> Drug and Alcohol Dependence
> Volume 78, Issue 3 , 1 June 2005, Pages 339-344
>
>
>
> This Document
>
>
> SummaryPlus
> Full Text + Links
> ·Full Size Images
> PDF (87 K)
>
>
>
> External Links
>
>
> Abstract + References in Scopus
>
>
>
>
> Actions
>
>
> Cited By
> Save as Citation Alert
> E-mail Article
> Export Citation
>
>
>
> doi:10.1016/j.drugalcdep.2004.12.004
> Copyright © 2004 Elsevier Ireland Ltd All rights reserved.
> Wine preference and related health determinants in a U.S. national sample
of
> young adults
>
> Mallie Paschall, and Robert I. Lipton
>
> Pacific Institute for Research and Evaluation Prevention Research Center,
> 1995 University Ave., Suite 450 Berkeley, CA 94704, USA
>
> Received 4 October 2004; revised 14 December 2004; accepted 20 December
> 2004. Available online 15 January 2005.
>
>
>
>
> Abstract
> This study examined relationships between wine preference and selected
> health determinants in a U.S. national sample of young adults to improve
> understanding of the association between light-moderate wine consumption
and
> long-term morbidity and mortality risk. Interview data collected from
12,958
> young adults who participated in the National Longitudinal Study of
> Adolescent Health were analyzed to determine whether wine preference was
> related to educational, health and lifestyle characteristics that are
> predictive of long-term morbidity and mortality. Wine drinkers generally
had
> more formal education, better dietary and exercise habits, and more
> favorable health status indicators (e.g., normal body mass) than other
> drinkers and non-drinkers. A larger proportion of wine drinkers were
> light-moderate drinkers compared to beer or liquor drinkers, and wine
> drinkers were less likely to report smoking or problem drinking than beer
or
> liquor drinkers. These findings indicate that wine preference in young
> adulthood is related to educational, health and lifestyle characteristics
> that may help to explain the association between light-moderate wine
> consumption and morbidity, and mortality risk in later adulthood.
>
> Keywords: Wine; Alcohol; Light-moderate drinking; Health; Young adulthood
>
>
> Article Outline
> 1. Introduction
> 2. Methods
> 2.1. Sample
> 2.2. Measures
> 2.2.1. Alcoholic beverage preference
> 2.2.2. Educational attainment and verbal ability
> 2.2.3. Subjective health
> 2.2.4. Body mass index
> 2.2.5. Dietary habits
> 2.2.6. Exercise habits
> 2.2.7. Depressive symptoms
> 2.2.8. Smoking history
> 2.2.9. Light-moderate drinking
> 2.2.10. Alcohol-related problems
> 2.2.11. Socio-demographic characteristics
> 2.3. Analysis
> 3. Results
> 3.1. Non-response attrition
> 3.2. Wine preference and related health determinants
> 3.3. Sex differences
> 4. Discussion
> Acknowledgements
> References
>
>
> 1. Introduction
> The J-shaped relationship between level of alcohol use and health is well
> documented, with the lowest morbidity and mortality risk occurring among
> light or moderate drinkers relative to abstainers and heavier drinkers
> (Gunzerath et al., 2004 and Klatsy, 1999). Light-moderate drinking is
> defined in the U.S. as no more than one standard alcoholic drink per day
for
> healthy non-pregnant women and no more than two drinks per day for healthy
> men (U.S. Department of Health and Human Services and U.S. Department of
> Agriculture, 2000). A number of prospective epidemiological studies also
> suggest that light-moderate wine consumption may confer greater health
> benefits than light-moderate consumption of other types of alcoholic
> beverages or abstinence from drinking (GrØnbæk et al., 1995, GrØnbæk et
al.,
> 2000, Klatsky et al., 2003, Prescott et al., 1999 and Truelsen et al.,
> 1998). These studies are supported by evidence for biological mechanisms
> (e.g., elevated levels of high density lipoprotein or "good cholesterol")
> through which wine may reduce risk for coronary disease and other health
> problems (Booyse and Parks, 2001, Fitzpatrick et al., 1993, Frankel et
al.,
> 1993, Hertog et al., 1993, Jang et al., 1997, Maxwell et al., 1994,
> Pace-Asciak et al., 1995 and Rimm, 1999). However, the notion that
> light-moderate wine consumption per se has a beneficial effect on health
> remains controversial due to possible confounding effects of
> socio-demographic, educational, health, and lifestyle factors that may
> affect level of alcohol use, alcoholic beverage preference, and mortality
> risk (GrØnbæk, 2001, GrØnbæk, 2004, Gunzerath et al., 2004 and Klatsy,
> 1999).
>
> In a recent prospective study of over 12,000 northern Californians,
Klatsky
> et al. (2003) found that light-moderate wine drinking was associated with
> the lowest risk for all-cause and coronary disease mortality, though lower
> mortality risks also were observed for light-moderate drinkers of beer and
> spirits relative to lifetime abstainers. The relative risk ratios were
> adjusted for a number of potential confounders, including age, sex, race,
> education, marital status, body mass index (BMI), smoking, and coronary
> disease risk/symptoms. The authors interpreted their results with caution
> however, acknowledging the possibility that other health-related lifestyle
> characteristics (e.g., exercise, diet) not examined in their study may
have
> attenuated the association between light-moderate wine consumption and
> mortality. Their cautionary interpretation is supported by a recent study
of
> University of North Carolina (UNC) alumni, which found that adults who
> preferred wine had healthier diets (more servings of fruit and vegetables,
> fewer servings of red or fried meats, less cholesterol and saturated fat,
> more fiber) than those who preferred beer or spirits or had no alcoholic
> beverage preference (Barefoot et al., 2002). Wine drinkers in the UNC
alumni
> study also were less likely to smoke and more likely to exercise regularly
> than other drinkers and non-drinkers. Additionally, wine drinkers had a
> lower mean body mass index than non-drinkers, but were similar to other
> drinkers in body mass. Several studies based in Europe have identified
other
> health-related factors (e.g., subjective health and well-being,
> intelligence, and social class) that are associated with wine preference
and
> therefore may confound the association between wine consumption and
> morbidity or mortality (GrØnbæk, 2001, GrØnbæk, 2004, GrØnbæk et al., 1999
> and Mortensen et al., 2001).
>
> Although prior studies have identified a number of potential confounding
> factors, most have been limited to Caucasian samples in the U.S. and
> elsewhere. Thus, little is known about whether and to what extent findings
> of prior studies may generalize to a more representative population
sample.
> Prior studies also have typically been based on older adult samples so
that
> conclusions could be drawn about the association between light-moderate
wine
> consumption and mortality risk. However, reliance on older adult samples
> limits our understanding of the developmental pattern of alcoholic
beverage
> preference, level of alcohol use, and related health determinants. Some
> prior studies have also been limited by not distinguishing between
lifetime
> abstainers and ex-drinkers, which may lead to erroneous conclusions about
> the beneficial health effects of light-moderate wine consumption (or
alcohol
> use in general) relative to abstinence. These limitations have been noted
in
> recent review articles (GrØnbæk, 2001, GrØnbæk, 2004, Gunzerath et al.,
2004
> and Peele and Brodsky, 2000), indicating the need for additional research
> with representative samples to better understand the association between
> light-moderate alcohol use in general, and wine consumption in particular,
> and morbidity and mortality risk.
>
> The present study examines relationships between wine and other types of
> alcohol preference (or non-preference), abstinence (lifetime and past
year),
> and long-term health determinants (i.e., socio-demographic, educational,
> health and lifestyle factors) with a national sample of young adults in
the
> U.S. who participated in the National Longitudinal Study of Adolescent
> Health (Add Health)(Harris et al., 2003). The Add Health study is being
> conducted to better understand how social contextual and individual
factors
> influence a variety of health behaviors in a large nationally
representative
> sample, and provides an excellent opportunity to determine whether
> relationships between alcoholic beverage preference, level of alcohol use,
> and other health determinants are observable in young adulthood. Specific
> health determinants examined in this study include socio-demographic
> characteristics (age, sex, race/ethnicity, marital status), educational
> attainment and verbal ability, subjective health, body mass index,
exercise
> and dietary habits, depressive symptoms, smoking history, light-moderate
> alcohol use and problem drinking. These potential confounders (or
> conceptually similar constructs) were identified in previous studies and
> review articles on the association between light-moderate wine consumption
> (or general alcohol use) and morbidity/mortality risk (e.g., Barefoot et
> al., 2002, GrØnbæk, 2001, GrØnbæk, 2004, Gunzerath et al., 2004, Klatsy,
> 1999 and Peele and Brodsky, 2000).
>
> 2. Methods
> 2.1. Sample
> This study uses computer-based in-home interview data collected in 1995
and
> 2002 from individuals who participated in the National Longitudinal Study
of
> Adolescent Health (Add Health) in the U.S. (Udry, 2003). Add Health
> participants (N = 15,197) were first interviewed in 1995 as middle or high
> school students and then in 2002 as young adults. The 1995 Add Health
sample
> was based on a stratified random sample of 134 middle and high schools in
80
> communities in 33 states. Sample weights were available for 14,322 (94%)
of
> the sample, of which 12,958 (90%) provided complete data for all study
> variables. More detail regarding the Add Health study design is provided
> elsewhere (Harris et al., 2003 and Chantala et al., 2003). Permission for
> use of the Add Health restricted public use dataset was obtained through a
> contract with the Add Health Project, which included a data security plan
> that was also approved by the Institutional Review Board of the Pacific
> Institute for Research and Evaluation.
>
> 2.2. Measures
> 2.2.1. Alcoholic beverage preference
> Respondents were asked whether they had one or more alcoholic beverages
> since 1995 when the first wave of Add Health data were collected, and
> whether they had one or more alcoholic beverages during the prior 12
months.
> Respondents who reported drinking at least one alcoholic beverage in the
> prior 12 months were also asked what type of alcoholic beverage they drank
> most often. Possible responses included "beer," "wine," "wine coolers,"
> "hard cider," "straight liquor," "mixed drinks," and "whatever is
> available." Wine coolers and hard cider were collapsed into a single
> category as were straight liquor and mixed drinks. Respondents who did not
> report any alcohol use prior to or after 1995 were classified as lifetime
> abstainers, while respondents who reported any alcohol use after 1995 but
> not in the prior 12 months were classified as ex-drinkers.
>
> 2.2.2. Educational attainment and verbal ability
> Respondents reported the number of years of formal education they had
> completed. They also participated in the Add Health Picture Vocabulary
Test
> (AHPVT), which is a measure of receptive (hearing) vocabulary or verbal
> cognitive ability. The AHPVT included a subset of 87 items from the
Peabody
> Picture Vocabulary Test-Revised (Dunn and Dunn, 1981). All interviewers
> received training in administration of the AHPVT and were required to pass
a
> pronunciation certification test for all words in the AHPVT. An
interviewer
> read a word aloud and the respondent selected the illustration that best
fit
> its meaning. Each word has four simple black-and-white illustrations
> arranged in a multiple-choice format. For example, the word "furry" has
> illustrations of a parrot, dolphin, frog, and cat from which to choose.
Raw
> AHPVT scores were standardized by age (mean = 100, S.D. = 15).
>
> 2.2.3. Subjective health
> Respondents were asked, "In general, how is your health?," with five
> response options: "poor," "fair," "good," "very good," "excellent" that
were
> coded from 1 to 5.
>
> 2.2.4. Body mass index
> Young adults provided information about their height and weight, which was
> used to calculate their body mass index (BMI). Respondents were then
> classified as having a BMI in the normal range (18.5-24.9) or out of the
> normal range.
>
> 2.2.5. Dietary habits
> Respondents were asked, "On how many of the past 7 days did you eat food
> from a fast food place, McDonalds, Kentucky Fried Chicken, Pizza Hut, Taco
> Bell, or a local fast food restaurant?," with responses ranging from 0 to
7.
> Young adults were also asked whether they considered themselves to be
> vegetarians (yes/no).
>
> 2.2.6. Exercise habits
> Young adults were asked how often in the past 7 days they engaged in a
> variety of physical activities such as going to a fitness center, riding a
> bicycle, roller blading or roller skating, and participating in team or
> individual sports (e.g., football, soccer, swimming, gymnastics), with
> response values ranging from 0 to 7. An exercise frequency measure was
> created for each respondent by summing response values.
>
> 2.2.7. Depressive symptoms
> Young adults responded to nine items from the Center for Epidemiological
> Studies-Depression Scale (CES-D), which was developed to assess depressive
> symptomatology in the general population, but is not a measure of clinical
> depression (Radloff, 1977). Respondents were asked how often in the past 7
> days they experienced depressive symptoms such as not being able to shake
> off the blues, being too tired to do things, and feeling depressed. Four
> possible responses to each question were "never or rarely," "sometimes,"
"a
> lot of the time," and "most of the time or all of the time," with
> corresponding values from 0 to 3. Response values were summed to create a
> depressive symptoms score for each respondent.
>
> 2.2.8. Smoking history
> Respondents were asked, "Have you ever smoked cigarettes regulary-that is,
> at least one cigarette every day for at least 30 days?" (yes/no).
>
> 2.2.9. Light-moderate drinking
> Respondents were asked how often in the prior 12 months they drank alcohol
> (none, 1 or 2 days, once a month or less, 2 or 3 days a month, 1 or 2 days
a
> week, 3-5 days a week) and how many drinks they usually had each time.
They
> were also asked how frequently they were drunk in the prior 12 months and
> how often they had consumed five or more consecutive alcoholic drinks in
the
> past 2 weeks. Respondents were classified as light-moderate drinkers if
they
> reported having no more than 1-2 drinks per occasion and did not report
any
> intoxication in the prior 12 months or heavy drinking in past 2 weeks.
>
> 2.2.10. Alcohol-related problems
> Young adults were asked how often in the past 12 months they had
experienced
> eight alcohol-related problems, including having a hangover, vomiting,
> having a regrettable sexual experience, and getting into a physical fight.
> Five possible responses ranged from "never" to "five or more times" and
were
> coded 0-4. An alcohol-related problems frequency measure was created by
> summing response values for each respondent.
>
> 2.2.11. Socio-demographic characteristics
> Young adults reported their age, sex, race/ethnicity, and marital status.
> Race/ethnicity was treated as a dichotomy (non-Hispanic White versus
> non-White) as the majority of respondents (66%) were non-Hispanic white.
>
> 2.3. Analysis
> Preliminary analyses were conducted to compare respondents with and
without
> complete data for all study variables with respect to socio-demographic
> characteristics and alcohol use. Unadjusted bivariate analyses (t-tests
and
> chi-square tests) were then conducted to compare socio-demographic
> characteristics (age, sex, race/ethnicity, marital status) of wine
drinkers
> versus other subgroups of drinkers and non-drinkers. Socio-demographic
> variables were then included as covariates in subsequent analyses of
> covariance comparing levels of health determinants (e.g., educational
> attainment, dietary and exercise habits) among wine drinkers versus other
> subgroups of drinkers and non-drinkers. Separate analyses were initially
> conducted for males and females, but very few sex differences were
observed;
> hence, only detailed analysis results for the full sample are reported.
All
> analyses were conducted with SUDAAN statistical software (Research
Triangle
> Institute, 2002) to accommodate the unequal weighting of the Add Health
> sample and to adjust for clustering effects that are attributable to the
> stratified probability sampling design (Harris et al., 2003 and Chantala
et
> al., 2003). The weights for analyses with 1995 and 2002 data provided in
the
> Add Health restricted public use dataset were applied to all of the
> statistical procedures used for this study.
>
> 3. Results
> 3.1. Non-response attrition
> Preliminary analyses indicated that respondents with and without complete
> data for all study variables were similar with respect to age, sex and
> marital status. However, respondents with incomplete data were more likely
> to be non-White and non-drinkers than respondents with complete data. Only
> respondents with complete data for all study variables (N = 12,958) were
> included in the analyses so that all results would be based on the same
> group of respondents.
>
> 3.2. Wine preference and related health determinants
> Results of analyses comparing young adults who preferred wine and other
> subgroups of drinkers and non-drinkers are provided in Table 1. Unadjusted
> socio-demographic comparisons at the top of Table 1 indicate that wine
> drinkers were older and more likely to be female than all of the other
> subgroups. Wine drinkers also differed from at least some of the other
> subgroups in their race/ethnicity and martial status. Age, sex, race, and
> marital status were therefore included as covariates in subsequent
analyses.
>
> Table 1.
>
> Sample characteristics and results of analyses to examine relationships
> between wine preference and selected health determinantsa Variable Total
> sample (N = 12,958) Alcoholic beverage preference Non-drinkers
> Beer (n = 4348) Wine (n = 584) Wine cooler/hard cider (n = 847)
> Liquor/mixed drinks (n = 3407) No preference (n = 347) Ex-drinker (n =
1601)
> Lifetime abstainer (n = 1824)
> Age, mean (S.D.) 21.8 (1.9) 21.9 (1.8)* 22.5 (1.7) 21.4 (1.9)*
21.8
> (1.8)* 20.9 (1.7)* 22.0 (1.9)* 21.5 (2.0)*
> Male (%) 51.2 69.3* 23.1 19.2 39.1* 55.4* 47.9* 48.3
> White (%) 65.8 75.6 71.2 63.7b 65.7 71.3 52.5* 49.1*
> Married (%) 16.5 13.0* 18.0 21.4 16.0 3.6* 25.5* 19.0
>
>
> Adjusted for age, sex, race, and marital status
> Education, mean (S.D.) 13.1 (2.0) 13.2 (2.0)* 14.0 (2.0) 13.0
> (1.8)* 13.3 (2.0)* 13.4 (1.8)* 12.5 (1.9)* 12.8 (1.8)*
> Vocabulary test score, mean (S.D.) 100.1 (15.0) 100.9 (12.5)*
105.3
> (12.0) 99.7 (13.6)* 102.7 (12.9)* 105.1 (12.4) 95.8 (17.6)* 94.5 (19.4)*
> Subjective health, mean (S.D.) 4.0 (0.9) 4.0 (0.8) 4.1 (0.9) 3.9
> (0.9)b 3.9 (0.9)b 4.0 (0.9) 3.9 (0.9)b 4.1 (0.9)
> Normal body mass (%) 48.6 49.9* 58.0 42.5* 46.7* 59.6 47.6* 46.9*
> Fast food, mean (S.D.) 2.4 (2.1) 2.4 (2.1)* 1.8 (1.7) 2.4 (1.8)*
> 2.5 (2.1)* 2.3 (2.0)b 2.6 (2.1)* 2.5 (2.1)*
> Vegetarian (%) 2.8 2.8 4.5 2.9 2.4 3.8 2.3 3.0
> Exercise, mean (S.D.) 7.0 (7.1) 7.0 (7.3)* 8.3 (7.1) 6.6 (5.9)*
7.0
> (6.9)* 7.2 (7.5) 6.4 (7.1)* 6.9 (7.5)*
> Depressive symptoms, mean (S.D.) 4.4 (4.0) 4.3 (3.7) 4.3 (4.0)
4.6
> (4.0) 4.5 (4.1) 5.4 (4.3)b 4.8 (4.6) 4.1 (3.9)
> Ever a regular smoker (%) 43.1 53.7* 35.6 32.1 43.9* 43.9 42.6b
> 19.9*
> Light-moderate drinker (%) 10.4 8.5* 32.3 39.3 14.8* 7.6* - -
> Alcohol-related problems, mean (S.D.) 2.7 (3.8) 4.2 (4.3)* 2.4
> (2.9) 1.6 (2.7)* 3.1 (3.7)* 4.1 (4.7)* - -
> a Wine drinkers are the referent group for all significance tests.
> b p < 0.05.
> * p < 0.01.
>
>
>
> Wine drinkers had a significantly higher mean level of formal education
than
> young adults in all of the other subgroups, and had a significantly higher
> mean picture vocabulary test score than all other subgroups except
drinkers
> who had no alcoholic beverage preference. The mean level of subjective
> health was significantly higher among wine drinkers compared to young
adults
> who preferred wine coolers/hard cider or liquor/mixed drinks and
> ex-drinkers. A larger percentage of wine drinkers had a normal body mass
> index than all other subgroups except drinkers with no alcoholic beverage
> preference. Wine drinkers had a significantly lower frequency of fast food
> consumption than all of the other subgroups. A larger percentage of wine
> drinkers considered themselves to be vegetarians compared to all of the
> other subgroups, but none of the observed differences were statistically
> significant. Exercise frequency was significantly higher among wine
drinkers
> compared to all of the other subgroups except young adults with no
alcoholic
> beverage preference. Although depressive symptoms were lower among wine
> drinkers compared to all other subgroups except lifetime abstainers, only
> drinkers with no alcoholic beverage preference had a significantly higher
> level of depressive symptoms. A smaller percentage of wine drinkers had
ever
> been regular smokers compared to young adults who preferred beer or
> liquor/mixed drinks and ex-drinkers. A larger percentage of wine drinkers
> were light-moderate drinkers compared to young adults who preferred beer
or
> liquor/mixed drinks and drinkers who had no alcoholic beverage preference.
> Wine drinkers also reported fewer alcohol-related problems in the past
year
> compared to young adults who preferred beer or liquor/mixed drinks and
> drinkers who had no alcoholic beverage preference. However, the number of
> alcohol-related problems among wine drinkers was significantly higher than
> the number of problems reported by young adults who preferred wine
> coolers/hard cider.
>
> 3.3. Sex differences
> Analysis results were generally very similar for males and females, but
two
> exceptions are noteworthy. The percentage of female wine drinkers with a
> normal body mass index (68%) was significantly higher than all of the
other
> subgroups, while those differences were less prominent and consistent for
> males. Female wine drinkers reported significantly fewer depressive
symptoms
> than all of the other subgroups, while male wine drinkers were similar or
> somewhat higher than other subgroups in depressive symptoms. Thus,
observed
> differences in normal body mass and depressive symptoms for the entire Add
> Health young adult sample are primarily driven by significant differences
in
> these health indicators among female wine drinkers versus other subgroups.
>
> 4. Discussion
> This study provides further evidence that light-moderate wine consumption
is
> more likely among adults who have more formal education, better health
> status indicators, and healthier lifestyle habits than adults who prefer
> other types of alcoholic beverages and non-drinkers. Finding this pattern
of
> relationships between wine preference and various health determinants in a
> young adult sample raises more questions about prospective studies that
have
> found beneficial effects of light-moderate wine consumption on long-term
> health and life expectancy, given the importance of other health
> determinants (e.g., education, diet, exercise, body mass, smoking)
examined
> in this study (Klatsky et al., 2003, GrØnbæk, 2001, Gunzerath et al., 2004
> and Mortensen et al., 2001). Our findings are compelling because such a
> large proportion of healthy young adults prefer other types of alcoholic
> beverages (e.g., beer) and engage in occasional heavy drinking-a
phenomenon
> that is most prevalent among college students (Windle, 2003). Thus, it
would
> not have been surprising to find no differences in educational, health and
> lifestyle characteristics between young adults who preferred wine versus
> those who preferred other types of alcoholic beverages and non-drinkers.
>
> Our reliance on cross-sectional data for most of the analyses leaves to
> question the temporal ordering of relationships between alcoholic beverage
> preference and other health determinants examined in this study. However,
> additional analyses conducted with 1995 measures of family SES (parent
> education, household income), academic achievement, college intentions,
and
> subjective health also indicated that wine preference in young adulthood
was
> positively associated with higher levels of these potential long-term
health
> determinants. These results suggest that light-moderate wine consumption
in
> adulthood is influenced by social class, academic achievement, and good
> health habits that are established in childhood and adolescence. It is
also
> possible that parents' wine preference has an influence on alcoholic
> beverage preference among their children, but this could not be determined
> with Add Health data.
>
> Study findings should be considered in light of several other potential
> limitations. Non-response attrition may have influenced analysis results
in
> unknown ways as respondents who did not provide complete data for all
study
> variables were more likely to be non-White and non-drinkers than
respondents
> who provided complete data. Self-report data are subject to social
> desirability and recall bias, which also may have influenced our results
in
> unknown ways. Finally, there may be other long-term health determinants
> (e.g., medical conditions) not examined in this study that are associated
> with wine preference and may therefore help to explain the association
> between light-moderate wine consumption and mortality in later adulthood.
>
> In conclusion, this study points to the importance of socio-demographic,
> educational, health and lifestyle factors that may act as confounders of
the
> relationship between light-moderate wine consumption, morbidity and
> mortality. Future studies on the health effects of light-moderate alcohol
> use and wine consumption should include a more comprehensive set of
> potential confounders such as those identified in this study to determine
> whether these types of alcohol use actually have beneficial long-term
health
> effects.
>
>
> Acknowledgements
>
> This study was supported by a grant from the National Institute on Alcohol
>
>
>
> Volume 78, Issue 3 , 1 June 2005, Pages 339-344
>
>
>
> This Document
>
>
> SummaryPlus
> Full Text + Links
> ·Full Size Images
> PDF (87 K)
>
>
>
> External Links
>
>
> Abstract + References in Scopus
>
>
>
>
> Actions
>
>
> Cited By
> Save as Citation Alert
> E-mail Article
> Export Citation
>
>
>
> doi:10.1016/j.drugalcdep.2004.12.004
> Copyright © 2004 Elsevier Ireland Ltd All rights reserved.
> Wine preference and related health determinants in a U.S. national sample
of
> young adults
>
> Mallie Paschall, and Robert I. Lipton
>
> Pacific Institute for Research and Evaluation Prevention Research Center,
> 1995 University Ave., Suite 450 Berkeley, CA 94704, USA
>
> Received 4 October 2004; revised 14 December 2004; accepted 20 December
> 2004. Available online 15 January 2005.
>
>
>
>
> Abstract
> This study examined relationships between wine preference and selected
> health determinants in a U.S. national sample of young adults to improve
> understanding of the association between light-moderate wine consumption
and
> long-term morbidity and mortality risk. Interview data collected from
12,958
> young adults who participated in the National Longitudinal Study of
> Adolescent Health were analyzed to determine whether wine preference was
> related to educational, health and lifestyle characteristics that are
> predictive of long-term morbidity and mortality. Wine drinkers generally
had
> more formal education, better dietary and exercise habits, and more
> favorable health status indicators (e.g., normal body mass) than other
> drinkers and non-drinkers. A larger proportion of wine drinkers were
> light-moderate drinkers compared to beer or liquor drinkers, and wine
> drinkers were less likely to report smoking or problem drinking than beer
or
> liquor drinkers. These findings indicate that wine preference in young
> adulthood is related to educational, health and lifestyle characteristics
> that may help to explain the association between light-moderate wine
> consumption and morbidity, and mortality risk in later adulthood.
>
> Keywords: Wine; Alcohol; Light-moderate drinking; Health; Young adulthood
>
>
> Article Outline
> 1. Introduction
> 2. Methods
> 2.1. Sample
> 2.2. Measures
> 2.2.1. Alcoholic beverage preference
> 2.2.2. Educational attainment and verbal ability
> 2.2.3. Subjective health
> 2.2.4. Body mass index
> 2.2.5. Dietary habits
> 2.2.6. Exercise habits
> 2.2.7. Depressive symptoms
> 2.2.8. Smoking history
> 2.2.9. Light-moderate drinking
> 2.2.10. Alcohol-related problems
> 2.2.11. Socio-demographic characteristics
> 2.3. Analysis
> 3. Results
> 3.1. Non-response attrition
> 3.2. Wine preference and related health determinants
> 3.3. Sex differences
> 4. Discussion
> Acknowledgements
> References
>
>
> 1. Introduction
> The J-shaped relationship between level of alcohol use and health is well
> documented, with the lowest morbidity and mortality risk occurring among
> light or moderate drinkers relative to abstainers and heavier drinkers
> (Gunzerath et al., 2004 and Klatsy, 1999). Light-moderate drinking is
> defined in the U.S. as no more than one standard alcoholic drink per day
for
> healthy non-pregnant women and no more than two drinks per day for healthy
> men (U.S. Department of Health and Human Services and U.S. Department of
> Agriculture, 2000). A number of prospective epidemiological studies also
> suggest that light-moderate wine consumption may confer greater health
> benefits than light-moderate consumption of other types of alcoholic
> beverages or abstinence from drinking (GrØnbæk et al., 1995, GrØnbæk et
al.,
> 2000, Klatsky et al., 2003, Prescott et al., 1999 and Truelsen et al.,
> 1998). These studies are supported by evidence for biological mechanisms
> (e.g., elevated levels of high density lipoprotein or "good cholesterol")
> through which wine may reduce risk for coronary disease and other health
> problems (Booyse and Parks, 2001, Fitzpatrick et al., 1993, Frankel et
al.,
> 1993, Hertog et al., 1993, Jang et al., 1997, Maxwell et al., 1994,
> Pace-Asciak et al., 1995 and Rimm, 1999). However, the notion that
> light-moderate wine consumption per se has a beneficial effect on health
> remains controversial due to possible confounding effects of
> socio-demographic, educational, health, and lifestyle factors that may
> affect level of alcohol use, alcoholic beverage preference, and mortality
> risk (GrØnbæk, 2001, GrØnbæk, 2004, Gunzerath et al., 2004 and Klatsy,
> 1999).
>
> In a recent prospective study of over 12,000 northern Californians,
Klatsky
> et al. (2003) found that light-moderate wine drinking was associated with
> the lowest risk for all-cause and coronary disease mortality, though lower
> mortality risks also were observed for light-moderate drinkers of beer and
> spirits relative to lifetime abstainers. The relative risk ratios were
> adjusted for a number of potential confounders, including age, sex, race,
> education, marital status, body mass index (BMI), smoking, and coronary
> disease risk/symptoms. The authors interpreted their results with caution
> however, acknowledging the possibility that other health-related lifestyle
> characteristics (e.g., exercise, diet) not examined in their study may
have
> attenuated the association between light-moderate wine consumption and
> mortality. Their cautionary interpretation is supported by a recent study
of
> University of North Carolina (UNC) alumni, which found that adults who
> preferred wine had healthier diets (more servings of fruit and vegetables,
> fewer servings of red or fried meats, less cholesterol and saturated fat,
> more fiber) than those who preferred beer or spirits or had no alcoholic
> beverage preference (Barefoot et al., 2002). Wine drinkers in the UNC
alumni
> study also were less likely to smoke and more likely to exercise regularly
> than other drinkers and non-drinkers. Additionally, wine drinkers had a
> lower mean body mass index than non-drinkers, but were similar to other
> drinkers in body mass. Several studies based in Europe have identified
other
> health-related factors (e.g., subjective health and well-being,
> intelligence, and social class) that are associated with wine preference
and
> therefore may confound the association between wine consumption and
> morbidity or mortality (GrØnbæk, 2001, GrØnbæk, 2004, GrØnbæk et al., 1999
> and Mortensen et al., 2001).
>
> Although prior studies have identified a number of potential confounding
> factors, most have been limited to Caucasian samples in the U.S. and
> elsewhere. Thus, little is known about whether and to what extent findings
> of prior studies may generalize to a more representative population
sample.
> Prior studies also have typically been based on older adult samples so
that
> conclusions could be drawn about the association between light-moderate
wine
> consumption and mortality risk. However, reliance on older adult samples
> limits our understanding of the developmental pattern of alcoholic
beverage
> preference, level of alcohol use, and related health determinants. Some
> prior studies have also been limited by not distinguishing between
lifetime
> abstainers and ex-drinkers, which may lead to erroneous conclusions about
> the beneficial health effects of light-moderate wine consumption (or
alcohol
> use in general) relative to abstinence. These limitations have been noted
in
> recent review articles (GrØnbæk, 2001, GrØnbæk, 2004, Gunzerath et al.,
2004
> and Peele and Brodsky, 2000), indicating the need for additional research
> with representative samples to better understand the association between
> light-moderate alcohol use in general, and wine consumption in particular,
> and morbidity and mortality risk.
>
> The present study examines relationships between wine and other types of
> alcohol preference (or non-preference), abstinence (lifetime and past
year),
> and long-term health determinants (i.e., socio-demographic, educational,
> health and lifestyle factors) with a national sample of young adults in
the
> U.S. who participated in the National Longitudinal Study of Adolescent
> Health (Add Health)(Harris et al., 2003). The Add Health study is being
> conducted to better understand how social contextual and individual
factors
> influence a variety of health behaviors in a large nationally
representative
> sample, and provides an excellent opportunity to determine whether
> relationships between alcoholic beverage preference, level of alcohol use,
> and other health determinants are observable in young adulthood. Specific
> health determinants examined in this study include socio-demographic
> characteristics (age, sex, race/ethnicity, marital status), educational
> attainment and verbal ability, subjective health, body mass index,
exercise
> and dietary habits, depressive symptoms, smoking history, light-moderate
> alcohol use and problem drinking. These potential confounders (or
> conceptually similar constructs) were identified in previous studies and
> review articles on the association between light-moderate wine consumption
> (or general alcohol use) and morbidity/mortality risk (e.g., Barefoot et
> al., 2002, GrØnbæk, 2001, GrØnbæk, 2004, Gunzerath et al., 2004, Klatsy,
> 1999 and Peele and Brodsky, 2000).
>
> 2. Methods
> 2.1. Sample
> This study uses computer-based in-home interview data collected in 1995
and
> 2002 from individuals who participated in the National Longitudinal Study
of
> Adolescent Health (Add Health) in the U.S. (Udry, 2003). Add Health
> participants (N = 15,197) were first interviewed in 1995 as middle or high
> school students and then in 2002 as young adults. The 1995 Add Health
sample
> was based on a stratified random sample of 134 middle and high schools in
80
> communities in 33 states. Sample weights were available for 14,322 (94%)
of
> the sample, of which 12,958 (90%) provided complete data for all study
> variables. More detail regarding the Add Health study design is provided
> elsewhere (Harris et al., 2003 and Chantala et al., 2003). Permission for
> use of the Add Health restricted public use dataset was obtained through a
> contract with the Add Health Project, which included a data security plan
> that was also approved by the Institutional Review Board of the Pacific
> Institute for Research and Evaluation.
>
> 2.2. Measures
> 2.2.1. Alcoholic beverage preference
> Respondents were asked whether they had one or more alcoholic beverages
> since 1995 when the first wave of Add Health data were collected, and
> whether they had one or more alcoholic beverages during the prior 12
months.
> Respondents who reported drinking at least one alcoholic beverage in the
> prior 12 months were also asked what type of alcoholic beverage they drank
> most often. Possible responses included "beer," "wine," "wine coolers,"
> "hard cider," "straight liquor," "mixed drinks," and "whatever is
> available." Wine coolers and hard cider were collapsed into a single
> category as were straight liquor and mixed drinks. Respondents who did not
> report any alcohol use prior to or after 1995 were classified as lifetime
> abstainers, while respondents who reported any alcohol use after 1995 but
> not in the prior 12 months were classified as ex-drinkers.
>
> 2.2.2. Educational attainment and verbal ability
> Respondents reported the number of years of formal education they had
> completed. They also participated in the Add Health Picture Vocabulary
Test
> (AHPVT), which is a measure of receptive (hearing) vocabulary or verbal
> cognitive ability. The AHPVT included a subset of 87 items from the
Peabody
> Picture Vocabulary Test-Revised (Dunn and Dunn, 1981). All interviewers
> received training in administration of the AHPVT and were required to pass
a
> pronunciation certification test for all words in the AHPVT. An
interviewer
> read a word aloud and the respondent selected the illustration that best
fit
> its meaning. Each word has four simple black-and-white illustrations
> arranged in a multiple-choice format. For example, the word "furry" has
> illustrations of a parrot, dolphin, frog, and cat from which to choose.
Raw
> AHPVT scores were standardized by age (mean = 100, S.D. = 15).
>
> 2.2.3. Subjective health
> Respondents were asked, "In general, how is your health?," with five
> response options: "poor," "fair," "good," "very good," "excellent" that
were
> coded from 1 to 5.
>
> 2.2.4. Body mass index
> Young adults provided information about their height and weight, which was
> used to calculate their body mass index (BMI). Respondents were then
> classified as having a BMI in the normal range (18.5-24.9) or out of the
> normal range.
>
> 2.2.5. Dietary habits
> Respondents were asked, "On how many of the past 7 days did you eat food
> from a fast food place, McDonalds, Kentucky Fried Chicken, Pizza Hut, Taco
> Bell, or a local fast food restaurant?," with responses ranging from 0 to
7.
> Young adults were also asked whether they considered themselves to be
> vegetarians (yes/no).
>
> 2.2.6. Exercise habits
> Young adults were asked how often in the past 7 days they engaged in a
> variety of physical activities such as going to a fitness center, riding a
> bicycle, roller blading or roller skating, and participating in team or
> individual sports (e.g., football, soccer, swimming, gymnastics), with
> response values ranging from 0 to 7. An exercise frequency measure was
> created for each respondent by summing response values.
>
> 2.2.7. Depressive symptoms
> Young adults responded to nine items from the Center for Epidemiological
> Studies-Depression Scale (CES-D), which was developed to assess depressive
> symptomatology in the general population, but is not a measure of clinical
> depression (Radloff, 1977). Respondents were asked how often in the past 7
> days they experienced depressive symptoms such as not being able to shake
> off the blues, being too tired to do things, and feeling depressed. Four
> possible responses to each question were "never or rarely," "sometimes,"
"a
> lot of the time," and "most of the time or all of the time," with
> corresponding values from 0 to 3. Response values were summed to create a
> depressive symptoms score for each respondent.
>
> 2.2.8. Smoking history
> Respondents were asked, "Have you ever smoked cigarettes regulary-that is,
> at least one cigarette every day for at least 30 days?" (yes/no).
>
> 2.2.9. Light-moderate drinking
> Respondents were asked how often in the prior 12 months they drank alcohol
> (none, 1 or 2 days, once a month or less, 2 or 3 days a month, 1 or 2 days
a
> week, 3-5 days a week) and how many drinks they usually had each time.
They
> were also asked how frequently they were drunk in the prior 12 months and
> how often they had consumed five or more consecutive alcoholic drinks in
the
> past 2 weeks. Respondents were classified as light-moderate drinkers if
they
> reported having no more than 1-2 drinks per occasion and did not report
any
> intoxication in the prior 12 months or heavy drinking in past 2 weeks.
>
> 2.2.10. Alcohol-related problems
> Young adults were asked how often in the past 12 months they had
experienced
> eight alcohol-related problems, including having a hangover, vomiting,
> having a regrettable sexual experience, and getting into a physical fight.
> Five possible responses ranged from "never" to "five or more times" and
were
> coded 0-4. An alcohol-related problems frequency measure was created by
> summing response values for each respondent.
>
> 2.2.11. Socio-demographic characteristics
> Young adults reported their age, sex, race/ethnicity, and marital status.
> Race/ethnicity was treated as a dichotomy (non-Hispanic White versus
> non-White) as the majority of respondents (66%) were non-Hispanic white.
>
> 2.3. Analysis
> Preliminary analyses were conducted to compare respondents with and
without
> complete data for all study variables with respect to socio-demographic
> characteristics and alcohol use. Unadjusted bivariate analyses (t-tests
and
> chi-square tests) were then conducted to compare socio-demographic
> characteristics (age, sex, race/ethnicity, marital status) of wine
drinkers
> versus other subgroups of drinkers and non-drinkers. Socio-demographic
> variables were then included as covariates in subsequent analyses of
> covariance comparing levels of health determinants (e.g., educational
> attainment, dietary and exercise habits) among wine drinkers versus other
> subgroups of drinkers and non-drinkers. Separate analyses were initially
> conducted for males and females, but very few sex differences were
observed;
> hence, only detailed analysis results for the full sample are reported.
All
> analyses were conducted with SUDAAN statistical software (Research
Triangle
> Institute, 2002) to accommodate the unequal weighting of the Add Health
> sample and to adjust for clustering effects that are attributable to the
> stratified probability sampling design (Harris et al., 2003 and Chantala
et
> al., 2003). The weights for analyses with 1995 and 2002 data provided in
the
> Add Health restricted public use dataset were applied to all of the
> statistical procedures used for this study.
>
> 3. Results
> 3.1. Non-response attrition
> Preliminary analyses indicated that respondents with and without complete
> data for all study variables were similar with respect to age, sex and
> marital status. However, respondents with incomplete data were more likely
> to be non-White and non-drinkers than respondents with complete data. Only
> respondents with complete data for all study variables (N = 12,958) were
> included in the analyses so that all results would be based on the same
> group of respondents.
>
> 3.2. Wine preference and related health determinants
> Results of analyses comparing young adults who preferred wine and other
> subgroups of drinkers and non-drinkers are provided in Table 1. Unadjusted
> socio-demographic comparisons at the top of Table 1 indicate that wine
> drinkers were older and more likely to be female than all of the other
> subgroups. Wine drinkers also differed from at least some of the other
> subgroups in their race/ethnicity and martial status. Age, sex, race, and
> marital status were therefore included as covariates in subsequent
analyses.
>
> Table 1.
>
> Sample characteristics and results of analyses to examine relationships
> between wine preference and selected health determinantsa Variable Total
> sample (N = 12,958) Alcoholic beverage preference Non-drinkers
> Beer (n = 4348) Wine (n = 584) Wine cooler/hard cider (n = 847)
> Liquor/mixed drinks (n = 3407) No preference (n = 347) Ex-drinker (n =
1601)
> Lifetime abstainer (n = 1824)
> Age, mean (S.D.) 21.8 (1.9) 21.9 (1.8)* 22.5 (1.7) 21.4 (1.9)*
21.8
> (1.8)* 20.9 (1.7)* 22.0 (1.9)* 21.5 (2.0)*
> Male (%) 51.2 69.3* 23.1 19.2 39.1* 55.4* 47.9* 48.3
> White (%) 65.8 75.6 71.2 63.7b 65.7 71.3 52.5* 49.1*
> Married (%) 16.5 13.0* 18.0 21.4 16.0 3.6* 25.5* 19.0
>
>
> Adjusted for age, sex, race, and marital status
> Education, mean (S.D.) 13.1 (2.0) 13.2 (2.0)* 14.0 (2.0) 13.0
> (1.8)* 13.3 (2.0)* 13.4 (1.8)* 12.5 (1.9)* 12.8 (1.8)*
> Vocabulary test score, mean (S.D.) 100.1 (15.0) 100.9 (12.5)*
105.3
> (12.0) 99.7 (13.6)* 102.7 (12.9)* 105.1 (12.4) 95.8 (17.6)* 94.5 (19.4)*
> Subjective health, mean (S.D.) 4.0 (0.9) 4.0 (0.8) 4.1 (0.9) 3.9
> (0.9)b 3.9 (0.9)b 4.0 (0.9) 3.9 (0.9)b 4.1 (0.9)
> Normal body mass (%) 48.6 49.9* 58.0 42.5* 46.7* 59.6 47.6* 46.9*
> Fast food, mean (S.D.) 2.4 (2.1) 2.4 (2.1)* 1.8 (1.7) 2.4 (1.8)*
> 2.5 (2.1)* 2.3 (2.0)b 2.6 (2.1)* 2.5 (2.1)*
> Vegetarian (%) 2.8 2.8 4.5 2.9 2.4 3.8 2.3 3.0
> Exercise, mean (S.D.) 7.0 (7.1) 7.0 (7.3)* 8.3 (7.1) 6.6 (5.9)*
7.0
> (6.9)* 7.2 (7.5) 6.4 (7.1)* 6.9 (7.5)*
> Depressive symptoms, mean (S.D.) 4.4 (4.0) 4.3 (3.7) 4.3 (4.0)
4.6
> (4.0) 4.5 (4.1) 5.4 (4.3)b 4.8 (4.6) 4.1 (3.9)
> Ever a regular smoker (%) 43.1 53.7* 35.6 32.1 43.9* 43.9 42.6b
> 19.9*
> Light-moderate drinker (%) 10.4 8.5* 32.3 39.3 14.8* 7.6* - -
> Alcohol-related problems, mean (S.D.) 2.7 (3.8) 4.2 (4.3)* 2.4
> (2.9) 1.6 (2.7)* 3.1 (3.7)* 4.1 (4.7)* - -
> a Wine drinkers are the referent group for all significance tests.
> b p < 0.05.
> * p < 0.01.
>
>
>
> Wine drinkers had a significantly higher mean level of formal education
than
> young adults in all of the other subgroups, and had a significantly higher
> mean picture vocabulary test score than all other subgroups except
drinkers
> who had no alcoholic beverage preference. The mean level of subjective
> health was significantly higher among wine drinkers compared to young
adults
> who preferred wine coolers/hard cider or liquor/mixed drinks and
> ex-drinkers. A larger percentage of wine drinkers had a normal body mass
> index than all other subgroups except drinkers with no alcoholic beverage
> preference. Wine drinkers had a significantly lower frequency of fast food
> consumption than all of the other subgroups. A larger percentage of wine
> drinkers considered themselves to be vegetarians compared to all of the
> other subgroups, but none of the observed differences were statistically
> significant. Exercise frequency was significantly higher among wine
drinkers
> compared to all of the other subgroups except young adults with no
alcoholic
> beverage preference. Although depressive symptoms were lower among wine
> drinkers compared to all other subgroups except lifetime abstainers, only
> drinkers with no alcoholic beverage preference had a significantly higher
> level of depressive symptoms. A smaller percentage of wine drinkers had
ever
> been regular smokers compared to young adults who preferred beer or
> liquor/mixed drinks and ex-drinkers. A larger percentage of wine drinkers
> were light-moderate drinkers compared to young adults who preferred beer
or
> liquor/mixed drinks and drinkers who had no alcoholic beverage preference.
> Wine drinkers also reported fewer alcohol-related problems in the past
year
> compared to young adults who preferred beer or liquor/mixed drinks and
> drinkers who had no alcoholic beverage preference. However, the number of
> alcohol-related problems among wine drinkers was significantly higher than
> the number of problems reported by young adults who preferred wine
> coolers/hard cider.
>
> 3.3. Sex differences
> Analysis results were generally very similar for males and females, but
two
> exceptions are noteworthy. The percentage of female wine drinkers with a
> normal body mass index (68%) was significantly higher than all of the
other
> subgroups, while those differences were less prominent and consistent for
> males. Female wine drinkers reported significantly fewer depressive
symptoms
> than all of the other subgroups, while male wine drinkers were similar or
> somewhat higher than other subgroups in depressive symptoms. Thus,
observed
> differences in normal body mass and depressive symptoms for the entire Add
> Health young adult sample are primarily driven by significant differences
in
> these health indicators among female wine drinkers versus other subgroups.
>
> 4. Discussion
> This study provides further evidence that light-moderate wine consumption
is
> more likely among adults who have more formal education, better health
> status indicators, and healthier lifestyle habits than adults who prefer
> other types of alcoholic beverages and non-drinkers. Finding this pattern
of
> relationships between wine preference and various health determinants in a
> young adult sample raises more questions about prospective studies that
have
> found beneficial effects of light-moderate wine consumption on long-term
> health and life expectancy, given the importance of other health
> determinants (e.g., education, diet, exercise, body mass, smoking)
examined
> in this study (Klatsky et al., 2003, GrØnbæk, 2001, Gunzerath et al., 2004
> and Mortensen et al., 2001). Our findings are compelling because such a
> large proportion of healthy young adults prefer other types of alcoholic
> beverages (e.g., beer) and engage in occasional heavy drinking-a
phenomenon
> that is most prevalent among college students (Windle, 2003). Thus, it
would
> not have been surprising to find no differences in educational, health and
> lifestyle characteristics between young adults who preferred wine versus
> those who preferred other types of alcoholic beverages and non-drinkers.
>
> Our reliance on cross-sectional data for most of the analyses leaves to
> question the temporal ordering of relationships between alcoholic beverage
> preference and other health determinants examined in this study. However,
> additional analyses conducted with 1995 measures of family SES (parent
> education, household income), academic achievement, college intentions,
and
> subjective health also indicated that wine preference in young adulthood
was
> positively associated with higher levels of these potential long-term
health
> determinants. These results suggest that light-moderate wine consumption
in
> adulthood is influenced by social class, academic achievement, and good
> health habits that are established in childhood and adolescence. It is
also
> possible that parents' wine preference has an influence on alcoholic
> beverage preference among their children, but this could not be determined
> with Add Health data.
>
> Study findings should be considered in light of several other potential
> limitations. Non-response attrition may have influenced analysis results
in
> unknown ways as respondents who did not provide complete data for all
study
> variables were more likely to be non-White and non-drinkers than
respondents
> who provided complete data. Self-report data are subject to social
> desirability and recall bias, which also may have influenced our results
in
> unknown ways. Finally, there may be other long-term health determinants
> (e.g., medical conditions) not examined in this study that are associated
> with wine preference and may therefore help to explain the association
> between light-moderate wine consumption and mortality in later adulthood.
>
> In conclusion, this study points to the importance of socio-demographic,
> educational, health and lifestyle factors that may act as confounders of
the
> relationship between light-moderate wine consumption, morbidity and
> mortality. Future studies on the health effects of light-moderate alcohol
> use and wine consumption should include a more comprehensive set of
> potential confounders such as those identified in this study to determine
> whether these types of alcohol use actually have beneficial long-term
health
> effects.
>
>
> Acknowledgements
>
> This study was supported by a grant from the National Institute on Alcohol
>
>
>
