹²

1.4.3 Daily alcohol consumption by socio-demographic factors

Socio-economic classification ( NS-SEC)

There is no consistent pattern between daily drinking levels and NS-SEC among men. Mean daily consumption and the proportion drinking more than 4 or more than 8 units was lower among men in managerial and professional households and those headed by small employers than in other types of household. For example, the prevalence of binge drinking in men was 37% among those in managerial and professional households and 34% in households headed by a small employer compared with between 43% and 47% among men in the other three types of household. There was a somewhat clearer pattern for women in terms of the proportions drinking more than 3 or 6 units a day, but not with mean daily consumption. Women in professional/managerial households and those in intermediate households were the most likely to exceed 3 units and to binge drink. The proportion of women in professional and managerial or intermediate households who drank more than 3 units was 67% in both cases, compared with around 61% for women in other types of household. The corresponding figures for binge drinking were 34% for women in managerial/professional households, 37% in intermediate households and between 30% and 31% in other households. In terms of mean units consumed, women in intermediate households had the highest daily amount (6.9 units) but no clear pattern exists for women in other groups.

This contrasts with the results found for usual weekly drinking levels described in Section 1.3.3, when the relationship with NS-SEC was more pronounced for both women and men, and was particularly so for women.
Table 1.10

Equivalised household income

Figure 1G shows that men's daily consumption declined as equivalised household income declined until the fifth, lowest, quintile which showed an increase to the level found among men in the highest income quintile. For example, Table 1.11 shows that daily consumption on the heaviest drinking day in the last week declined from 9.7 units in the first (highest income) quintile to 9.2 units in the second highest, 9.0 units in the third and 8.1 in the fourth quintile, but then increased to 9.5 units in the fifth (lowest income) quintile. Similarly, binge drinking showed a marked decrease from 46% in the first quintile to 32% in the fourth, before rising to 43% in the fifth. This is different to the pattern found for usual weekly consumption described in Section 1.3.3, when the proportion of men exceeding 21 units per week decreased consistently from the highest to the lowest quintiles. Though it is worth noting that the patterns for daily and weekly mean unit consumption and income are a little more similar, with both seeing a slight upward increase in the fifth quintile.

Women's daily drinking patterns followed a similar pattern to men's. As shown in Figure 1H, consumption declines as income declines until the fifth income quintile where there is an increase, although not as steep an increase as found for men. As a consequence women in the highest income households were the most likely to have consumed more than 3 or 6 units on their heaviest drinking day, and their mean daily consumption was the highest.
Table 1.11, Figure 1G, Figure 1H

Figure 1G Proportion of men who had drunk in the past week who drank more than 4 units, and more than 8 units, on their heaviest drinking day (observed), by equivalised household income

Figure 1H Proportion of women who had drunk in the past week who drank more than 3 units, and more than 6 units, on their heaviest drinking day (observed), by equivalised household income

Scottish Index of Multiple Deprivation ( SIMD)

Among men, the pattern for SIMD was the opposite to that found for income, with consumption increasing as deprivation increased. For example, mean daily consumption increased from 8.1 units in the least deprived quintile to 10.4 units in the most deprived quintile. This relationship was not found among women where there was no consistent pattern in the proportions drinking more than 3 or more than 6 units per day, or in mean units. The proportions exceeding 3 units were very similar (64% or 65%) in all quintiles except the middle (61%). The proportions reporting binge drinking were the lowest in the 3rd and 4th least deprived quintiles (28% and 29% respectively), and ranged between 34% and 37% in the remaining three. In terms of mean units consumed, those in the middle quintile stand out as having the lowest number (5.7) and those in the most deprived quintile have the highest number (6.8).
Table 1.12, Figure 1I

Figure 1I Binge drinking among adults who had drunk alcohol in the past week (observed), by Scottish Index of Multiple Deprivation quintile and sex

1.4.4 Factors associated with exceeding the daily limit and binge drinking

The original alcohol consumption chapter published in 2005 ¹² used multivariate logistic regression to examine the factors associated with exceeding the daily limit and also with binge drinking. The regressions were run separately for men and women, and examined the following independent variables: age group, NS-SEC, equivalised household income quintile, parental NS-SEC when the informant was 14, and SIMD. Only age and SIMD were found to be significantly associated with exceeding the daily limit, and with binge drinking, once all others were controlled in the models. This analysis was replicated using the data based on the new unit conversion factors and revised thresholds and preliminary results suggest that income and NS-SEC are also associated with daily drinking. Further analysis of this needs to be carried out before firm conclusions can be presented. (Data not shown.)

1.4.5 Daily drinking levels in Scotland and England

Table 1.13 in the original chapter based on the 2003 data ¹² compares daily drinking levels for adults in Scotland and England, based on the 2003 Health Survey for England ( HSE). As the HSE is a continuous survey data is available on an annual basis and the most recent results available are those from 2006. ¹⁷ The HSE times series data has been updated for all years to reflect the new thresholds but only the 2006 data has been analysed using the new conversion factors. It has not been possible, therefore, to update the comparative analysis between Scotland and England previously published.

1.4.6 Impact of the unit recalculations on estimates of daily consumption

As outlined above in Section 1.3.6 in relation to weekly drinking, the following text uses the percentage point change in estimates to gauge the impact of the revised conversion factors, rather than the percentage changes presented in the associated tables.

Alcohol consumption on the heaviest drinking day in the last week by age and sex

Table 1.8A presents estimates of daily drinking based on the following:

A. the original thresholds and original unit conversion factors;

B. the revised thresholds and original unit conversion factors; and

C. the revised thresholds and revised conversion factors.

The data in column A describes the original 2003 results as published previously; the data in column C is described above. Presenting these three sets of data makes it possible to differentiate the impact of the new thresholds (by comparing A and B), the impact of the new method of converting volumes to units (by comparing B and C), and the overall impact of both sets of changes (by comparing A and C). The following discussion, however, focuses solely on the impact of the new conversion factors (B to C).

Table 1.8A shows that, using the previous method to calculate units, 54% of men who had drunk alcohol during the preceding week said they drank more than 4 units of alcohol and 30% said they drank more than 8 units on their heaviest drinking day. The corresponding figures using the updated method are 63% and 40%. The impact of the revised conversion factors on the proportion of men exceeding 4 units follows no consistent pattern; it is greatest for men aged 35-44 and those aged 75 and over. The impact on binge drinking is a little clearer: the percentage point increase in the proportion of men drinking more than 8 units a day is greatest for those in the age groups 25-34 to 45-54 (up 12 points), while the impact on men aged 75 and over is much less (up just 4 points).

Amongst women who had drunk alcohol during the preceding week, the proportion who had drunk more than 3 units was 44% using the previous method, while 19% had drunk more than 6 units on their heaviest drinking day. With the revised conversion factors these figures increased to 64% and 33% respectively.

In line with what was found in relation to weekly drinking, the impact of changing the method of calculating units is greater for women's daily drinking estimates than for men's. Once again this is related to the different kinds of alcohol women drink compared to men. The following table presents the proportions who had consumed wine or normal beer on their heaviest drinking day to illustrate the way in which alcohol consumption varies by sex. Across all age groups, women were more likely than men to have drunk wine on their heaviest drinking day in the past week. In addition, women aged between 25 and 64 were more likely than other women to have drunk wine. These patterns correspond with the findings in relation to the overall impact of the new conversion factors on women's daily drinking levels. As Figure 1K illustrates, in relation to the proportion who drank more than 3 units, the impact is greatest for women in each of the age groups 25-34 to 55-64 (with increases of between 21 and 26 percentage points) and is much less pronounced for women in the youngest and oldest age groups (up 9 and 10 points respectively). The impact on binge drinking levels is greatest among those in the age groups 25-34 to 45-54, the proportion in these groups who report drinking more than 6 units a day increases by 17 points using the updated unit calculation method. The corresponding increase for those in the 16-24 and 55-64 age groups is 12 points and it is only 5 points for those aged 65-74. The updated method makes no difference to estimates of binge drinking among women aged 75 and over, the proportion is just 1% with both methods.
Table 1.8A, Figure 1J, Figure 1K

Proportion who had drunk wine and normal beer on their heaviest drinking day in past week, by age and sex

Figure 1J Proportion of men who had drunk in the past week who drank more than 4 units, and more than 8 units, on their heaviest drinking day, by age - original and revised units

Figure 1K Proportion of women who had drunk in the past week who drank more than 3 units, and more than 6 units, on their heaviest drinking day, by age - original and revised units

Daily alcohol consumption by socio-demographic factors

The estimates based on the previous unit conversion factors, presented in Table 1.10A, show there is a relationship between NS-SEC and daily drinking among men. Those in professional and managerial households and those headed by small employers are less likely to have drunk more than 4 or more than 8 units than men in other types of household. The impact of the revised conversion factors does not vary greatly across the NS-SEC groups, for example the proportion of men drinking more than 8 units a day increases by between 7 and 11 percentage points in each group, with the smallest impact being for men in semi-routine and routine households and those headed by small employers. As a consequence the overall pattern between NS-SEC and men's daily drinking is fairly similar regardless of which conversion factors are used.

In contrast, whereas there is no consistent pattern between NS-SEC and women's daily drinking based on the previous method to calculate units, the impact of the new conversion factors is greatest overall for women in professional and managerial and intermediate households. For example, the proportion of women in these two groups who drank more than 6 units on their heaviest drinking day increased by 17 percentage points compared with just 7 for women in semi-routine and routine households. The increase in mean units per day was also highest for these two groups. The only slight deviation from this pattern was in relation to the increase in the proportion exceeding 3 units where the percentage point increase was highest for women in professional and managerial households,
24 points, as per the pattern described above, followed by those in small employer households (22 points). As a consequence there is a more consistent relationship between NS-SEC and women's daily drinking than revealed by the previous conversion method; women in managerial and professional and intermediate households are the most likely to have drunk more than 3 or 6 units on their heaviest drinking day.
Table 1.10A

The impact of the new conversion factors on the relationship between household income and daily drinking is slightly different to that found with NS-SEC. Using the previous method, daily drinking does not appear to be consistently related to household income for either men or women. For example, mean units and binge drinking do not follow a clear pattern although the proportions of men and women exceeding 4 or 3 units respectively are highest in the highest and lowest quintiles. However, the impact of the new conversion factors is greatest in the highest income quintile for both sexes which results in a clearer pattern emerging (as described above in section 1.4.3). Between the first and fourth quintiles drinking declines as income declines before increasing again in the fifth quintile.
Table 1.11A

A further difference emerges when looking at the impact of the new conversion factors on the relationship between SIMD and daily drinking. Based on the previous estimates, men in the two most deprived SIMD quintiles are the most likely to drink more than 4 or 8 units a day and to consume the highest number of mean units on their heaviest drinking day. As the impact of the revised conversion factors is greatest for men in the least deprived quintile the pattern is now slightly less pronounced, but still remains. The impact of the updated method for calculating units is also greatest for women in the least deprived quintile.
For example, 17% of women in the least deprived areas drank more than 6 units on their heaviest drinking day using the previous method compared with 37% with the updated method. As a result, there is no consistent pattern between women's binge drinking and area deprivation when the new conversion factors are applied; this is not to say that no relationship exists, rather that its nature is somewhat complex (as described in Section 1.4.3).
Table 1.12A, Figure 1L

Figure 1L Binge drinking among adults who had drunk alcohol in the past week (observed), by Scottish Index of Multiple Deprivation quintile and sex - original and revised units

1.5 PROBLEM DRINKING AND PHYSICAL DEPENDENCY ON ALCOHOL

This section reports various drinking behaviours, rather than alcohol consumption, therefore it is unaffected by the revised conversion factors used in the previous sections.

1.5.1 Problem drinking by age and sex

Adults who were classified as current drinkers in the main interview (i.e., excluding those who said they never drank nowadays) were asked to fill in the CAGE problem drinking self-completion questionnaire. This contains a series of eight statements about drinking and informants were asked to judge whether each applied to them over the course of the last three months. Six of the items are indicators of problem drinking, three of which indicate physical dependency. Table 1.14 presents the proportion of current drinkers who agreed with each statement.

The most common problem drinking indicator agreed with was "I have felt that I ought to cut down on my drinking", which 23% of men and 16% of women agreed with. The next most common indicators were "I have felt ashamed or guilty about by drinking" (7% men and 5% women), and "People have annoyed me by criticising my drinking" (8% men and 3% women). Physical dependency indicators were less commonly reported: 6% of men and 3% of women found that their hands were shaking in the morning after drinking the previous night. 5% of men and 2% of women had had occasions when they felt unable to stop drinking, and 5% of men and 1% of women had had a drink first thing in the morning to steady their nerves or get rid of a hangover.

The prevalence of agreement with each of the six items tended to be lowest in the two oldest age categories (although this was not the case for every single item within both men and women).

Agreement with two or more of the six CAGE items is a possible indicator of 'problem drinking'. Overall, this was more prevalent among men (12%) than women (7%). The proportion agreeing with two or more items declined with age for women, from 10% in the youngest age category to 3% in the oldest. This relationship was also found for men, except that the decline started after the second youngest age category (25-34), from 19% for this group to 4% for men aged 75 and over.

The proportion of men aged 16-74 agreeing with two or more CAGE items was 12% in 1998 and 13% in 2003, suggesting little change over time; among women aged 16-74, the proportion increased slightly from 5% to 7%.
Tables 1.14, 1.15

1.5.2 Drunkenness by age and sex

Adults who were current drinkers were also asked if they had, on average over the last three months, been drunk once a week or more. Those who had not were asked if they had been drunk at all within the last 3 months. Table 1.14 shows the results.

Among current drinkers, men were more likely than women to have been drunk in the past 3 months (52% of men and 39% of women). The prevalence of drunkenness was highest among the youngest age group for women (69%), and it then decreased within each age category so that by the oldest group it was only 4%. Among men, it was the 25-34 age group who were most likely to say they had been drunk in the last 3 months (79%), and this proportion then decreased with age to 8% of men aged 75 and over. It was the youngest age group among both men and women who were the most likely to say they had been drunk at least once a week in the last three months: 40% of men and 28% of women aged 16-24 said this; the proportion reporting this then decreased with age for both sexes.
Table 1.14

1.6 SOCIAL CONTEXT OF DRINKING

Questions about the social context of drinking were included in the 2003 survey for the first time. The questions asked current drinkers where they usually were, and who they were usually with, when they drank alcohol. The locations presented to informants included: pub/bar; restaurant; club/disco; party with friends; at home; at someone else's home; a street/park/other outdoor area. Potential drinking companions included: boyfriend/ girlfriend/ partner/spouse; friends of the same sex; friends of the opposite sex; group of friends of both sexes; family/relatives; on own. Informants could give as many locations and companions from these lists as they wished.

1.6.1 Usual places to drink

The most common drinking location mentioned by current drinkers was in their own home (71% of men and 72% of women). There were some notable differences by sex. Women were more likely than men to report drinking in someone else's home (35% and 29% respectively), and were less likely to drink in a pub or bar (35% and 47% respectively). 30% of women reported drinking in restaurants compared with 22% of men.

There were significant generational differences in drinking locations. The most common places to drink mentioned by those aged 16-24 were a pub or bar (68% of men and 70% of women), and a club or disco (43% and 51% respectively). Drinking in these places decreased significantly from the age of 25 onwards for both sexes so that by the age of 75 and over these two locations were the least popular for drinking in (aside from drinking outdoors): among the 75+, only 14% of men and 4% of women said they drank in a pub/bar, and 10% of men and 5% of women drank in a club/disco. Drinking at home was the most commonly mentioned location for the oldest age category (81% for men and 68% for women).

Very few drinkers of any age mentioned drinking in a street, park or other outdoor area (mentioned by just 1% of men and women).
Table 1.16

1.6.2 Usual drinking company

Drinking companions were similar for both sexes, and most often included partner/spouse (56% of both men and women), other family/relatives (40% men, 47% women), and groups of friends (43% men, 40% women). 11% of men and 8% of women said they usually drank on their own.

As would be expected, drinking companions change with age. Among both sexes in the youngest age category, drinking companions were most likely to be groups of friends of both sexes (76% men, 67% women). However, women in this age category were much more likely than men to report drinking with a partner (48% and 28% respectively). By the 25-34 age category, drinking with a partner was the most common response for both sexes (61% men, 67% women), with groups of friends still being mentioned by over half of male and female informants. Drinking with groups of friends, however, declined with age, so that by the oldest age group it was mentioned by only 18% of men and 19% of women. Aside from the youngest age group, drinking with a partner was the response given most often by men, although it declined for the two oldest age groups. It was also the response most often mentioned by women up to the 55-64 age group, when it declined to 41%, and drinking with family/relatives was more commonly mentioned (55%). Women in the oldest age group were far more likely to drink with family/relatives (60%) than with their partner (19%).
Table 1.17

1.7 CHILDREN'S ALCOHOL CONSUMPTION

This section presents updated figures for alcohol consumption amongst children aged
13-15. The report based on the 2003 survey contained separate volumes for adult and child data and contained a chapter covering children's alcohol consumption. ¹⁸ As only one table from that chapter has been updated using the revised unit conversion factors it is being presented here rather than in a separate publication.

1.7.1 Data collection methods

The 2003 Scottish Health Survey used a self-completion booklet for children aged 13-15 so that questions about alcohol consumption and smoking behaviour could be asked of this age group without their parents/guardians hearing their responses. The questions followed a similar format to that used for those aged 16 and over by asking about the types and volumes of alcohol consumed in the previous week. Historically, information about children's drinking behaviour has been collected via the Scottish Health Survey and a schools based survey of children aged 12 and 15 (the Scottish Schools Adolescent Lifestyle and Substance Use Survey - SALSUS). Data collected in the home about children's drinking behaviour is likely to be prone to under-reporting; despite the self-completion format children may nevertheless have concerns about revealing behaviours that might meet with parental disapproval. It is often thought that the over-reporting that school based surveys can be prone to is of a lesser scale than the under-reporting in home based surveys. The 2003 SHeS consumption estimates for children aged 13-15 are certainly lower than those ascertained in the 2002 and 2004 SALSUS surveys. ¹⁸ The other key advantage that SALSUS has over SHeS is its much larger sample size; as a consequence the trend data based on SALSUS is more robust than that available via SHeS. Although there are advantages to measuring alcohol consumption in a survey such as SHeS, for example the ability to investigate health behaviours within families and across a wider range of other topics within the survey, from 2008 onwards it has been decided that data on children's alcohol consumption will be collected solely via SALSUS.

1.7.2 Revisions to 2003 data

Table 1.18 presents the estimates of children's alcohol consumption based on the original and revised conversion factors. The methods used to revise the units are identical to those set out in Section 1.2.2. However, additional changes were also made to the way in which the units are presented ¹⁹ and the method for handling cases in which children reported having consumed alcohol but did not state the associated volume. The data presented in the previous report used imputation to estimate the volumes for children with missing data. Although it only affected a small number of cases this method is not wholly optimal. Instead the new table presents these children as having consumed an unknown quantity of alcohol.

1.7.3 Weekly consumption by age and sex

As shown in Table 1.18, most children had drunk either no alcohol, or less than one unit, in the preceding week. Boys and girls were equally likely to have drunk one or more units of alcohol in the last week, 12% and 10% respectively. The proportion of children who had drunk alcohol in the past week increased with age from 5% of boys and 5% of girls aged 13 to 19% of boys and 20% of girls aged 15.

Boys aged 13-15 reported drinking 1.0 mean units of alcohol in the past week, the corresponding figure for girls was 1.3 units. The amount of alcohol children reported drinking also increased with age. Among children aged 13, mean consumption in the last week was 0.1 units for boys and 0.2 units for girls. This rose to 2.2 units in the last week for boys aged 15 and 3.7 units for girls aged 15.

Among all children who said they had drunk alcohol in the last week, mean consumption was 10.7 units for boys and 15.0 units for girls. (Both these figures should be treated with caution since they are each based on fewer than 50 children.) Because of the small numbers of children who had drunk alcohol in the preceding week, it was not possible to compare consumption within age groups.
Table 1.18

1.7.4 Impact of the unit recalculations on children's weekly alcohol consumption

The previous method to calculate units estimated boy's mean units per week to be 0.8, the updated method shows this to be 1.0 units, an increase of 25%. For girls the impact was slightly greater, their mean weekly units increased by 30% from 1.0 to 1.3 units. This reflects the fact that girls were more likely than boys to have consumed wine (4% and 2% respectively) and alcopops (7% and 3% respectively) - the products whose unit estimates have increased the most. In contrast, boys were more likely to have consumed beers, lager, cider or shandy (9% of boys versus 3% of girls). [Data from Table 1.11 in the original chapter that reported the 2003 results.] ¹⁸

« Previous | Contents | Next »