White Cider Consumption and Heavy Drinkers: A Low-Cost Option but an Unknown Price

Aims: To compare characteristics of heavy drinkers who do, or do not, drink white cider during their typical drinking week and to contrast white cider drinkers' behaviour with a similar group recruited in comparable settings 4 years previously. To consider if excessive white cider consumption poses a specific health risk. Methods: Cross-sectional survey of alcohol purchasing and consumption by heavy drinkers consuming white cider in Edinburgh and Glasgow during 2012; comparison of purchasing patterns within Edinburgh in 2008–2009 and 2012. Participants were 639 patients (in- and out-patient settings) with serious health problems linked to alcohol, 345 in Glasgow, 294 in Edinburgh in 2012, and 377 in Edinburgh in 2008–2009. Results: In 2012 white cider consumption was reported by 25% of participants (median consumption (all alcohol) was 249 UK units per week—1 UK unit being 8 g of ethanol). They were more likely to be male and younger. They drank more units of alcohol than non-white cider drinkers and reported more alcohol-related problems. The median price paid for white cider in 2012 was 17 ppu. The period 2008–2012 was associated with decreasing affordability of alcohol, but consumption levels amongst the heaviest drinkers were maintained, associated with an increased proportion of units purchased as white cider. Conclusion: White cider makes an important contribution to the weekly intake of heavy drinkers in Scotland, likely facilitated by low price per unit of alcohol. We suggest these characteristics permit this drink to act as a buffer, supporting the continuation of a heavy drinking pattern when affordability of alcohol falls.

INTRODUCTION

Potential strategies to reduce the health and societal costs of alcohol misuse continue to occupy the minds of governments and those concerned with public health. There is debate amongst health professionals, policy makers and the alcohol drinks industry about the effectiveness of measures to restrict the sale and raise the price of alcohol. The Scottish government has been pro-active. It has instigated policy and legislative change in recent years culminating in the Alcohol (Minimum Pricing) (Scotland) Act 2012 (Scottish Parliament, 2012). This act proposes a minimum unit price for alcohol (one UK unit equalling 8 g of ethanol) of 50 pence (£0.5, US $0.83). However, this proposal is currently subject to legal challenge by drink producers: the implementation date remains uncertain. Nevertheless, evidence has emerged of a recent decrease in the alcohol affordability index in the period 2007–2011 within the UK (Beeston et al., 2013) partly linked to falling disposable incomes and some rises in alcohol prices.

The magnitude of the public health issue facing Scotland is illustrated by alcohol-related mortality rates which in 2011 were almost double those of England and Wales (Beeston et al., 2013). Difference in pricing and uptake of the various alcoholic beverages might contribute, as well as the distribution of drinkers, which in Scotland is weighted towards more heavy consumers (Beeston et al., 2013, p. 39).

Annual reports produced by NHS Health Scotland and the Information Services Division in Scotland (Monitoring and Evaluation Scotland's Alcohol Strategy (MESAS) (Beeston et al., 2013)) integrate data on alcohol retail sales, UK alcohol duty clearances and population survey reports.

While much research literature reports alcohol consumption as one entity, MESAS provides detail on the different types of alcoholic beverage. This reveals important differences between England & Wales and Scotland. Since 2008 per adult annual sales of alcohol in Scotland have been consistently around 20% higher in Scotland than in England & Wales. Spirit sales are key contributors. For example, in 2012, 2.4 times more off-trade vodka was sold ‘cheaply’, i.e. in the 35–49.9 ppu (pence per unit) range in Scotland than in England and Wales. MESAS also reported the cheapness of cider in Scotland: 26% of all sales were at <30 ppu. (For beer the corresponding figure was much lower; 5%, for spirits, 1% and for wine 1% (Beeston et al., 2013)). (Off-trade refers to sales of alcohol in supermarkets, independent shops, off-licences.)

Cider accounts for proportionately little (7%) of the total alcohol sold in Scotland, with most (70%) as off-trade sales (Beeston et al., 2013). It is sold in different formats: traditional apple cider (amber), pear cider (perry) and ‘white cider’ so-called because of its lack of colour. White cider (typical alcohol content by volume (ABV) 7.5%) is usually sold in 2l volumes containing 15 UK units. It has received adverse media attention; its poor image partially ascribed to high ABV and low price (Doward and Pemberton, 2011; Robinson, 2014; Scrimshire, 2014).

That white cider is popular in Scotland amongst heavy drinkers whose health has been harmed was demonstrated by Black et al. (2010). They found that consumption of white cider was reported by 17.5% of such drinkers and was the cheapest drink they purchased (mean unit price of 15 ppu). Anecdotal evidence presented by Goodall (2011) focussing on homeless and street drinkers in London, Leeds and West Sussex, suggested that white cider consumption was linked to greater risk of stomach problems with intake producing intense pain. However, with substantive reports linking white cider directly to physical harm being absent, it may be more correct to apportion any harm to the high dose of alcohol often consumed in a short period of time rather than any attribute or congener of the drink itself. Low price may support this pattern of consumption. Forsyth et al. (2013) concluded from a survey of off-sales outlets in one Scottish city (Glasgow) that those shops which reported the cheapest beverages as best sellers were located in neighbourhoods with higher levels of alcohol-related hospitalizations. He found that all best-selling products below 30 ppu were white cider.

The Aims of this study are to:

1. Describe the consumption and purchasing pattern of contemporary heavy drinkers consuming white cider in two Scottish cities, Edinburgh and Glasgow.

2. Contrast white cider drinkers with heavy drinkers not consuming white cider but recruited simultaneously at the same sites.

3. Consider associations between specific harms and white cider consumption.

4. Compare white cider consumption by heavy drinkers across time, relative to changing affordability.

METHODS

We will present two analyses of white cider consumption data. The first analysis (Aims 1–3) refers to a cross-sectional study conducted in 2012 in Edinburgh and Glasgow, cities which account for one-third of the Scottish population. The second analysis will address Aim 4 and will consider changes in white cider consumption over time during a period of decreasing affordability comparing white cider drinkers recruited in Edinburgh in 2012 with an earlier cohort recruited in 2008.

Patients were recruited from NHS alcohol services outpatient and day patient clinics and from amongst patients admitted to hospital with a diagnosis of a physical or psychiatric alcohol-related illness. Exclusion criteria were: being under 18 years old, unable to understand the questions or give understandable answers in English, evidence of clinically significant memory impairment, e.g. Korsakov's Dementia, being unwilling to be contacted for three further follow-up interviews (this relates to a follow-up study not reported here). In addition, advice from clinicians at each site was taken where patients were unsuitable for inclusion due to separate clinical issues.

Research interviewers administered a questionnaire (Black et al., 2010) which documented the participant's most recent 7 days of drinking using the time line follow-back method (Sobell and Sobell, 1996) or their most typical week. Participants self-reported alcohol consumption and expenditure including the type, volume, brand (when known) of beverage, cost and location of purchase, estimations of time spent drinking and number of drinking days in the week. In addition the questionnaire permitted capturing the reasons for the choice of any drink consumed. The interviews were not time limited and interviewers were able to probe and clarify detail where necessary. Age, gender and postcode were documented, the latter acting as a proxy for socioeconomic status using the Scottish Index of Multiple Deprivation (SIMD) (Scottish Government, 2012). The 2012 SIMD divides Scotland into 6505 small geographical areas called datazones containing ∼350 households identified by postcode. Each datazone is assigned a rank of relative deprivation based on several domains (employment, income, health, education, geographic access to services, crime and housing). We used our participants' postcode to record the SIMD rank by quintile. Participants were also asked to self-complete the Alcohol Related Problems Questionnaire (ARPQ). This is an eleven point questionnaire used to assess severity of alcohol-related problems (Patience et al., 1997). Scores range from zero to eleven with the highest indicating greatest problems. During the interview, participants were asked to self-report any illness or condition associated with their drinking. This was usually partly or wholly connected to presentation at health care services. (Due to ethical constraints, it was not possible to verify self-reported illness with clinical notes.)

Favourable ethical opinion was granted by NHS Lothian Regional Ethics Committee (REC reference 08/S1101/9) and approval was gained from the relevant Caldicott Guardians. Interviews were conducted with 639 patients. In addition, 89 patients identified by clinicians refused to participate prior to receiving detail relating to the study, 61 refused after this point, one refused during the interview and, in 20 cases, the researcher had concerns and terminated the interview. In total, 170 (21%) of those deemed eligible by clinicians, did not participate.

Data were analysed using SPSS version 19. Differences between groups were compared using the t-test and ANOVA, parametric tests, and where required the Mann–Whitney U-test, non-parametric test. The Chi-square tests were employed for categorical variables. Multivariable logistic regression analyses, using the stepwise procedure, were employed to investigate the associations of various factors (gender, age, SIMD, cigarette consumption, self-reported haematemesis, units consumed, time, ARPQ score, and city), on white cider drinking or not. The final model presents the odds ratio (OR) and the 95% confidence interval for those remaining significant factors.

Bonferroni corrections were applied to adjust for multiple comparisons. An alpha value of 0.05, two-sided, was considered significant.

RESULTS

Analysis of 2012 data

In the 639 patients (Edinburgh, n = 294, Glasgow, n = 345) 161 (25%) reported consuming any white cider in their most recent or ‘typical’ week of drinking. Within this latter group 72 participants drank white cider exclusively (WCE), with 89 drinking white cider in addition to other drink types (AWC). A total of 478 participants consumed no white cider (NWC).

Consumption

Expressed as a proportion of total consumption for the entire sample, white cider drinkers (any) represented 25.2% of participants yet consumed 33.0% of the total units.

Aspects of the recorded consumption of these three groups of drinkers are summarized in Tables 1 and 2.

For the entire sample (n = 639), the median unit consumption was 184.8 UK units (IQR = 161.3). Those drinking any white cider (the combined AWC and WCE groups) drank significantly more alcohol (median = 249.38 UK units, (IQR = 207.9)) than the NWC group (median = 173.6 UK units (IQR = 162.9)) (P < 0.001) with the AWC group drinking more units than the WCE group (P = 0.04).

Gender

Amongst all participants (n = 639), there was a greater proportion of men than women, men outnumbering women almost 3:1. Men consumed significantly more alcohol than women: median unit consumption (UK units) for men was 196.0 (IQR = 164.9) and for women was 157.6 (IQR = 159.8) (P < 0.001). However, while this gender difference was also found within the NWC group; men (median = 182.0 units (IQR 148.6)) and women (median = 137.8 units (IQR 138.1)) (P = 0.001), it was not evident within AWC or WCE groups.

Age

An independent sample t-test revealed a significant difference in age between the combined white cider groups (M = 40.3, SD = 9.0) and NWC drinkers (M = 47.4, SD = 10.7), (P < 0.001). The WCE and AWC groups did not differ with respect to mean age.

ARPQ

Participants were invited to self-complete the Alcohol Related Problems Questionnaires (ARPQ) (a higher score indicating greater problems). The median score for the combined groups of white cider drinkers was higher than that of the NWC group (P < 0.001). The median score for the AWC group was also higher than that of the exclusive WCE drinkers (P = 0.009).

Smoking

Smoking was defined in terms of tobacco use and equivalencies were applied for those who smoked tobacco in the form of roll-ups rather than filter cigarettes (NHS Greater Glasgow and Clyde, 2013). Of the total sample 70% (n = 447) reported tobacco use, smoking the equivalent of a mean of 18.6 (SD = 11.5) cigarettes per day (see Table 2). Within the NWC group 66% of drinkers were also smokers. Within each of the white cider groups proportionately more people reported smoking, 83% in the AWC group and 81% in the WCE group. When estimates were made of tobacco smoked in roll-ups, there was no significant difference between the three groups in relation to the number of cigarettes smoked per day (P = 0.102).

Place of purchase

All white cider was purchased exclusively at off-sales outlets: 77.25% of white cider units were purchased at independent licenced grocers, 12.7% from supermarkets and 10.0% from other off-licences (drinks retailers, garages and newsagents).

Expenditure

Table 3 shows the weekly expenditure and median unit price paid by NWC drinkers, AWC drinkers and WCE drinkers. For white cider the median unit price paid was 17.0 pence (IQR 2.0); the lowest price paid was 10 pence per unit. During the 2012 data collection phase reported here (9 months) the median price paid by participants per unit of white cider changed from 13 to 18 ppu in Edinburgh (16 to 17 ppu in Glasgow).

Participants' perceptions of white cider

Participants were asked to state their reasons for purchasing white cider (all responded). The majority reported that it was chosen because of its cheapness (82.5%). One participant advised that it was cheaper than heroin while others alluded to its role as a ‘buffer’ being used as a fallback drink when funds were low. Oft-repeated comments linking white cider consumption with an adverse effect on the stomach, prompted testing whether there was an association between white cider consumption and self-reported haematemesis; this was significant (P = 0.008).

Comparison of non-white cider drinkers and ‘any’ white cider consumers

Exploratory logistic regression suggested an association between greater unit intake and white cider consumption (see Table 4). Females had approximately half the odds of males of being in the any white cider group. For every increase of a year in age, the odds of being in the any white cider group were reduced slightly (for an increase of 10 years the odds are halved approximately). No significant associations were found for city, ARPQ score, weekly cigarette consumption, self-report of haematemesis, or SIMD (overall) (although there was a significant effect for the least deprived compared with most deprived, with the least deprived having a fifth of the odds of consuming any white cider).

Comparison of Edinburgh white cider drinkers recruited in 2012 with those recruited in 2008

The characteristics of these two groups of drinkers are summarized in Table 5. In the intervening 4 years there is evidence of a rise in the median price they paid for white cider (from 14 to 17 pence per unit) and, as might be predicted, median expenditure on white cider. Median intake of white cider in these white cider drinkers increased from 157.5 units to 174.4 units; however, this increase was not significant (P = 0.326). (The proportion of women recruits in these two intakes was similar; around one third.)

However amongst the whole samples white cider accounted for 22.67% of all units bought in Edinburgh in 2012 compared with 15.93% in 2008. When comparing the WCE drinkers in each study sample, the total number of participants who fell into this sub category increased from 4.2% (n = 16) in 2008 to 9.2% (n = 27) in 2012.

DISCUSSION

This is one of the first reports to detail drink choices and characteristics of a sub group of heavy drinkers whose purchasing power is directed to the cheapest available drinks on sale. One quarter of the participants in this Scottish sample of heavy drinkers reported consumption of white cider and of those around 45% drank it exclusively. Their median weekly alcohol intake was around 10 times that recommended for males in the UK (DoH, 1995). Overall cider (white and cheap amber varieties) accounted for 83% of all off-sale units sold below 30 ppu. White cider was predominantly purchased at independent licensed grocers at a median price of 17 ppu.

When compared with other heavy drinkers in our sample, white cider drinkers were more likely to be younger, male, and drink more alcohol within the week of study. (However, it is noteworthy that female white cider drinkers did not drink significantly fewer units than males, contrary to the significant gender difference evident in the non-white cider drinkers.) Logistic regression analysis suggests that the higher ARPQ score (indicative of higher alcohol-associated problems) characterizing white cider drinkers likely links to increased alcohol consumption rather than a specific aspect of white cider. We also found, not unsurprisingly, evidence of an association between being a consumer of white cider and living in the areas of highest deprivation. The preference for white cider seems almost certainly influenced by economics, with most of our participants volunteering that they purchased it due to its cheapness, some stating that it is a drink to fall back on when funds are low.

Smoking prevalence amongst the white cider drinkers was around 80% with those participants smoking the equivalent of around six packs of 20 cigarettes per week. (In the total sample around 70% smoked an average of 19 cigarettes per day.) Within the Scottish general population, in 2012, around 25% of adults smoked the equivalent of 13.5 cigarettes per day (http://www.ashscotland.org.uk/ash/4320). The frequent preference for ‘roll-ups' was likely influenced by finances, and creates a risk of error when attempting to compare use of tobacco between groups.

Affordability of alcohol in the UK (calculated from UK consumer price indices and data on real disposable household income) decreased by 4.7% during the time between our two studies (2008–2012). This resulted from a fall in disposable income and increasing off-sales prices. The average price for off-trade sales of alcohol in Scotland in 2008–2009 was 42 ppu, but in 2012 this was 49 ppu, an increase of 7 ppu (Beeston et al., 2013). Our data for white cider suggest an increase of 3 ppu in the same time period. When comparing our two samples recruited at similar NHS settings in Edinburgh four years apart, we find that quantities of units being consumed in this type of population were maintained and that there was an increase (5%) in the proportion consuming white cider exclusively. Additionally, the proportion that white cider contributed to the total units consumed by the sample was 6.7% higher in 2012 than in 2008. Although, the numbers are small and this finding should be interpreted with caution, we suggest that our data are consistent with the self-reports of participants that white cider, due to its low price, provides a buffer permitting the maintenance of heavy drinking patterns when economic circumstances decline.

We have explored the theory suggested from previous work (Gill et al., 2010; Goodall, 2011) that white cider consumption may be linked to increased risk of medical harm. Certainly derogatory comments volunteered by our participants are consistent with Goodall's report, but the results of logistic regression imply that our preliminary finding of an association between consumption of this drink and haematemesis is more likely linked to high alcohol intake rather than white cider specifically. It must also be acceded that drinkers' pre-conceptions and awareness of the negative connotations linked to white cider consumption may have influenced their comments reported to researchers.

Nevertheless we would hypothesize that several inter-related factors may coalesce to increase the risk of alcohol-induced harm for white cider drinkers in the short, and longer, term and are worthy of future investigation. A high dose of alcohol was ingested, arguably facilitated by cheap price. We have previously reported that ‘white’ drinks, including white cider, contain lower levels of antioxidant and cyto-protective compounds compared with amber drinks of equivalent alcohol content (Gill et al., 2010). Acetaldehyde is a mutagenic and carcinogenic chemical present to differing degrees in alcoholic drinks. It is also produced from alcohol by microflora within the mouth (Lachenmeier and Monakhova, 2011). Several factors associated with chronic drinkers may act to exacerbate the production, accumulation and thereby pathological impact of oral acetaldehyde; high alcohol intake, lengthy periods of continual drinking, poor saliva production due to alcohol-induced atrophy of the parotid glands (Salaspuro, 2003), poor oral hygiene, poor nutritional status and, as reported by Salaspuro (2007), smoking. A greater proportion of our sample of white cider drinkers smoked than non-white cider drinkers and tobacco is well known to interact with alcohol to produce aerodigestive cancer. The acetaldehyde content of various alcoholic drinks has been reported by Lachenmeier and Sohnius (2008). In terms of standard drinks, highest acetaldehyde content was noted for sherry, apple wine/cider (and some South American spirits). Taking these points together with the availability of particularly cheap cider in Scotland, its demonstrated contribution to the consumption pattern of ill, heavy drinkers in our cities leads us to suspect that the higher rate of alcohol-related mortality in Scotland (mainly liver and cancer deaths) might be in part accounted for.

Several limitations to our work suggest the need for caution in the interpretation of our findings. Their generalisability cannot be assured: the unknown number of ill, heavy drinkers not seen by services; those attending the services too ill to be interviewed and/or not referred by staff; patients admitted, but discharged, over the weekend and some declined to participate. All would be excluded.

All data were self-reported. We cannot be certain about accuracy of recall and honesty of reporting, although one of the strengths of our study is that the interview took as long as required by each individual participant. We took time to explore any ambiguities in recall and achieved a low rate of missing data. Where drink unit content or price was unclear, manufacturers' and supermarket websites were checked. In so doing we have contributed to the call by the MESAS authors (Robinson et al., 2013) for ‘A better understanding of the relationship between beverage-specific alcohol consumption and alcohol-related harm in Great Britain, as well as the role of drink preference across different types of drinker’ (page 15).

In conclusion, the comments of Brown (2014) are pertinent; she lamented the fact that the ban on ‘below cost sales' (for England & Wales) introduced in April 2014 may ironically permit cheap drinks to become even cheaper. She argues that while this new legislation prohibits retailers selling alcoholic drinks for less than the total sum of duty and value added tax, the favourable alcohol tax rates for cider could permit strong white cider to be sold for as little as 6 ppu. On the other hand, implementing a minimum price per unit of alcohol at 40 ppu or greater would eliminate this cheap choice. What is not known, however, is what would be the fallback response of the drinkers such as we studied.

Funding

This study was supported financially by the Chief Scientist Office, Scotland; Alcohol Research UK; NHS Health Scotland; NHS Lothian Foundation Trust and in kind by the Scottish Mental Health Research Network.

Conflict of interest statement

Professor Chick is medical director, Castle Craig Hospital, Scotland; Advisor: Alcoholics Anonymous UK, H. Lundbeck A/S, Drinkaware Trust, Institute for Alcohol Studies.

Acknowledgements

We gratefully acknowledged the assistance of NHS staff and patients in the various hospital settings.

REFERENCES