Vol.1, No.3, 80-87 (2011)
doi:10.4236/ojpm.2011.13012
C
opyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/OJPM/
Open Journal of Preventive Medicine
Alcohol consumption patterns during transition and
economic growth in Estonia: Results from the 1996 and
2006 Health Interview Surveys
Kersti Pärna*, Inge Ringmets
Department of Public Health, University of Tartu, Tartu, Estonia.
*Corresponding Auth or: kersti.parna@ ut.ee
Received 15 July 2011; revised 7 September 2011; accepted 20 October 2011.
ABSTRACT
Aims: To investigate alcohol consumption pa-
tterns in Estonia in 1996, which represents the
transition period after the collapse of the Soviet
Union, and in 2006, which is characterized by
economic growth. Methods: Data from all 25 - 6 4
years old from the nationally representative
1996 and 2006 Estonian Health Interview Sur-
veys were analysed. The frequency of alcohol
consumption and heavy episodic drinking (HED)
were examined. The odds ratios of at least week-
ly alcohol consumption and HED according to
socio-demographic characteristics were calcu-
lated using logistic regression models. Results:
In comparison with 1996, at lea s t weekly alcoh ol
consumption was nearly twice as high among
men in 2006. No association between HED and
the study year was found. Alcohol consumption
was lower among respondents in the oldest age
group. Weekly alcohol consumption was lower
among non-Estonian men in both study years.
HED was higher among non-Estonian men and
women in 1996 but this difference disappeared
in 2006. Weekly alcohol consumption was lower
among less educated men in 1996 but not in
2006. No clear tendency in weekly alcohol con-
sumption by education was found among wo-
men. HED was higher among less educated men
in both study years; however, this association
was weaker in 2006. While HED was not asso-
ciated with education among women in 1996, an
inverse relationship was found in 2006. Conclu-
sions: The study confirmed that rapid societal
changes had an effect on alcohol consumption
patterns in Estonia. Estonia needs the imple-
mentation of a comprehensive alcohol policy to
decrease harmful alcohol consumption in the
whole society.
Keywords: Alcohol Consumption; Hea vy Episodic
Drinking; Socio-Demographic Factors; Adults;
Estonia
1. INTRODUCTION
Alcohol consumption per capita doubled in Estonia
from 1996 to 2006. While in 1996 the total adult per
capita pure alcohol consumption was 5.6 litres, by 2006
it reached 11.4 litres [1,2]. The proportion of men drin-
king more than 280 g pure alcohol per week increased
from 7.4% to 16.2%, and the proportion of women drink-
ing more than 140 g pure alcohol per week increased
from 2.7% to 4.7% between 1996 and 2006 [3]. In Estoni a
alcohol-related mortality was the lowest during 1988 -
1991 (3.5% of all death s) and increased to 9.1% in 2002
-2005 [4]. At the same time, in Estonia age standardized
mortality rates of alcoholic liver cirrhosis increased from
9.7 to 37.5 per 100,000 men and from 2.2 to 16.1 per
100000 women aged 25 - 64 in 1996-2006 [5].
One of the factors in the growth of alcohol con-
sumption and alcohol-related premature mortality in
Estonia could be related to societal changes. Estonia
regained its independence in August 1991 after the col-
lapse of the Soviet Union. The following societal re-
forms brought major changes to the residents of Estonia.
1996 represents a period of overwhelming transition. In
terms of economic development, it was characterized by
significant unemployment (9.9%) and the relatively low
gross domestic product (GDP) (4890 Euros per person)
[6]. On the other hand, 2006 is an example of economic
growth and relative prosperity. By 2006 unemployment
had dropped to 5.9% and the GDP had almost doubled
(9022 Euros per person) [6].
In order to gain some insight into how rapid societal
reforms and economic growth could be related to ch an g es
in alcohol consumption, the present paper compares
alcohol consumption patterns in Estonia in 1996 and
2006.
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/OJPM/
81
2. MATERIALS AND METHODS
The data of the cross-sectional, nationally represen-
tative Estonian Health Interview Surveys (EHIS) of
1996 and 2006 were used. The full details of the survey
methodologies have been described in published reports
[7,8].
2.1. Study Population
In short, the target population of the 1996 EHIS
consisted of men and women aged 15 - 79 on 1 January
1996 who had been permanent residents of Estonia at the
time of the 1989 population census. The sampling frame
was divided into 16 sampling units, and each of them
was stratified by gender and five-year age groups. All in
all 6019 eligible respondents were selected and for-
warded to the interviewers. Of these 4711 respondents
(2131 men and 2580 women) were interviewed. There-
fore, the crude response rate was 78.3%, and the cor-
rected response rate was 84.3% [7].
The sampling frame of the 2006 EHIS consisted of
men and women aged 15 - 84 on 1 January 2006 who
had been permanent residents of Estonia at the time of
the 2000 popu1ation census. The sampling frame was
divided into five regional sampling units. Each of them
was stratified by gender and five-year age groups. All in
all 11023 eligible respondents were forwarded to the
interviewers. Of these 6434 individuals (3110 men and
3324 women) completed the interviews. While the crude
response rate was 58.4%, the corrected response rate was
60.2% [8].
The present study focused on 25 - 64 year s old adults;
the total number of respondents was 2690 (1298 men
and 1392 women) in 1996 and 3316 (1543 men and
1773 women) in 2006. Participants younger than 25
years of age were excluded because they were more
likely to still be in higher educatio n. Th e upper limit was
used because of selective mortality in the older age
groups.
2.2. Procedure
Face-to-face interviews were conducted in the Esto-
nian and Russian languages, and trained interviewers
were used. EHIS gathered data on a wide range of vari-
ables related to health status, life styles, health services
uti- lization, household income, socio-economic status,
and societal attitudes.
2.3. Alcohol Consumption
The 1996 EHIS consisted of 13 questions about alco-
hol consumption. Overall alcohol consumption was
measured by the question: “Have you ever in your life
drunk alcohol more than ju st to taste? ” (yes/no). Alcoho l
consumption during the previous four weeks was mea-
sured by the question: “Thinking back to the previous
four weeks, have you used alcohol during this period?”
(yes/no). A detailed assessment of alcohol consumption
was acquired for the previous four weeks. Frequency of
any alcoholic beverage consumption was assessed with
the following question: “How frequently did you use
alcohol during the last four weeks?” The options
included “every day (almost every day)”, “3 - 4 times a
week”, “1 - 2 times a week”, “1 - 3 times a month”.
Drink-specific questions focused on the consumption of
the typical amount of each beverage per occasion: “What
kind of alcohol and how much did you yourself usually
used at one sitting (p er day) during the last four weeks? ”
The possible responses were “beer__drinks”, “wine or
other light alcohol__drinks”, “vodka or other strong
alcohol__drinks”.
The 2006 EHIS dataset measured alcohol con-
sumption by means of 22 questions. Overall alcohol
consumption was measured by the question: “Have you
ever in you life drunk alcohol more than one shot of
strong alcohol, a glass of wine or a bottle of beer?”
(yes/no). Alcohol consumption during the previous four
weeks was measured by the question: “Thinking back to
the last four weeks, have you used alcohol during this
period?” (yes/no). Frequency of consumption of all types
of alcoholic drinks was assessed with the following
question: “How frequently have you drunk beer/light
alcoholic drink/win e/vodk a, or some othe r kind of strong
alcohol during last four weeks?” The possible responses
were “every day/almost every day”, “3 - 4 times a week”,
“1 - 2 times a week”, “1 - 3 times a month”, “I haven’t
drunk (this beverage) at all”. A typical amount of alcohol
per one occasion was asked by the following question:
“How much (in drinks/doses) beer/light alcohol/wine/
vodka, or some other kind of strong alcohol did you
usually drink at o ne sitting during the last four weeks?”
In both study years the information concerning the
amount of alcohol consumed was converted from litres
to grams of pure alcohol per week assuming that an 0.5l
bottle of beer or light alcohol, 100 ml of wine or 30 ml
(one shot) of spirits contained 10 grams of pure alcohol
(=one drink) and 0.5 bottle of strong beer (over 6%)
contained 35 grams of pure alcohol. For both study years,
the respondents who had reported the consumption of
any type of alcohol at least weekly were classified as
weekly drinkers. Drinking any kind of alcohol at least
seven drinks per occasion among men and at least five
drinks per occasion among women were classified as
heavy episodic drinking (HED).
2.4. Socio-Demographic Factors
The analysis used the following socio-demographic
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. http://www.scirp.org/journal/OJPM/Openly accessible at
82
factors: age, ethnicity, and education. Age at the time of
the interview was used and aggregated into 10-year age
groups: 25 - 34, 35 - 44, 45 - 54, and 55 - 64. Ethnic
origin was measured by self-identified n ationality (Esto-
nian/Russian/other) an d was dichotomized as “Estonian”
and “non-Estonian”. Education levels were divided into
three cat egori es: “bas ic ” (prima ry le vel e ducati on), “secon-
dary” (secondary education), and “higher” (university or
academic education).
2.5. Statistical Analysis
As gender is a strong determinant of drinking rates,
the data for men and women were analysed separately.
The characteristics of the study sample, frequency, and
volume of alcohol consumption were described by fre-
quency tables.
Associations between alcohol consumption and socio-
demographic variables were estimated using logistic re-
gression models. The odds ratios (OR) with confidence
intervals (95% CI) of at least weekly consumption of
alcohol and HED during the previous four weeks were
estimated using multiple logistic regression analysis with
adjustment for all the socio demographic variables si-
multaneously. The models used dichotomized alcohol
consumption as a dependent variable and socio demo-
graphic factors as explanatory variables.
The data were analyzed using the statistical package
Stata 10 [9].
3. RESULTS
3.1. General Overview of the Sample
Table 1 shows the socio-demographic characteristics
of the sample. In both study years Estonians comprised
two thirds and non-Estonians (mainly Russians) about
one third of the sample. More than 50% of the respon-
dents had secondary education. Compared to 1996, the
proportion of respondents with higher education was
higher in 2006 (p < 0.00 1) .
Among the men in 1996, 94.0% had drunk in their life
more than just to taste it) and 93.8% in 2006 (p > 0.05).
Among the women the respective proportions were
73.8% and 85.2% (p < 0.001). During the previous four
weeks 77.7% of the men had consumed alcohol in 1996
and 77.6% in 2006 (p > 0.05). Among women, the res-
pective proport i ons we re 67 .2% and 6 6. 9% (p > 0.0 5) .
3.2. At Least Weekly Alcohol Consumption
by Socio-Demographic Factors
While in 1996 37.7% of men consumed alcohol at
least weekly, in 2006 54.4% did so (p < 0.001). At the
same time, the proportion of women consuming alcohol
at least weekly remained at about the same level (18.1%
and 17.9%, respectively).
In both study years more than half of the men and
women drinking alcohol weekly were from the two
younger age groups (Table 2). In 1996 and 2006 Es-
tonians constituted more than two thirds of the adults
who consumed alcohol weekly. Compared to 1996, in
2006 the proportion of university-educated women who
drank weekly was significantly higher (39.7% and 24.8% ,
respectively), but the proportion of women with a secon-
dary education significantly lower (50.8% and 69.6%,
respectively).
Among men age was not associated with frequency
Table 1. Socio-demographic characteristics of 25 - 64 years old respondents, by gender, 1996 and 2006 Estonian Health
Interview Surveys.
Men Women
1996 2006 1996 2006
Characteristics
n % n % n % n %
Age group
25 - 34 273 28.8 290 26.0 190 27.6 246 24.5
35 - 44 262 27.6 313 28.0 223 32.4 272 27.2
45 - 54 220 23.2 281 25.2 165 23.9 278 27.8
55 - 64 193 20.4 233 20.8 111 16.1 205 20.5
p-value 0.493 0.009
Ethnicity
Estonian 620 65.4 725 64.9 449 65.2 677 67.6
non-Estonian 328 34.6 392 35.1 240 34.8 324 32.4
p-value 0.814 0.291
Education
higher 156 16.5 197 21.1 151 21.9 306 30.6
secondary 566 59.7 684 61.2 457 66.3 605 60.4
basic 226 23.8 197 17.7 81 11.8 90 9.0
p-value <0.001 <0.001
Total 948 100 1117 100 689 100 1001 100
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/OJPM/
83
Table 2. Consumption of any type of alcohol at least once a week during the previous four weeks among men and
women by socio-demographic factors, 1996 and 2006 Estonian Health Interview Surveys (n, %, 95% CI).
Men Women
Characteristics 1996 2006 1996 2006
Age groups 357 608 125 179
25 - 34 30.5 (25.8 - 35.3) 24.7 (21.2 - 28.1) 28.0 (20.1 - 35.9) 30.2 (23.4 - 36.9)
35 - 44 28.3 (23.6 - 33.0) 27.1 (23.6 - 30.7) 37.6 (29.1 - 46.1) 26.3 (19.8 - 32.7)
45 - 54 22.7 (18.3 - 27.0) 25.5 (22.0 - 29.0) 21.6 (14.4 - 28.8) 28.5 (21.9 - 35.1)
55 - 64 18.5 (14.5 - 22.5) 22.7 (19.4 - 26.0) 12.8 (0.69 - 18.7) 15.1 (0.98 - 20.3)
Ethnicity
Estonian 72.0 (67.3 - 76.6) 68.8 (65.1 - 72.4) 67.2 (59.0 - 75.4) 69.8 (63.1 - 76.6)
non-Estonian 28.0 (23.4 - 32.7) 31.3 (27.6 - 34.9) 32.8 (24.6 - 41.0) 30.2 (23.4 - 36.9)
Education
higher 21.6 (17.3 - 25.8) 21.9 (18.6 - 25. 2) 24.8 (17.2 - 32.4) 39.7 (32.5 - 46.8)
secondary 54.3 (49.2 - 59. 5) 58.2 (54.3 - 62. 1) 69.6 (61.5 - 77.7) 50.8 (43.5 - 58.2)
basic 24.1 (19.7 - 28.5) 19.9 (16.7 - 23.1) 5.6 (1.6 - 9.6) 9.5 (5.2 - 13.8)
of alcohol consumption (Table 3). Compared to women
in the youngest age group, the odds ratio of alcohol
consumption at least once a week was slightly lower in
the oldest age group of women in 2006. In both study
years alcohol consumption at least weekly was lower
among non-Estonian men. Weekly alcohol consumption
was not associated with ethnicity among women. While
weekly alcohol consumption was lower among less
educated men in 1996, no such an association was found
in 2006. Compared to the university-educated women,
weekly alcohol consumption was slightly lower among
women with a basic education in 1996 but lower among
women with a secondary education in 2006 .
3.3. Heavy Episodic Drinking (HED) by
Socio-Demographic Factors
The proportion of men drinking seven or more drinks
at one sitting during the previou s four weeks was 42.7%
in 1996 and 39.0% in 2006 (p > 0.05). Among women,
11.2% reported having drunk five or more drinks per
occasion in 1996 and 12.1% in 2006 (p > 0.05). In com-
parison with 2006, in 1996 the proportion of HED was
higher among women in the younger age groups (Table
4). In both study years nearly two thirds of HED men
were Estonians. While in 1996 52.0% of HED women
were Estonians, in 2006 this increased to 63.6%. Com-
pared to the first study year, the proportion of univer-
sity-educated HED men and women was higher in 2006
(6.7% and 16.7% for men and 14.3% and 21.5% for
women, respectively). Compared with 1996, in 2006 the
proportion of HED men with basic education was lower,
but the proportion of HED among less educated women
was higher.
Logistic regression analysis demonstrated that com-
pared to the 25 - 34 years old adults, HED was lower in
the oldest age group of men and women in both study
years (Table 5). Compared to Estonian respondents, the
odds of being HED was significantly higher among non-
Estonian men and women in 1996 but not in 2006.
Compared to university-educated men, the odds of
drinking at least seven drinks per occasion was signi-
ficantly higher among less educated men in both study
years. While in 1996 no association between HED and
education was found among women, in 2006 HED was
higher among lower educated women.
4. DISCUSSION
This article focused on alcohol consumption in dif-
ferent socio-demographic groups among 25 - 64 years
old adults in 1996 and 2006. The strength of this study
was that data was based on the large scale nationally
representative Estonian Health Interview Survey. At the
same time, all the questionnaire based surveys of alcohol
consumption have several limitations.
4.1. Limitations
People usually underestimate their consumption and
tend to disregard less common frequency and heavy drink-
ing occasions. Also, heavy drinkers are generally less
likely to participate in surveys. Moreover, alcohol con-
sumption is on average much higher in those (e.g. non-
household) subgroups of population that are excluded
from survey sampling frames [10]. Compared to 1996,
the response rate was much lower in 2006, and this
could contribute to underestimation of heavy drinking
especially in the latter study year. Furthermore, the ques-
tions concerning alcohol consumption varied slightly in
different study years in this However, a major survey.
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/OJPM/
84
Table 3. Prevalence odds ratio (POR) and 95% confidence intervals (CI) for consuming any type of alcohol at least once a
week among men and women, 1996 and 2006 Estonian Health Interview Surveys.
Men Women
Adjusted ORa (95% CI) Adjusted ORa (95% CI)
Characteristics 1996 2006 1996 2006
Age groups
25 - 34 1.00 1.00 1.00 1.00
35 - 44 0.93 (0.65 - 1.32) 1.06 (0.77 - 1.46) 1.18 (0.72 - 1.93) 0.78 (0.50 - 1.21)
45 - 54 0.80 (0.55 - 1.17) 1.19 (0.85 - 1.66) 0.93 (0.53 - 1.63) 0.86 (0.56 - 1.34)
55 - 64 0.69 (0.45 - 1.04) 1.35 (0.95 - 1.92) 0.88 (0.46 - 1.72) 0.57 (0.34 - 0.95)
Ethnicity
Estonian 1.00 1.00 1.00 1.00
non-Estonian 0.65 (0.49 - 0. 87) 0.69 (0.54 - 0. 89) 0.86 (0.56 - 1.30) 0.92 (0.65 - 1.32)
Education
higher 1.00 1.00 1.00 1.00
secondary 0.52 (0.36 - 0.75) 0.86 (0.64 - 1.17) 0.92 (0.58 - 1.46) 0 .61 (0.43 - 0.87)
basic 0.65 (0.43 - 1.00 ) 1.22 (0.83 - 1. 80) 0.39 (0.16 - 0.96) 0.82 (0.45 - 1.50)
Table 4. Heavy episodic drinking (HED) during the previous four weeks among men and women by socio-demographic factors,
1996 and 2006 Estonian Health Interview Surveys (n, %, 95% CI).
Men Women
Characteristics 1996 2006 1996 2006
Age groups 405 436 77 121
25 - 34 31.1 (26.6 - 35.6) 27.1 ( 22.9 - 31.2) 40.3 (29.3 - 51.2) 29.8 (21.6 - 37.9)
35 - 44 28.2 (23.8 - 32.5) 29.8 ( 25.5 - 34.1) 31.2 (20.8 - 41.5) 24.8 (17.1 - 32.5)
45 - 54 23.2 (19.1 - 27.3) 26.2 ( 22.0 - 30.3) 20.8 (11.7 - 29.8) 32.2 (23.9 - 40.6)
55 - 64 17.5 (13.8 - 21.2) 17.0 ( 13.4 - 20.5) 7.8 (1.8 - 13.8) 13.2 (7.2 - 19.3)
Ethnicity
Estonian 57.8 (53.0 - 62. 6) 62.4 (57.8 - 66.9) 52.0 (40.8 - 63.1) 63.6 (55.1 - 72. 2)
non-Estonian 42.2 (37.4 - 47.0) 37.6 (33.1 - 42.2) 4 8 .1 (36.9 - 59.2) 36.4 (27.8 - 44. 9)
Education
higher 6.7 (4. 2 - 9.1) 16.7 (13.2 - 20.2) 14.3 (6.5 - 22.1) 21.5 (14.2 - 28.8)
secondary 65.4 (60.8 - 70.0) 61.7 (57.1 - 66. 3) 76.6 (67.2 - 86.1) 65.3 (56.8 - 73.8)
basic 27.9 (23.5 - 32.3) 21.6 (17.7 - 25.4) 9.1 (2.7 - 15.5) 13.2 (7.2 - 19.3)
Table 5. Prevalence odds ratio (POR) and 95% confidence intervals (CI) for heavy episodic drinking (HED) among men and
women, 1996 and 2006 Estonian Health Interview Surveys.
Men Women
Adjusted OR* (95% CI) Adjusted ORa (95% CI)
Characteristics 1996 2006 1996 2006
Age groups
25 - 34 1.00 1.00 1.00 1.00
35 - 44 0.88 (0.62 - 1.26) 1.07 (0.77 - 1.49) 0.60 (0.33 - 1.06) 0.72 (0.43 - 1.22)
45 - 54 0.84 (0.57 - 1.23) 0.99 (0.71 - 1.39) 0.53 (0.27 - 1.02) 0.91 (0.55 - 1.50)
55 - 64 0.60 (0.39 - 0.91) 0.64 (0.45 - 0.93) 0.29 (0.11 - 0.75) 0.45 (0.24 - 0.84)
Ethnicity
Estonian 1.00 1.00 1.00 1.00
non-Estonian 1.87 (1.41 - 2. 49) 1.21 (0.94 - 1.56) 1.90 (1.17 - 3.10) 1.15 (0.77 - 1.73)
Education
higher 1.00 1.00 1.00 1.00
secondary 3.78 (2.40 - 5.95) 1.41 (1.02 - 1.93) 1.74 (0.88 - 3.43) 1.67 (1.04 - 2.67)
basic 5.49 (3.32 - 9.08) 2.15 (1.44 - 3.20) 1.66 (0.59 - 4.67) 2.49 (1.25 - 4.95)
*Each OR was adjusted for all the other characteristics in the table.
limitation of this survey was related to the conversion of
alcohol consumption from litres to grams of pure alcohol
per week assuming that an 0.5 l bottle of beer contains
10 grams of pure alcohol (=one drink). Typically, an 0.5
l bottle of beer (4.2% - 4.7%) contains 20 gra ms of pure
alcohol. Thus, the present study underestimated HED.
Unfortunately, the questionnaires defined the amount of
different types of alcohol consumed per one occasion in
terms of drinks; therefore, it was impossible to reconvert
the data without information concerning the strength of
beer. Finally, a detailed assessment of alcohol consum-
ption was acquired for previous four weeks only. Thus,
the respondents who had not consumed any alcohol
during the previous four weeks were left out from the
analysis. With these caveats several observations can be
made.
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/OJPM/
85
4.2. Socio-Demographic Differences in
Alcohol Consumption
Proportion of abstainers was, not surprisingly, higher
among women in both study years in Estonia. In com-
parison with 1996, the proportion of lifetime abstainers
was the same among men, but significantly lower among
women in 2006. This cou ld be explained by traditionally
low alcohol consumption among women during the
Soviet period, increasing after its collapse [3,11]. At the
same time, the share of propor tion of drinkers during the
previous four weeks was stable in the two study years in
Estonia.
Frequency and the amount of drinking constitute an
essential part of alcohol consumption patterns. Com-
pared to 1996, the proportion of alcohol consumption at
least weekly among men was nearly twice as high in
2006. In 1996 one third of men but in 2006 one in two
men consumed alcohol at least weekly. Among women,
nearly one fifth were weekly alcohol drinkers in both
study years. Compared to Finbalt study in 2000 - 2006,
consumption of any type of alcohol beverage at least
weekly was similar among men in Estonia and Finland,
but it was much lower among women in Estonia than in
Finland [3]. At the same time McKee et al. [12] found
that among 19 - 64 years old, 61% of men and 26% of
women were weekly drinkers in Estonia in 1997. The
inconsistency in the findings might be explained by
different age groups used in these studies. Moreover, the
survey of McKee et al. [12] was primarily designed to
obtain data on nutrition, and therefore the questions
about alcohol drinking were rather basic. In 2001, the
proportion of weekly drinkers among men 18 years old
and more was 58% in Estonia, 45% in Latvia, 59% in
Lithuania, and 56% in Poland (respectively 22%, 15%,
23%, and 23% among women) [13]. Again, the results
are not comparable because of different age groups and
years used in these surveys.
A simple measure of HED has been related to
increased risk of mortality [14]. In this study ‘at risk’
HED was defined as consumption of seven and more
drinks per occasion for men and five or more for women.
Compared to the women, proportion of HED was much
higher among men in both study years. Also Pomerlau et
al. reported much higher proportion of heavy episodic
drinking among men in eight countries of the former
Soviet Union. No difference in HED was found between
1996 and 2006 in Estonia. Nevertheless, the proportion
of men drink ing 280 g r ams and th e p ropor tio n o f wo men
drinking more than 140 grams pure alcohol per week
increased significantly between 1996 and 2006 in
Estonia [3]. This inconsistency might be related to the
different methodology: in EHIS was measured HED per
occasion, but in Finbalt study in pure alcohol per week.
Compared to the youngest age group, in 2006 weekly
alcohol consumption was lower in the oldest age group
of women, but HED was lower among 55 - 64 years old
adults in both study years. These findings are similar to
those of the Finbalt surveys on health behaviour, where
drinking more than six drinks per occasion was more
common among younger women in all three Baltic
countries and Finland [15,16]. The higher proportion of
alcohol drinking among adults in the younger age groups
could reflect liberal alcohol policy and attitudes towards
alcohol consumption in Estonia [17].
In general, the proportion of non-Estonian men, but
not women, who drink at least weekly, was lower in both
study years. At the same time, the proportion of HED
non-Estonian men and women was higher in 1996, but
the ethnic gradient disappeared in 2006. These findings
are in accordance with earlier studies in Estonia [3].
Higher frequency of alcohol drinking among Estonian
men could be explained by the preference for beer while
non-Estonian men prefer to drink strong alcohol. Pärna
et al. [3] reported that in Estonia more than half of men
are weekly beer drinkers and more than a fifth drink
spirits weekly. During 1994-1999 the consumption of
beer doubled among men in Estonia [18]. In 1996-2006
the consumption of beer nearly tripled (from 33.2 to 85.8
litres per capita) while the consumption of spirits nearly
doubled (from 6.4 to 12.3 litres per capita) among adults
in Estonia [1,2]. Moreover, in terms of pure alcohol the
consumption of beer increased the most, too. In 1996
consumption of beer was 2 litres in pure alcohol per
capita, but in 2006 it reached 4.4 litres (2.9 litres of
spirits and 4.7 litres of pure alcohol, respectively) [1,2].
Disappearance of ethnic differences in HED in 2006
could be explained by the integration (including drinking
habits) of non-Estonians into Estonian society during 15
years after Estonia regained independence in 1991.
Moreover, compared to 1996, in 2006 a diver se selection
of light alcoholic beverages was available. This might
also influence the preferences for different types of
alcohol among both ethnic groups. At the same time
non-Estonians were more likely to be surrogate alcohol
(legal manufactured ethanol-containing liquids not in-
tended for consumption) drinkers [19,20] and more
likely to die from alcohol-related causes than Estonians
in 1996-2006 [3 ,4].
In general, the educational gradient among men and
women showed that better-educated people were more
likely to drink alcohol at least weekly, but less educated
were more often HED in Estonia. While less educated
men drank alcohol with lower frequency in 1996, then
no such an association was found in 2006. At the same
time no clear tendency in weekly alcohol consumption
by education was found among women. Compared to
university-educated men, HED was more common among
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/OJPM/
86
less educated men in both study years. Nevertheless, in
2006 the association between HED and education was
weaker. On the other hand, while no association was
found between HED and education among women in
1996, then HED was more common among less edu-
cated women in 2006. Also, the Finbalt survey reported
similar results for men and women in Estonia, Latvia,
Lithuania and Finland [3,15]. At the same time Po-
merleau et al. [11] did not find educat ional di ff erences in
HED among adults in eight former Soviet countries
which might be explained by different definition of HED
in these studies. In 1996-2006 there was an inverse
relationship between the educational level and the risk of
alcohol-related death [3,4]. Nevertheless, educational
and socio-economic differences exposed by self-reported
questionnaire studies in HED and heavy alcohol con-
sumption are less marked than the differences in re-
gister-based studies of alcohol-related mortality [5].
4.3. Societal Context in 1996 and 2006
Considering the frequency and amount of alcohol
drinking in 1996 and 2006, it is worth analysing them in
the context of societal changes. In Estonia th e year 1996
presented a period of transition and tumultuous changes
in many areas, but 2006 was characterized by economic
boom and prosperity. Compared to the year 1996,
unemployment was twice lower, but the GDP was nearly
twice as high in 2006.
During 1996-2006 alcohol policy was virtually non-
existent in Estonia. There was a national alcoho lism and
drug abuse pr evention programme for 1997 -2007, which
was continued since 2004 under national drug abuse
prevention strategy (1994-2012) [21]. This programme
mainly focused on the creation of a nation-
wide information system for evaluating the damage
caused by alcohol and drug abuse. In 1996-2006 the
share of alcohol excise in excise revenue decreased 13%
(from 43% in 1996 to 30% in 2006) constituting 6% in
1996 and 4% in 2006. This had an effect on alcohol
consumption in Estonia and on alcohol policy and
alcohol drinking in the neighbouring countries, such as
Finland. When Estonia joined the European Union on 1
May 2004, Estonia became a destination of alcohol
tourism, and citizens imported low-priced alcohol from
Estonia to Finland on a large scale [22,23]. In turn, it
brought about reduced alcohol prices and increased
drinki ng i n Fi nland [16].
An increase in alcohol consumption has been shown
to be related to increased income and wealth in society
[24]. Moreover, reduced alcohol prices bring about
increased alcohol consumption [16]. In 1996-2006 the
prices of alcoholic beverages increased in Estonia (1.3
times for domestically produced beer and 1.4 times for
vodka). However, the average price increase has been
slower than the increase in the consumer price index, as
well as slower than the increased income of inhabitants,
which most likely has also contribu ted to the increase of
alcohol consumption [1]. Thus, affordability of alcohol
increased more than 50% in Estonia in 1996-2004,
which was o ne of the highes t i n the E uropean Union [23].
While in 2000 one could by 199 litres of beer or 28 litres
of vodka, then in 2006 this amount nearly doubled (348
and 54 litres, respectiv ely) [1,2]. While in 1996 the sales
of alcoholic beverages in litres of pure alcohol per cap ita
(including purchases by tourists) was 2.5 for strong
alcohol, 1.9 for beer and 0.5 for wine in Estonia, the
respective figures for 2006 were 5.7, 5.6 and 1.1
(additionally 1.7 litres for light alcoholic beverages)
[1,2]. In general, alcohol was easily available in Estonia.
In 2006 there were 195 retail shops per 100,000
inhabitants in Estonia, 6.3 in Finland, and 4.5 in Sweden
[1,2]. Thus, policy actions in Estonia should include
reduction of the density of alcohol outlets and clearer
separation of alcoholic beverages from other goods in
retail stores [17].
On the basis of this situation, recommendations for
the improvement of alcohol policy were developed in
2004 in Estonia. The most appropriate actions were the
following: increase of alcohol taxation by 50% (com-
pared to the 2004 level), complete ban of alcohol adver-
tising, further restrictions on alcohol sales (with regard
to both place and time), and brief counselling in primary
health care [25]. The first step of strengthening the
alcohol policy in Eston ia was a 5% increase in the excise
tax on alcohol with the exception of wine in 2005 [17].
4.4. Conclusions
The present study confirmed that economic upturn
had an important role in the increase of alcohol con-
sumption in Estonia. Also, rapid societal changes had an
effect on alcohol consumption patterns. Estonia needs
the implementation of a comprehensive alcohol policy
where price regulation is used in a combination with
other effective and cost-effective measures against harm-
ful alcohol consumption [26]. Special attention should
be paid to the less educated subgroup.
5. ACKNOWLEDGEMENTS
The study was supported by the Estonian Science foundation (grant
ETF7416) and the Norwegian Financial Mechanism/EEA (grant
EE0016) and the Estonian Ministry of Education and Science (target
funding SF0180060s09).
REFERENCES
[1] Estonian Institute of Economic Research (2008) Alcohol
market, consumption and harms in Estonia. Yearbook,
K. Pärna et al. / Open Journal of Preventive Medicine 1 (2011) 80-87
Copyright © 2011 SciRes. http://www.scirp.org/journal/OJPM/Openly accessible at
87
Tallinn.
[2] Estonian Institute of Economic Research. (2010) Alcohol
market, consumption and harms in Estonia. Yearbook,
Tallinn.
[3] Pärna, K., Rahu, K., Helakorpi, S. and Tekkel, M. (2010)
Alcohol consumption in Estonia and Finland: Finbalt
survey 1994-2006. BMC Public Health, 10, 261.
[4] Rahu, K., Pärna, K., Palo, E. and Rahu, M. (2009) Con-
trasts in alcohol-related mortality in Estonia: education
and ethnicity. Alcohol & Alcoholism, 44, 517-522.
doi:10.1093/alcalc/agp035
[5] Pärna, K. and Rahu, K. (2010) Dramatic increase in al-
coholic liver cirrhosis mortality in Estonia in 1992-2008.
Alcohol & Alcoholism, 45, 548-551.
[6] Statistics Estonia (2011) Statistical database. http://stat.ee
[7] Leinsalu, M., Grintshak, M., Noorkõiv, R. and Silver, B.
(1998) Estonian Health Interview Survey. Methodologi-
cal Report. Institute of Experimental and Clinical Medi-
cine, Tallinn.
[8] Oja, L., Matsi, A. and Leinsalu, M. (2008) Methodologi-
cal report. Estonian Health Interview Survey 2006. Na-
tional Institute for Health Development, Tallinn.
[9] StataCorp (2007) Stata statistical software: Release 10.
College Station, StataCorp LP, College Station.
[10] Mäkelä, P. and Huhtanen, P. (2010) The effect of survey
sampling frame on coverage: the level of and changes in
alcohol-relate d mortality in Finland as a test case. Addic-
tion, 105, 1935-1941.
doi: 10.1111/j .1 360-0443.2010.03069.x
[11] Pomerleau, J., McKee, M., Rose, R., Haerpfer, C.W.,
Rotman, D. and Tumanov, S. (2008) Hazardous alcohol
drinking in the former Soviet Union: A cross-sectional
study of eight countries. Alcohol & Alcoholism, 43, 351-
-359. doi:10.1093/alcalc/agm167
[12] McKee, M., Pomerleau, J., Robertson, A., Pudule, I.,
Grinberga, D., Kadziauskiene, K., Abaravicius, A. and
Vaask, S. (2000) Alcohol consumption in the Baltic Re-
publics. Journal of Epidemiology & Community Health,
54, 361-366. doi:10.1136/jech.54.5.361
[13] Reitan, T.C. (2004) On the verge of EU membership:
alcohol cultures in the Baltis Sea region in a European
perspective. Contemporary Drug Problem, 31, 287-327.
[14] Laatikainen, T., Manninen, L., Poikolainen, K. and Var-
tiainen, E. (2003) Increased mortality related to heavy
alcohol intake pattern. Journal of Epidemiology & Com-
munity Health, 57, 379-384. doi:10.1136/jech.57.5.379
[15] Helasoja,V., Lahelma, E., Prättälä, R., Petkeviciene, J.,
Pudule, I. and Tekkel, M. (2007) The sociodemographic
patterning of drinking and binge drinking in Estonia,
Latvia, Lithuania and Finland, 1994-2002. BMC Public
Health, 7, 241. doi:10.1186/1471-2458-7-241
[16] Helakorpi, S., Mäkelä, P. and Uutela, A. (2010) Alcohol
consumption before and after a significant reduction of
alcohol prices in 2004 in Finland: Were the effects dif-
ferent across population subgroups? Alcohol & Alcohol-
ism, 45, 286-292. doi:10.1093/alcalc/agq007
[17] Lai, T. and Habicht, J. (2011) Decline in alcohol con-
sumption in Estonia: Combined effects of strengthened
alcohol policy and economic downturn. Alcohol & Alco-
holism, 46, 200-203. doi:10.1093/alcalc/agr001
[18] Brunovskis, A. and Ugland, T. (2003) Alcohol consump-
tion in the Baltic States. Developments from 1994 to
1999.Institute for Applied Social Science (Fafo), Nor-
way.
[19] Pärna, K., Lang, K., Raju, K., Väli, M. and McKee, M.
(2007) A rapid situation assessment of the market for sur-
rogate and illegal alcohols in Tallinn: Estonia. Interna-
tional Journal of Public Health, 52, 402-404.
doi:10.1007/s00038-007-6112-z
[20] Pärna, K. and Leon, D.A. (2011) Surrogate alcohol drink -
ing in Estonia. Alcoholism: Clinical and Experimental
Research, 35, 1-4.
[21] Ministry of Social Affairs (1997) Alcoholism and drug
abuse prevention programme for 1997-2007.
http://www.tai.ee/failid/Alcoholism_and_drug_abuse_pre
vention_programme_for_1997.pdf
[22] Mäkelä, P. and Österberg, E. (2009) Weakening of one or
more alcohol control pillar: a review of the effects of the
alcohol tax cuts in Finland in 2004. Addiction, 104,
554-563. doi:10.1111/j. 1360-0443.2009.02517.x
[23] Rabinovich, L., Brutscher, P.B., de Vries, H., Tiessen, J.,
Clift, J. and Reding, A. (2009) The affordability of alco-
holic beverages in the European Union: Understanding
the link between alcohol affordability, consumption and
harms. RAND, European Commission, Cambridge.
[24] Chaloupka, F.J., Grossman, M. and Saffer, H. (2002) The
effects of price on alcohol consumption and alcohol-re-
lated problems. Alcohol Research & Health, 26, 22-34.
[25] Lai, T., Habicht, J., Reinap, M., Chisholm, D. and Bal-
tussen, R. (2007) Costs, health effects and cost-effec-
tiveness of alcohol and tobacco control strategies in Es-
tonia. Health Policy, 84, 75-88.
doi:10.1016/j.healthpol.2007.02.012
[26] WHO Regional Office for Europe. (2009) Evidence for
the effectiveness and cost-effectiveness of interventions
to reduce alcohol-related harms. World health Organiza-
tion Regional Office for Europe, Copenhagen.