Alcohol Drinking Patterns and Work Areas: epidemiological study of factory and rural workers in Florence, Italy

British Journal of Addiction (1988) 83, 1169-1178

Alcohol Drinking Patterns and Work Areas:epidemiological study of factory and ruralworkers in Florence, Italy

ALLAMAN ALLAMANI, FRANCESCO CIPRIANI,STEFANO INNOCENTI, CANIO LOMUTO, MARCO MARCHI& ANTONIO MORETTINI

Divisione di Gastroenterologia, Ospedale di Careggi, USL 10/D, 50100 Firenze, Italy

SummaryTwo epidemiological studies have been recently carried out in Florence and the surrounding area in Italy, totest a hypothesis that work conditions affect alcohol drinking patterns. This hypothesis is based both onavailable health statistics and on a previous epidemiological study carried out in the same area in the late197O's. Two hundred and fourteen factory employees and 184 agricultural workers were interviewed using aquestionnaire that addresses sociodemographic characteristics, psychological and social aspects, and alcoholdrinking patterns. The main results include: average per capita per day anhydrous alcohol consumption is34.9 g among factory employees, and 87.9 among agricultural workers. Wine is the most common alcoholicbeverage, and it is drunk essentially together with other people. In the factory, workmen drink significantlymore than clerks. There is a drinking culture that is more rooted in the country than in urban areas. No othersignificant relationship was found between drinking pattern and work features.

IntroductionIt is a growing opinion among health professionalsin Italy that a significant percentage of the popula-tion should be considered at risk for excessivealcohol consumption and that alcohol-related bio-logical and psycho-social problems are rising. In-deed, Italy often scores as one of the top countriesboth in alcohol consumption and in alcohol pro-duction. According to national data of the CentralInstitute of Statistics (ISTAT), in 1981 Italyproduced 70,580,000 hi of wine, which represented25% of total world production. When this figure iscompared with the annual wine production figuressince 1950, a slight increase is noted (Modonutti et

Address for correspondence and offprints: Dr AUaman AUamani,Via P. Toselli, 140, 50144 Firenze, Italy.

al., 1984). It should be pointed out that data derivedfrom the producers' official statements are onlyapproximates of the real situation.

On the other hand, the overall consumption ofalcohol in the years following the second World Warhas notably increased during the decade 1960-1970,having reached 16.08 litres of anhydrous alcohol percapita per year in 1970 (Modonutti et al., 1984). Forthe following years, data show a slight drop to 14.11litres in 1981. This decrease is noticeable for wine,which in keeping with Italian drinking patterns(Cottino & Morgan, 1985), remains the mostpopular drink by far and a central part of Italiandietary culture. When alcohol consumption is consi-dered by region through the years, Friuli-VeneziaGiulia, Veneto, Val d'Aosta, Marche and Tuscanyscore highest.

As characteristics of wine-producing countries.

1169

1170 AUaman AUamani et al.

wine is the alcohol drunk most, usually duringmeals. However, due to the so-called 'internationali-zation of drinking habits' and to the changingbehavior of contemporary youth, other types ofalcoholic beverages and drinking patterns arespreading in recent years (Walsh, 1982). A commonsource for the above-mentioned data is a periodicsurvey carried out by ISTAT on a sample offamilies.

With respect to a traditional index of alcoholism,i.e. hepatic cirrhosis mortality, the ISTAT statisticsfrom medical death certificates show an increasebetween 1955 (when there were 13.7 deaths per100,000 population) and 1977 (33.8 deaths per100,000) (Rausa et al, 1984). The contribution ofalcoholism to such deaths seems to vary from 30% to70%, depending on the region. When cirrhosis deathrate is compared with alcohol consumption invarious Italian regions, we find, as may be expected,that the regions with the greatest consumption havea higher death rate. From this point of view,Tuscany's death rate from cirrhosis differs from thenational tendency because, while among the groupswith the highest amount of alcohol drinking, thedeath rate is below average (Morettini et al., 1982).

Notwithstanding the increasing attention to alco-hol and alcoholism beginning in Italy during the197O's (RoUi, 1985), accurate epidemiological in-formation was and remains lacking. Data on pro-duction and consumption, either from public orprivate agencies, are often contradictory or not evenavailable. For example, there is a remarkableabsence of data collected by ISTAT since 1973 ondrinking liquor and on regional distribution of beerand liquor consumption. The same Institute pro-vides information which sometimes differs fromthat derived from other sources, e.g. the MinistryDepartment of Agriculture or the National Institutefor Protection of Italian Brandy (Morettini et al.,1982). Similar problems exist for the data onalcohol-related problems.

It is thus necessary to have a clearer descriptionof the phenomena. It is obvious that studies of thehospital population (e.g. Iandolo & Capuano, 1968a;Iandolo & Capuano, 1968b; Pedrazzoli et al., 1974;Saraceni, 1975; Aimone et al., 1986) do notaccurately represent the general population.

An Alcohol Study of the General Population inFlorence (1977)It was with these concerns in mind that particularlyin the last years of the 197O's a few epidemiological

research groups in different areas in Italy began topay attention to alcohol consumption in theircommunities. The first of these studies—in whichtwo of the authors of this article participated (A.M.and A.A.)—was undertaken in 1977. It consisted ofa sample of 401 adult subjects equally dividedaccording to sex, age (20-30 and 45-55 years old)and place of residence, and randomly chosen fromthe general population in four areas in the provinceof Florence, Tuscany. This province has approxi-mately 1,200,000 inhabitants (430,000 in the city ofFlorence). The purpose of the study was to identifyboth the quantity and the patterns of alcoholconsumption and to determine the correlationbetween alcohol and liver damage. Subjects wereinterviewed using a psychological and alcohologicalquestionnaire and were submitted to a medicalexamination, including a blood test.

The survey found (Table 1) that wine was themost common drink, the consumption being 37.3 gof anhydrous alcohol per capita per day, followed ata considerable distance by liquor (3.0 g) and beer(0.9 g). While drinking beer and bitters waspositively associated with the younger group, noneof the other alcoholic beverages were correlatedwith age. Thirteen percent of the subjects had ananhydrous alcohol daily intake of more than 80 g.The members of these latter subgroups werepredominantly adult, male, married, with largefamilies, residents in the country or the province ofFlorence, and had few years of education. Theywere either employed as workmen or workedindependently in agricultural or industrial activities(Sirigatti, 1982; Morettini, AUamani, Lomuto,1982).

Table 1. Average Anhydrous Alcohol Intake (in grams) perCapita per Day, According to the Type of Drink, in 1977(general population), in 1983 (factory employees) and in

1984 (agricultural workers) in Florence area

1977Year1983 1984

WineBeerLiquors

37.30.93.0

28.91.13.2

80.81.55.5

Rather than being psychologically motivated, theuse and abuse of alcoholic beverages appeared torefiect a response to environmental needs as well associal custom. Wine was drunk with daily meals, athome and with the family. Some variation was foundin traditional drinking patterns, as for instance the

Alcohol Drinking Patterns and Work Areas 1171

increased consumption of beer by the younger agegroup, which possibly reflected the impact ofcultural changes occurring in family and society.These findings have been partly confirmed by othernational studies (Raimondi et al., 1987).

Liver damage (indicated by at least elevations ofserum transaminases) showed up in 10.2% of thepeople. They mainly were men of a lower educa-tional level working in agriculture or industry. Liverdamage was found in 18% of those with a dailyintake of anhydrous alcohol of more than 60 g. Apositive association was found between alcoholintake and liver pathology, even though the twophenomena could exist independently (Bozza, 1982;Morettini et al., 1982).

Research on Alcohol Consumption Patterns inWork Areas in FlorenceThe aforementioned study raised a number ofquestions regarding the relationship between workarea and drinking pattern. The researchers consi-dered the hypotheses that work stress, requiredwork skills, and work culture would affect con-sumption patterns. A further interest was in evalu-ating changes in Italian drinking patterns throughthe years. To these ends a second series of studies(1983-1987) was planned by the Division Gastro-enterology of Careggi Hospital, Florence, which isnow nearing completion. The studies' researcherscome from Epidemiology, Medicine, Psychiatry,Health Education, and Information Systems.

MethodologyThis research focuses on four work areas, eachconsidered a social system. We assume that indivi-dual participation and social changes are greater insituations defined as systems, where components arein some mutual interaction, than among the generalpopulation. Thus among the work areas identified asbest fitting such characteristics, the following werechosen: (1) a medium-sized factory; (2) an agricul-tural district; (3) a group of elementary and highschools; and (4) an area in the town wherecraftsmen mainly reside. Within each system sampl-ing is taken according to work sector. A similarprocedure is used in the four areas to include theparticipation of a few of the system's key personnelwho, together with the research personnel, agree toadapt the basic questionnaire on alcohol to the needsand characteristics of the system.

With minor variations the questionnaire is thesame one used in the 1977 survey. It is divided intothree parts: (1) sociodemographical; (2) alcohologi-cal—questions on the interviewee's opinions aboutthe amount, the frequency, the length (years) ofdrinking, and the drinking changes in the last 3years for the more common alcoholic beverages;(3) and psychological/cultural—questions on theinterviewee's perceptions of drinking.

A list of questions on the subjects' eating habits isincluded in the questionnaire, in keeping with theexpectation of a strong connection between eatingand drinking alcohol in Italy.

Table 2. Values Attributed to Drinking Frequencies to beused to Compute Average Alcohol Daily Intake According to

Formula (1)

Drinking frequency Value

Never or less than once per month 0Once per month 1Once every 15 days 2Once per week 4Two or three times per week 12Every day 30

Table 3. Alcohol Degree Values Attributed to DifferentTypes of Drinks, used to computer average alcohol daily

intake according to formula (1)

Type of drink

WineBeer'Vin Santo'AperitifBitters'Dry' Liquors'Sweet' Liquors

Alcohol degree

124

1620344042

The quantity of alcohol consumption during anaverage day is inferred by showing each intervieweethree glasses of sizes commonly used—respectively250, 150 and 35 cm', and recording both the size ofthe glass used (G), and the number of glasses drunk(nG). Also, the frequency of consumption (F) ofeach kind of beverage is considered (see Table 2).The amount of alcohol consumed for each type ofbeverage is translated into grams of anhydrousalcohol according to the degrees of alcohol (,AD)shown in Table 3. Further, in computing volume

1172 Allaman AUamani et al.

Table 4. Perception of Alcohol Drinks (Wine, Beer, Spirits)According to their Attributes

Whets appetiteFacilitates social relationshipsHelps digestionCheers upIs gentlemanlikeQuenches thirstRelaxesMakes companyIs for connoisseursMakes for spontaneityPrice is accessibleTastes good/is good for the healthReassuresIs harmful

into weight, it should be remembered that thespecific weight of alcohol is 0.8. On these premises,the quantity of alcohol daily consumption (Q) iscalculated as follows:

ADxFxnGxO.S

30(1)

In the 1977 survey and in the present studies aquestionnaire on perceptions of various kinds ofdrinks according to 14 attributes is administered tothe interviewees (Table 4). The importance of eachattribute is determined by the score (from 0 to 9) itreceives. Thus each alcoholic beverage is defined bya series of attributes, classified along a scaleaccording to their average importance.

Interviews are preceded by a general assembly towhich all workers are called, and are carried out onthe selected samples by specially prepared inter-viewers. Following data collection and data process-ing, the results are presented to a final assembly.

Research AimsThe general aim of this study is to evaluate alcoholconsumption patterns of workers in the followingwork areas: industry, agriculture, schools, andcrafts. According to the hypotheses that stress, workskills, and culture are factors affecting alcoholdrinking, the study also attempts to correlate thework components within each system to alcoholdrinking patterns.

Another aim is to simultaneously initiate aprocess of health education by involving all theworkers in the work area, on the assumption that theeffect of a research on alcohol operating in a worksystem would possibly be to moderate alcohol intake

both during the study and after communication ofits findings. The effect of this education is to beassessed in the following years.

We report here on the first two areas: industryand agriculture.

Alcohol Consumption Pattern Among FactoryEmployeesCharacteristics of the FactoryThe 'Officine Galileo' factory was chosen as theobject of our study because historically, socially, andeconomically it is representative of medium-sizedfactories in the province of Florence. We must notethat this province is characterized by numeroussmall or family-based plants, some medium-sizedones, and very few large ones. The Officine Galileofactory is situated in the northern industrial out-skirts of Florence and has 1,350 employees. At onetime it was known for its optical products. Today itsproducts are several and there are three work sectors:mechanical, optical, and electronic.

In addition, there are a number of work environ-ments which have been differentiated according tothe following criteria: the presence/absence ofsteam, powder, thermic excursion, noise, and thephysical closeness/distance to other employees (as-sumed to correlate to the social pressure on thecontrol of drinking).

The employee work categories are basically two:workmen (blue collar) and clerks (white collar).

Interviewing and SamplingAfter subdivision and numbering of personnel, arandom sample was taken that was proportional tosectors, environments and categories. A total of 214adults (106 workmen, 108 clerks) between 21 and60 years of age (average age=39.4) were inter-viewed. Of these, 85% were men and 15% werewomen; 81.8% were married, 68.1% were bom inFlorence or the province of Florence whereas 86.9%were born in the region of Tuscany. Only 11.8%came from other Italian regions while four inter-viewees were born abroad. In the sample, 24.6% hadelementary education; 27.5% had junior high schooland 44.1% had senior high school education; 3.3%had university education while 0.5% were illiterate.

The research personnel, in co-operation withthose responsible for the factory council, preparedthe questionnaire. Before starting the survey, theproject co-ordinator held a general assembly confer-ence in the factory with participation of most of the

Alcohol Drinking Patterns and Work Areas 1173

employees in order to explain the general aims of theproject and to solicit the involvement of employees.

The survey took place between January 1, 1983and March 12, 1983. The non-respondents wereseven (3.3%).

HypothesesThe study hypotheses were based on the followingcultural assumptions about alcohol consumption:

(i) alcohol was assumed to be consumed on thejob to a relatively high degree—mostly duringfactory afternoon meals—in keeping with knownTuscan and Italian tradition.

(ii) alcohol was assumed to be drunk more: (1) inwork sectors requiring less precision (routine me-chanical skills); (2) in work sectors more exposed tosteam, powder, thermic excursion (thus creatingmore thirst); (3) in noisier work sectors (creatingmore psychological tension); (4) among workerswho work in relative physical/social isolation wherethere would be less social control on drinking.

(iii) in keeping with the above-mentioned issues,and with our previous findings, workmen wereexpected to consume more alcohol than clerks.

ResultsAlcohol consumption on the job is lower thanexpected, and in fact only 23.8% of the interviewedsubjects (51 people) indicated that they consumedalcohol at work—mainly wine at meals. The averageconsumption per person per day is 2.3 g ofanhydrous alcohol (SD=6.2), which rises to 10.0 gif only drinkers are considered. Thus the assump-

tion that many workers drink at the afternoon mealin the factory was not supported.

Regarding the total daily consumption of alcohol(both in and out of the work setting), the averageconsumption per person per day is 34.9 g ofanhydrous alcohol (SD = 28.4), which rises to 36.4 gif only drinkers are considered (Table 5).

It is the subjects' opinion that this amount hasremained essentially the same during the last 3years.

Non-drinkers constitute 4.3% of the sample,while consumers of 60 g and more are 15.3%.

Wine is by far the most drunk, followed by liquorand beer (see Table 1). Also, wine is the firstalcoholic beverage drunk during the life cycle, inkeeping with its prominence in Italian drinkingtraditions. The average beginning age—the subject'sage less the reported length of time he/she has beendrinking—is 17.8 years old (SD = 7.4) for wine,while it is around 20 years old for all other types ofdrinks.

Regarding interactional patterns, wine is mosttypically drunk in the family context, while beer andspirits are especially drunk with friends, outside ofthe home. Beer is sometimes consumed alone.

As far as the relationship between alcohol con-sumption and demographical characteristics is con-cerned, our results show the following: (1) Womendrink less than men (average grams of anhydrousalcohol per person per day are 14.7 among womenand 38.5 among men (p<0.001). (2) Age, civilstatus, education are not correlated with meaningfuldifferences in the amount of drinking. (3) Inconsidering the distribution of drinking according tothe place of residence, we find that the residents ofFlorence and its outskirts (Fiesole, Sesto Fioren-

Table 5. Prequency Distribution of Galileo Pactory Employees (Plorence,1983) According to six classes of anhydrous alcohol daily intake on the job and

during the whole day

Anhydrous alcoholintake

Intake(g per day)

00.1-40

40.1-6060.1-8080,1-120>120

TOTAL

On

N

16350

1———

214

the job

%

76.223,30,5———

100.0

During thewhole day

N

91264614172

214

%

4,358,921.5

6.57.90,9

100.0

1174 Allaman AUamani et al.

tino, Scandicci) mostly fall into the category ofminor drinkers. This confirms what others havenoted (see also the 1977 survey in Florence)—thatdrinking is more rooted in the culture of non-urbanresidents than in that of city-dwellers (Fig. 1).

Periferical Area D

0 0.1-20 20.1-40 40.1-60 60.1-80 80.1-120 > 120

Intake Igl

Figure 1. Prequency distribution of Galileo factory employ-ees (Florence, 1983) according to residence and to 7 classes

of anhydrous alcohol intake (g per day).

Considering the study hypotheses our results arethe following:

(i) Workmen show a significantly higher con-sumption than clerks (Fig. 2).

(ii) No significant correlation is found betweenalcohol consumption and employees' specific worksectors, including those who worked in settings highin steam, who on first observation, seemed to beamong the class of heavy drinkers.

(iii) A tendency towards an inverse association isfound between drinking water and drinking alcoholamong the workmen and seems to suggest that theneed to drink alcohol is part of the need to consumeliquids in general.

Smokers constitute 51.4% of the interviewedsample. No difference is found when distinguishingwomen from men, workmen from clerks, or differ-ent age groups. Furthermore, while not statisticallysignificant, it is noteworthy that those who drank

= E = ni n= ^ =0 0.1-20 20.1-40 40,1-60 60.1-80 80.1-120 > 120

Intake (g)

Figure 2. Prequency distribution of Galileo factory clerksand workmen (Plorence, 1983) according to 7 classes of

anhydrous alcohol intake (g per day).

more, smoke less. Thus no heavy drinker (i.e.drinking more than 80 g of anhydrous alcohol percapita per day) is a heavy smoker (smoking morethan 20 cigarettes per capita per day); among theheavy smokers, 52% do not drink or drink little (lessthan 20 g) (Fig. 3).

This finding may support the idea that depen-dency is a behavioural condition of the individualthat can be satisfied by an alcoholic substance or byanother, e.g. smoking, which may be interchangeable(Glatt, 1974; Ziegler-DriscoU, 1979).

Qualities Attributed to BeveragesIn exploring what qualities the interviewees attri-buted to different types of beverages, we find thefollowing. Whereas in the 1977 responses wine wasconsidered as 'not harmful' ('harmfulness' scored14th, the last position on the evaluation scale), the1983 responses show a noticeable increase inattributing harmfulness to all alcoholic beverages, aswell as wine to the point that its harmfulness is nowscored as second. This may be related to the effectthat the 1983 research had on the factory workers,even if the researchers were careful to avoid anti-alcoholic crusader attitudes. Other possible explana-tions are either that the attitudes of people changed

Alcohol Drinking Patterns and Work Areas 1175

1-10 11-20 > 20Number of

Cigarenes

Figure 3. Prequency distribution of Galileo factory employ-ees (Plorence, 1983) according to 4 classes of smokingcigarettes (number of cigarettes per day) and to 4 classes of

anhydrous alcohol intake (g per day).

after 6 years; or that this is a particularly health-oriented factory culture. However, the three mostfrequently attributed qualities of wine in the 1977research—'to stimulate cheerfulness', 'to quenchthirst', 'to have an accessible price'—scoring in thefirst four positions in the evaluation scale, receivesimilar top positions 6 years later.

The most frequent ascriptions of 'thirst-quench-ing' quality to beer, of 'helping digestion' to bitters,and of 'whetting appetite' to aperitifs are found bothin 1977 and in 1983.

Finally, comparing ascriptions of 'gentlemanli-ness' and of 'being a question of connoisseurs' towine in both studies we find that they are scoredmuch less by the interviewees in the more recentresearch than 6 years earlier (their position in theevaluation scale being respectively 1 and 6 in 1977and 5 and 14 in 1983). This might be due to aspecific culture of the industrial plant, in contrastwith other work settings.

ConclusionsWithin the limits of our research instruments, wemay conclude the following.

(1) Drinking alcohol does not seem connected tothe physical-chemical variables of the environment.

(2) Alcohol consumption is not correlated withparticular psychological tension or skills in particu-lar work areas.

(3) Alcohol consumption is not associated withpsychosocial variables in the work area.

(4) Alcohol consumption is correlated with dif-ferences in type of work across the different worksectors, that is with blue collar or white collar (theformer drinking alcoholic beverages more than thelatter).

(5) Non-urban residents drink more than city-dwellers.

(6) Alcohol intake may be negatively correlatedwith smoking.

(7) Unexpectedly, wine is considered to beharmful by the majority of those sampled.

Alcohol Consumption Pattern among RuralWorkersCharacteristics of the Rural AreaGreve in Chianti, south of Florence, was chosen forthis study since it represents a typical rural area inTuscany. Its economy is based on cultivating vinesand on producing wine. Indeed 61% of rural Tuscanfirms cultivate vines as their main activity (Cianfer-oni & Pisani, 1987). The population residing inGreve and its surrounding country is 10,500. Whilea few of the workers are craftsmen, most (1^340)work in agriculture. They are predominantly male,working until old age, especially if they own theirown fields. Agricultural work varies along severaldimensions and skills. We distinguished four impor-tant work roles: independent cultivators (coltivatoridiretti); dependent workers (braccianti); tractordrivers; and cellar workers.

Interviewing and SamplingThe basic questionnaire was adapted to the realitiesof the area with the co-operation of the townauthority and trade union representatives; its essen-tial characteristics were maintained however. Theprocess of informing the research participants andthe collection of data were the same as thosedescribed for the Galileo factory.

After subdivision and numbering, a randomsample was taken proportional to the different roles.The interviewed sample consisted of 186 adults (31independent cultivators, 59 dependent workers, 26tractor drivers, 19 cellar workmen; 51 persons

1176 AUaman AUamani et al.

working at two or more of the above types of work),between 24 and 86 years of age (average age, 51.1years). The survey took place from September 4,1983 to April 20, 1984. Non-respondents were 9(4.8%).

The sample was made up of 96.8% men and only3.2% women (a proportion reflecting the fewwomen who work in Greve in Chianti), mostlymarried (86.5%), born within the Florence province(73.2%) or Tuscany region (96.2%). A majoritylived in isolated houses (56.6%), as is to be expectedfrom the social and geographical features of thearea. The general level of education was low with88.2% having elementary school education only;3.2% are illiterate while 8.6% have attended schoolbeyond the elementary level. It is of note that asmall percentage of interviewees (5.5%) had workedin industry before working in rural activities, apattern opposite the more usual movement fromrural to industrial work.

HypothesesIn keeping with traditional rural Italian drinkinghabits, it was expected that wine would be con-sumed on the job often, together with water, forrelief from thirst because of the hard field work.Consumption was expected to differ among thedifferent work sectors: specifically, tractor workerswere expected to drink less because of the moreexacting skills required.

ResultsConsumption at Work. First we must note that inthis context, 'at work' refers to activity in the place

of work. Lunch and dinner are excluded sincepeople generally eat and drink alcohol at homeduring the day. Nearly two-thirds of the sample(64% or 119 interviewees) drink while working.Average alcohol intake is 16.3 g of anhydrousalcohol per capita per day (SD = 22.4). If onlydrinkers are considered, average consumption risesto 25.4 (SD=23.4). Over a third (36% of inter-viewees) do not drink at work.

No significant differences in alcohol consumptionare found among the aforementioned work sectors.

Drinking water seems to be inversely correlatedwith wine drinking. This seems to suggest thatduring worktime wine drinking may have a thirst-quenching function.

The average total daily alcohol consumption (bothin and out of the work context), is as high as 87.9 gof anhydrous alcohol per capita per day(SD = 44.8). Considering drinkers only, the figure is88.9 g (SD=44.1).

According to the interviewees' opinion, no realchange in such figures has occurred during the last 3years. Only 2 persons (1.1% of interviewees)describe themselves as an abstainer. The majority(53.2%) instead indicate that they drink more than80 g per capita per day of anhydrous alcohol (Table6).

Wine is preferred to beer and to liquor (Table 1)and is usually the first drunk in the lifecycle. Theaverage beginning age (the subjects' age less thereported number of years of alcohol intake) for wineis 16.8 years old (SD= 10.2); for all the other drinksit is around 30 years old.

As far as interactional patterns are concerned,wine and beer are especially drunk in the familycontext while liquors are consumed among friends.

Table 6. Frequency Distribution of Greve in Chianti Agricultural Workers(Florence, 1984) according to six classes of anhydrous alcohol daily intake on

the job and during the whole day

Anhydrous alcoholintake

Intake(g per day)

00.1-40

40.1-6060.1-8080.1-120>120

TOTAL

N

67931835

—

186

On the job

%

36.050.09.71.62.7—

100.0

Duringwhole

N

22131336237

186

theday

%

1.111.316.717.733.319.9

100.0

Alcohol Drinking Patterns and Work Areas 1177

Moreover no statistically meaningful relationshipis found between drinking alcohol and any of themore common socio/demographic variables.

The few women sampled drink large quantities ofalcohol (i.e. 61.7 g of anhydrous alcohol per capitaper day) even if male drinking is much more (89.8 gper capita per day).

The amount of alcohol drunk is consistently highin all age groups, in different civil statuses andamong different levels of education.

Considering our hypothesized association betweenthe amount of alcohol consumed and aspects ofwork, this high level of drinking is found to hold forall the working sectors identified.

Finally, non smokers are 60.5% of the sample.Among smokers, the major subgroup smokes 11-20cigarettes per person per day; they constitute 20.5%of the interviewees. No significant correlation isfound between smoking on the one hand and alcoholconsumption on the other. However, 51.4% ofdrinkers of more than 120 g per day do not smokeand 52.7% of non-smokers drink more than 80 g perday.

Drinking water and drinking alcohol during thewhole day are not correlated; this differs from whatis found between water intake and alcohol consump-tion at work.

Qualities Attributed to BeveragesRegarding the qualities attributed by the inter-viewees to the different types of beverages, theresults are the following:

(i) Wine is considered a 'connoisseurs drink': bothin the study of the general population in 1977 and inthe 1984 rural sample this quality scored as topamong 14. It is perceived as 'not harmful' ('harm-fulness' is the quality holding the last position bothin 1977 and 1984); on the contrary, it is perceived tobe 'good for the health'.

(ii) In keeping with common cultural values therural interviewees as well as those in the 1977sample of the general population perceive beer asbeing 'thirst quenching', bitters as 'helping diges-tion', and aperitifs as 'whetting the appetite'.

Conclusions(1) People drink mostly out of work hours andmuch more than in 1977. Thus drinkers of morethan 80 g during the whole day are 53.2% of thesample, as opposed to 13% in the general populationsurveyed in 1977.

(2) As could be expected, wine is the alcoholpredominantly consumed. Drinking wine seems tobe very rooted among the people in Greve. Such abeverage seems to be endowed with time-honouredpositive qualities.

(3) The characteristics of different work sectoractivities do not appear to affect the level of alcoholconsumption.

(4) Drinking wine may have the function of thirstquenching during work time.

Conclusions: alcohol consumption patterns inwork areas in FlorenceThere are some methodological difficulties in com-paring the 1977 survey with the 1983 study offactory employees and the 1984 one on ruralworkers. For example, the 1977 study is of thegeneral population while the latter two concentrateon two work areas following a systems approach.Moreover, in the first study the sample is propor-tional according to sex and age. In the latter ones thesample is proportional to work sectors or roles, andconsequently age groups are not equally repre-sented, and females are fewer than men (refiectingthe small number of working women).

Within the above-mentioned limitations in com-paring the three surveys, we may however draw thefollowing general conclusions:

(1) wine is the principle beverage consumed,conforming to the wine-growing culture of theTuscany region; its contribution to the total alcoholconsumed remains around 90% in the variouscontexts surveyed.

(2) drinking is preferred in a social context: in thefamily or with friends, almost never alone;

(3) the hypothesized relationship between spe-cific jobs within a given work area surveyed and theconsumption of alcohol is not supported; howeverclerks as a group reveal themselves as drinking lessthan blue collar workers;

(4) there is a culture of drinking more rooted inthe country and particularly among the agriculturalworkers and residents of the rural area; especiallyhere wine is considered the drink endowed withhighly positive qualities;

From data on the perceptions of beverages, anactual restrictive tendency toward drinking, seeingwine as harmful, emerges in the factory context; thismight be part of a broader health movement that isappearing in Western culture, according to whichthe ideal is the care of the body and abstinence fromtobacco and alcohol.

1178 AUaman AUamani et al.

AcknowledgementsRoberto Foschini, Luciano Gennai, GiovanniMosca, Roberto Pellegrini, Anna Sodi, Paola Giuntiparticipated in the present studies by interviewingthe 1983 and 1984 samples.

We are very grateful to Deborah Ruth Gordonand to Raimondo Maillart for the English transla-tion of this paper and for their helpful comments onit. We also thank Jonathan Chick for his invaluablesupport and suggestions on the manuscript.

The three epidemiological studies presented inthis paper have been funded by OsservatorioEpidemiologico della Regione Toscana, Italy.

ReferencesAIMONE, A., ALLEMANO, P., CORZANI, A. & SOLAVAGIONE,

L. (1986) Alcolismo e Ospedalizzazione, £/)idemto/o^ae Prevenzione, 26, pp. 28-31.

BOZZA, G. F. (1982) Aspetti medici del consumo di alcol.Danno epatico, in: REGIONE TOSCANA—DipartimentoSicurezza Sociale (Ed.) Alcol, Aspetti economici, sociali,medici, psicologid, pp. 199-206 (Firenze, Nuova SerieAtti Regione Toscana).

ClANFERONl, R. & PiSANi, P. L. (1987) Situazione eprospettive della viticultura toscana, in: REGIONE Tos-CANA (Ed.) Alcol, Aspetti nutrizionali, medico sociali edeconomici, pp. 5-44 (Roma, II Pensiero Scientiflco).

CoTTlNO, A. & MORGAN, P. (1985) Four country profiles:Italy, in: M. GRANT (Ed.) Alcohol Policies, pp. 83-92(Copenhagen, WHO Regional Publications).

IANDOLO, C. & CAPUANO, P. (1968a) II piccolo e grandealcolismo in Italia, Policlinico Pratico, 75, 36, pp.1157-1175.

IANDOLO, C . & CAPUANO, P. (1968b) II piccolo e grandealcolismo in Italia, Policlinico Pratico, 75, 37, pp.1189-1214.

GLATT, M . M . (1974) Drugs, Society and Man: A Guide toAddiction and its Treatment (Lancaster, MTP).

MODONUTTI, G . B., BENUSSI, G . , NEDOCLAN, G . , GALLIM-BERTI, L., & GASPARINI, V. (1984) Disponibilita,consumi di bevande alcoliche e mortalita alcol-correlatain Italia, Minerva Medica, 74, pp. 1-9.

MORETTINI, A., ALLAMANI, A. & LOMUTO, C . (1982)Epidemiologia delle epatopatie da etanolo, in: C. A.BOGGIANO (Ed.) La malattia cronica di fegato, pp.179-191 (Pisa, Pacini).

PEDRAZZOLI, M . , TAFNER, G . & BAJOCCHI, E . (1974)L'alcolismo nel Trentino, Minerva Medica, 659, pp.778-789.

RAIMONDI, A. , MODONUTTI, G . B., ARTITO, F . (1987)Aspetti epidemiologici e tendenze di consumo dellebevande aicoliche in Italia, in: REGIONE TOSCANA (Ed.)Alcol, Aspetti nutrizionali, medico sociali ed economici,pp. 108-136 (Roma, II Pensiero Sciemifico).

RAUSA, G. , MODONUTTI, G . B., GALLIMBERTI, L . , CIBIN,M., BENUSSI, G. , GUIDI, E. , FEDELI, L . , ORLANDINI, D. ,CESSI, C . & GASPARINI, V. (1984) Consumo di alcol emortalita alcol correlata in Italia, Rivista Italiana diAlcologia, 3(1), pp. 3-11.

ROLLI, A. (1985) Un panorama critico della ricerca suU,alcol in Italia (1960-1981), in: Gruppo di ricerca suiproblemi dell 'alcolismo e delle tossicodipendenze (Ed.)L'alcol nella societa: sdema, cultura, e controllo sociale,pp. 57-85 (Torino, CELID).

SARACENI, G . (1975) L'alcolismo a Venezia, Con partico-lare riguardo alia diffusione dell 'epatopatia alcoolica.Res Medicae.

SIRIGATTI, S. (1982) Aspetti psicosociali del consumo dialcol. Note conclusive, in: Regione Toscana-Diparti-mento Sicurezza Sociale (Ed.) Alcol, Aspetti economici,sociali, medici, psicologici, pp. 189-198 (Firenze, NuovaSerie Atti Regione Toscana).

WALSH, D . (1982) Alcohol-related medicosocial problemsand their prevention (Copenhagen, WHO).

ZIEGLER-DRISCOLL, G . (1979) The Similarities in Fami-lies of Drug Dependents and Alcoholics, in: E. KAUF-MAN & P. KAUFMANN (Eds) Family Therapy of Drugsand Alcohol Abuse, pp. 19-39 (New York, GardnerPress).