ArticlePDF Available

Effects of Initiating Moderate Alcohol Intake on Cardiometabolic Risk in Adults With Type 2 Diabetes: A 2-Year Randomized, Controlled Trial

Authors:

Abstract and Figures

Background: Recommendations for moderate alcohol consumption remain controversial, particularly in type 2 diabetes mellitus (T2DM). Long-term randomized, controlled trials (RCTs) are lacking. Objective: To assess cardiometabolic effects of initiating moderate alcohol intake in persons with T2DM and whether the type of wine matters. Design: 2-year RCT (CASCADE [CArdiovaSCulAr Diabetes & Ethanol] trial). (ClinicalTrials.gov: NCT00784433). Setting: Ben-Gurion University of the Negev-Soroka Medical Center and Nuclear Research Center Negev, Israel. Patients: Alcohol-abstaining adults with well-controlled T2DM. Intervention: Patients were randomly assigned to 150 mL of mineral water, white wine, or red wine with dinner for 2 years. Wines and mineral water were provided. All groups followed a Mediterranean diet without caloric restriction. Measurements: Primary outcomes were lipid and glycemic control profiles. Genetic measurements were done, and patients were followed for blood pressure, liver biomarkers, medication use, symptoms, and quality of life. Results: Of the 224 patients who were randomly assigned, 94% had follow-up data at 1 year and 87% at 2 years. In addition to the changes in the water group (Mediterranean diet only), red wine significantly increased high-density lipoprotein cholesterol (HDL-C) level by 0.05 mmol/L (2.0 mg/dL) (95% CI, 0.04 to 0.06 mmol/L [1.6 to 2.2 mg/dL]; P < 0.001) and apolipoprotein(a)1 level by 0.03 g/L (CI, 0.01 to 0.06 g/L; P = 0.05) and decreased the total cholesterol-HDL-C ratio by 0.27 (CI, -0.52 to -0.01; P = 0.039). Only slow ethanol metabolizers (alcohol dehydrogenase alleles [ADH1B*1] carriers) significantly benefited from the effect of both wines on glycemic control (fasting plasma glucose, homeostatic model assessment of insulin resistance, and hemoglobin A1c) compared with fast ethanol metabolizers (persons homozygous for ADH1B*2). Across the 3 groups, no material differences were identified in blood pressure, adiposity, liver function, drug therapy, symptoms, or quality of life, except that sleep quality improved in both wine groups compared with the water group (P = 0.040). Overall, compared with the changes in the water group, red wine further reduced the number of components of the metabolic syndrome by 0.34 (CI, -0.68 to -0.001; P = 0.049). Limitation: Participants were not blinded to treatment allocation. Conclusion: This long-term RCT suggests that initiating moderate wine intake, especially red wine, among well-controlled diabetics as part of a healthy diet is apparently safe and modestly decreases cardiometabolic risk. The genetic interactions suggest that ethanol plays an important role in glucose metabolism, and red wine's effects also involve nonalcoholic constituents. Primary funding source: European Foundation for the Study of Diabetes.
Content may be subject to copyright.
Effects of Initiating Moderate Alcohol Intake on Cardiometabolic Risk
in Adults With Type 2 Diabetes
A 2-Year Randomized, Controlled Trial
Yftach Gepner, MPH*; Rachel Golan, RD, PhD*; Ilana Harman-Boehm, MD; Yaakov Henkin, MD; Dan Schwarzfuchs, MD;
Ilan Shelef, MD; Ronen Durst, MD; Julia Kovsan, MSc; Arkady Bolotin, PhD; Eran Leitersdorf, MD; Shoshana Shpitzen, MA;
Shai Balag, MD; Elad Shemesh, MD; Shula Witkow, RD, MPH; Osnat Tangi-Rosental, BA†; Yoash Chassidim, PhD;
Idit F. Liberty, MD; Benjamin Sarusi, MSc; Sivan Ben-Avraham, RD, MPH; Anders Helander, PhD; Uta Ceglarek, PhD;
Michael Stumvoll, MD; Matthias Blu¨ her, MD; Joachim Thiery, MD; Assaf Rudich, MD, PhD; Meir J. Stampfer, MD, DrPH; and
Iris Shai, RD, PhD
Background: Recommendations for moderate alcohol con-
sumption remain controversial, particularly in type 2 diabetes
mellitus (T2DM). Long-term randomized, controlled trials (RCTs)
are lacking.
Objective: To assess cardiometabolic effects of initiating mod-
erate alcohol intake in persons with T2DM and whether the type
of wine matters.
Design: 2-year RCT (CASCADE [CArdiovaSCulAr Diabetes &
Ethanol] trial). (ClinicalTrials.gov: NCT00784433)
Setting: Ben-Gurion University of the Negev–Soroka Medical
Center and Nuclear Research Center Negev, Israel.
Patients: Alcohol-abstaining adults with well-controlled T2DM.
Intervention: Patients were randomly assigned to 150 mL of
mineral water, white wine, or red wine with dinner for 2 years.
Wines and mineral water were provided. All groups followed a
Mediterranean diet without caloric restriction.
Measurements: Primary outcomes were lipid and glycemic
control profiles. Genetic measurements were done, and patients
were followed for blood pressure, liver biomarkers, medication
use, symptoms, and quality of life.
Results: Of the 224 patients who were randomly assigned, 94%
had follow-up data at 1 year and 87% at 2 years. In addition to
the changes in the water group (Mediterranean diet only), red
wine significantly increased high-density lipoprotein cholesterol
(HDL-C) level by 0.05 mmol/L (2.0 mg/dL) (95% CI, 0.04 to 0.06
mmol/L [1.6 to 2.2 mg/dL]; P < 0.001) and apolipoprotein(a)
1
level by 0.03 g/L (CI, 0.01 to 0.06 g/L; P = 0.05) and decreased
the total cholesterol–HDL-C ratio by 0.27 (CI, 0.52 to 0.01;
P = 0.039). Only slow ethanol metabolizers (alcohol dehydroge-
nase alleles [ADH1B*1] carriers) significantly benefited from the
effect of both wines on glycemic control (fasting plasma glucose,
homeostatic model assessment of insulin resistance, and hemo-
globin A
1c
) compared with fast ethanol metabolizers (persons
homozygous for ADH1B*2). Across the 3 groups, no material
differences were identified in blood pressure, adiposity, liver
function, drug therapy, symptoms, or quality of life, except that
sleep quality improved in both wine groups compared with the
water group (P = 0.040). Overall, compared with the changes in
the water group, red wine further reduced the number of com-
ponents of the metabolic syndrome by 0.34 (CI, 0.68 to
0.001; P = 0.049).
Limitation: Participants were not blinded to treatment
allocation.
Conclusion: This long-term RCT suggests that initiating moder-
ate wine intake, especially red wine, among well-controlled dia-
betics as part of a healthy diet is apparently safe and modestly
decreases cardiometabolic risk. The genetic interactions suggest
that ethanol plays an important role in glucose metabolism, and
red wine's effects also involve nonalcoholic constituents.
Primary Funding Source: European Foundation for the Study
of Diabetes.
Ann Intern Med. 2015;163:569-579. doi:10.7326/M14-1650 www.annals.org
For author affiliations, see end of text.
* Mr. Gepner and Dr. Golan contributed equally to this work.
Deceased.
This article was published online first at www.annals.org on 13 October
2015.
T
he risk–benefit balance of moderate alcohol con-
sumption in persons with diabetes is controversial
(1, 2). Epidemiologic studies suggest that the incidence
of type 2 diabetes mellitus (T2DM) is reduced among
moderate alcohol drinkers (3–7). Among healthy per-
sons (8, 9) and diabetic patients (10 –12), moderate al-
cohol consumption is linked to lower cardiovascular
and total mortality rates. Yet, whether to recommend
initiation of moderate alcohol consumption to patients
with T2DM is questionable. The American Diabetes As-
sociation (13) leaves moderate alcohol consumption to
personal preference. The American Heart Association
(14) recommends that alcohol use be discussed by the
physician and patient, given the lack of long-
term trials that support causal cardioprotective effects
(15).
A recent systematic review summarized short-term
clinical trials (16) that mostly involved healthy partici-
pants and suggested that moderate alcohol consump-
tion is associated with favorable patterns of biomarkers
of cardiovascular risk. The effect of moderate alcohol
See also:
Celebrating the ACP Centennial: From the Annals
Archive ..................................639
Summary for Patients.......................I-34
Annals of Internal Medicine ORIGINAL RESEARCH
© 2015 American College of Physicians 569
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
consumption on blood pressure (BP) (17) and adiposity
remains controversial (18, 19).
Some studies (20–23) suggest similar beneficial as-
sociations for different alcoholic beverages and impli-
cate ethanol as the primary mediator. In contrast, other
studies (24–27) propose that red wine may induce su-
perior benefits; whether red wine confers any advan-
tage over white wine is unclear.
After a 3-month feasibility trial (28) in 109 patients
with T2DM, we performed the 2-year CASCADE
(CArdiovaSCulAr Diabetes & Ethanol) trial among 224
diabetic patients. We hypothesized that initiating mod-
erate wine consumption would improve cardiometa-
bolic risk mainly because of the ethanol component.
Therefore, we predicted similar effects of red and white
wine. Because of genetic variability in alcohol metabo-
lism, we further hypothesized that the effects of wine on
the metabolic variables would vary by ADH1B
(rs1229984) genotype.
METHODS
Design Overview
The 2-year CASCADE trial involved alcohol-
abstaining diabetic participants who were randomly as-
signed in a parallel design (1:1:1) to mineral water,
white wine, or red wine (150 mL at dinnertime). Recruit-
ment of participants began in November 2009. Partici-
pants who provided informed consent were weighed
and measured at baseline and were randomly assigned
to a single-phase, 2-year intervention (June 2010 to
May 2012). The protocol did not change after trial com-
mencement (earlier modifications are detailed in Ap-
pendix Figure 1 and the Appendix, available at www
.annals.org). Further, CASCADE was conducted at 2
centers: Ben-Gurion University of the Negev–Soroka
Medical Center (BGU–SMC) and Nuclear Research Cen-
ter Negev (NRCN) in Israel. The study was approved
and monitored by the human subjects committees of
SMC and BGU, which also cover the NRCN. No financial
compensation was provided to participants.
Setting and Participants
During recruitment, we intentionally avoided any
emphasis on the alcohol component to correctly iden-
tify alcohol abstainers. Candidates were screened by a
physician for eligibility. We included men and women
aged 40 to 75 years with T2DM diagnosed according
to the American Diabetes Association criteria (29). Ex-
clusion criteria were as follows: more than 1 alcoholic
drink per week; personal or family history of addiction,
smoking, stroke, or myocardial infarction; major sur-
gery within the past 3 months; using more than 2 insu-
lin injections per day or an insulin pump; triglyceride
level greater than 4.52 mmol/L (400 mg/dL), hemoglo-
bin A
1c
(HbA
1c
) level less than 6.4% or 10% or more;
women with first-degree relatives with breast cancer; or
pregnant women.
Randomization and Intervention
We performed the randomization (detailed in the
Appendix) within strata of patients by recruitment site
and planned substudies; we used the PROC PLAN pro-
cedure in SAS software, version 9.2 (SAS Institute). At
NRCN, participants were randomly assigned to receive
water or red wine (1:1 ratio). At BGU–SMC, participants
in additional substudies were randomly assigned to wa-
ter or red wine (1:1 ratio) and the remaining partici-
pants to water, red wine, or white wine (1:1:3 ratio). The
participants were instructed to consume 150 mL (5
ounces) of the randomly assigned beverage with din-
ner by using a standard 150 mL measuring glass we
provided. The randomized beverages were dry red
wine (from Golan Heights Winery; 16.9 g of ethanol
[14.2% by volume], with 270.1 mg of gallic acid equiv-
alent of total phenols; 120 kcal/150 mL), dry white wine
(from Golan Heights Winery; 15.8 g of ethanol [13.3%
by volume], with 38.5 mg of gallic acid equivalent of
total phenols; 111 kcal/150 mL), or mineral water (from
Mey Eden). The red wine had 7-fold higher levels of
total phenols and between 4- to 13-fold higher levels of
the specific resveratrol group compounds than the
white wine. We provided the beverages at no charge
for 2 years (mineral water, 18.9 L/mo; white wine, 750
mL/mo [6 bottles]; or red wine, 325 mL/mo [14 bot-
tles]). Patients assigned to consume wine were in-
structed to initiate drinking gradually over the first
month and avoid driving after drinking. The partici-
pants were asked to return the empty bottles at each
visit to monitor use.
Mediterranean Dietary Guidelines
In an attempt to achieve a comparable healthy diet
and provide an incentive to participate, we provided all
participants with guidelines to follow a Mediterranean
diet as per our previous 2-year DIRECT (Dietary Inter-
EDITORS' NOTES
Context
The long-term benefits and risks of moderate alcohol
intake among patients with type 2 diabetes mellitus
(T2DM) are unclear.
Contribution
Alcohol-abstaining patients with T2DM were randomly
assigned to 150 mL (5 ounces) of red wine, white wine,
or mineral water with dinner for 2 years. Primary study
outcomes included lipid and glycemic control
measures.
Caution
The trial did not include a grape juice control group.
Implication
Moderate alcohol intake, particularly red wine, among
patients with T2DM was associated with decreased car-
diometabolic risks and no significant adverse events.
Genetic typing for alcohol dehydrogenase may identify
patients who may benefit clinically from moderate
alcohol consumption.
ORIGINAL RESEARCH Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes
570 Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Table 1. Baseline Characteristics of the CASCADE Study Population*
Variable Mineral Water
(n 83)
White Wine
(n 68)
Red Wine
(n 73)
Total
(n 224)
Age, y 59.1 (6.7) 60.6 (6.8) 59.3 (7.8) 59.7 (7.1)
Men, % 65 65 77 69
BMI, kg/m
2
29.7 (4.0) 30.4 (5.1) 30.0 (4.1) 30.0 (4.4)
Ethanol intake, g/d 2.1 (2.6) 2.5 (3.0) 2.5 (3.4) 2.3 (3.0)
Metabolic syndrome components
HDL-C level
Total
mmol/L 1.09 (0.31) 1.11 (0.27) 1.18 (0.33) 1.12 (0.34)
mg/dL 42.1 (12.1) 43.0 (10.6) 45.7 (12.9) 43.5 (13.0)
Men
mmol/L 1.03 (0.26) 1.04 (0.24) 1.12 (0.32) 1.06 (0.28)
mg/dL 39.7 (10.0) 40.0 (9.4) 43.3 (12.5) 41.1 (10.9)
Women
mmol/L 1.21 (0.37) 1.25 (0.28) 1.38 (0.30) 1.27 (0.33)
mg/dL 46.6 (14.3) 48.3 (10.9) 53.4 (11.6) 48.9 (12.7)
Triglyceride level
mmol/L 1.7 (1.3) 1.6 (0.7) 1.5 (0.7) 1.6 (1.0)
mg/dL 150.4 (118.5) 141.0 (65.8) 133.2 (65.7) 141.9 (88.8)
FPG level
mmol/L 8.3 (2.3) 8.5 (2.1) 8.3 (1.8) 8.3 (2.1)
mg/dL 149.5 (40.6) 153.3 (38.2) 148.9 (32.9) 150.4 (37.3)
BP, mm Hg
Systolic 136.3 (17.4) 136.3 (19.2) 139.7 (19.0) 137.4 (18.5)
Diastolic 77.2 (10.3) 77.4 (11.1) 79.4 (11.1) 78.0 (10.8)
Waist circumference, cm
Total 104.4 (9.7) 105.2 (13.4) 105.3 (9.6) 104.9 (10.9)
Men 104.1 (8.9) 108.5 (13.7) 106.8 (8.4) 106.3 (10.5)
Women 104.8 (11.3) 98.7 (10.4) 100.2 (11.5) 101.7 (11.2)
Mean positive metabolic syndrome criteria (SD), n 3.0 (1.2) 3.2 (1.3) 3.0 (1.2) 3.1 (1.2)
Other variables
Glycemic biomarkers
HbA
1c
level, %
6.9 (1.1) 6.9 (1.0) 6.8 (0.9) 6.9 (1.0)
Fasting insulin level, pmol/L 91 (49) 105 (68) 96 (53) 97 (57)
HOMA-IR score 4.8 (3.4) 5.7 (4.2) 5.0 (3.2) 5.2 (3.6)
Lipid biomarkers
LDL-C level
mmol/L 2.43 (0.79) 2.33 (0.80) 2.45 (0.81) 2.41 (0.80)
mg/dL 93.9 (30.5) 90.0 (30.9) 94.7 (31.2) 93.0 (30.8)
Apolipoprotein(a)
1
level, g/L
1.4 (0.2) 1.4 (0.2) 1.4 (0.2) 1.4 (0.2)
Apolipoprotein(b)
100
level, g/L
0.92 (0.20) 0.90 (0.20) 0.91 (0.20) 0.91 (0.20)
Total cholesterol–HDL-C ratio 4.3 (1.4) 4.0 (1.1) 4.0 (1.2) 4.1 (1.3)
Apolipoprotein(b)
100
–apolipoprotein(a)
1
ratio
0.68 (0.19) 0.67 (0.19) 0.64 (0.19) 0.66 (0.19)
Liver function biomarkers
ALT level, U/L 28.6 (4.2) 28.0 (11.5) 30.9 (13.9) 28.8 (13.3)
AST level, U/L 24.8 (10.1) 24.35 (10.7) 25.1 (8.4) 24.8 (9.7)
ALP level, μkat/L 1.1 (0.3) 1.0 (0.3) 1.1 (0.3) 1.1 (0.3)
Bilirubin level
μmol/L 8.21 (5.13) 8.03 (5.13) 8.38 (5.11) 8.21 (5.13)
mg/dL 0.48 (0.30) 0.47 (0.30) 0.49 (0.30) 0.48 (0.30)
Genotype
ADH1B polymorphism, %
CC (ADH1B*1) 27.6 45.0 36.4 35.6
CT (ADH1B*1*2) 44.7 46.7 37.9 43.1
TT (ADH1B*2, rs1229984) 27.6 8.3 25.8 21.3
Number of different medications in current use, n (%)
Oral glycemic control
0 19 (22.9) 14 (20.6) 20 (27.4) 53 (23.7)
1 38 (45.8) 33 (48.5) 25 (34.2) 96 (42.9)
2 23 (27.7) 16 (23.5) 21 (28.8) 60 (26.8)
3–4 3 (3.6) 5 (7.4) 7 (9.6) 15 (6.7)
Insulin treatment‡
0 73 (88.0) 60 (88.2) 63 (86.3) 196 (87.5)
1 10 (12.0) 7 (10.3) 10 (13.7) 27 (12.1)
2 0 (0) 1 (1.5) 0 (0) 1 (0.4)
Lipid-lowering therapy
0 26 (31.3) 21 (30.9) 26 (35.6) 73 (32.6)
1 49 (59.0) 38 (55.9) 39 (53.4) 126 (56.3)
2 8 (9.6) 9 (13.2) 8 (11.0) 25 (11.2)
Continued on following page
Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes ORIGINAL RESEARCH
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 571
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
ferase, alkaline phosphatase, or bilirubin) in any of the
3 groups (Appendix Figure 6, available at www.annals
.org). The ADH1B polymorphism had no significant ef-
fect on changes in liver enzymes (Appendix Figure 7,
available at www.annals.org).
Overall Effect on the Metabolic Syndrome
Compared with water, only the red wine group had
an overall further significant decrease in the number of
variables of the metabolic syndrome by 0.34 (CI, 0.68
to 0.001; P = 0.049) (Table 2)(Appendix Figure 8,
available at www.annals.org).
DISCUSSION
There are several clinical implications of this 2-year
trial. First, among patients with well-controlled T2DM
and a low risk for alcohol abuse, initiating moderate
alcohol consumption in the context of a healthy diet is
apparently safe and may modestly reduce cardiometa-
bolic risk. Second, red wine may be somewhat superior
in improving lipid variables, which indicates the poten-
tial synergy of moderate alcohol intake with specific
nonalcoholic wine constituents. Third, differential ef-
fects on the glycemic control we saw were based on
ADH1B genetic variants, which indicate that ethanol
may play a role in the glycemic effects of the wine in-
tervention. Finally, ADH1B variants may assist in identi-
fying patients with T2DM for whom moderate wine con-
sumption may be clinically beneficial.
The study has several limitations. First, study partic-
ipants were not blinded to group assignment. Second,
the gold standard for assessing insulin resistance is the
hyperinsulinemic euglycemic clamp (37), but we used
fasting insulin and glucose levels to estimate insulin re-
sistance (32). The HOMA-IR score, however, correlates
reasonably well with results of clamp studies, including
in diabetic patients (38). In addition, change in HDL-C
level may not necessarily correlate with HDL functional-
ity (39). Despite the known differences in pharmacoki-
netics of alcohol by sex (40), we provided similar alco-
hol doses for both sexes for safety and simplicity. Thus,
a 2-glass-per-day regimen among men might have
yielded larger changes. We relied on self-reported al-
cohol intake to assess adherence, but we also moni-
tored use by reviewing returned empty bottles. Finally,
we cannot dissect the contribution of nonalcoholic red
wine constituents from their combined effect with eth-
anol.
Strengths of the study include its long duration, use
of comprehensive measurements, and high percentage
of participants who completed the 2-year follow-up.
The nutritional education sessions and free mineral wa-
ter supply to the control group allowed equal intensity
of intervention and enabled assessment of wine-
specific effects within the setting of a Mediterranean
diet. The CASCADE design benefited from initiating
the intervention in a population that generally has low
consumption of alcohol (41) and reducing the risk for
abuse by including participants older than 40 years
Figure 3.
Effect of long-term consumption of 150 mL of mineral water, white wine, or red wine per day on glycemic control
and BP variables in type 2 diabetes mellitus according to genetic variation in ADH1B.
FPG Level Fasting Insulin Level
−20
−10
0
10
20
30
40
Change, mg/dLChange, mg/dL
Wine
Water
CC CT TT
HOMA-IR Score HbA
1c
Level
−6
−5
−4
−3
−2
−1
0
1
2
Change, µU/mL
Change, µU/mL
−20
−10
0
10
20
30
40
CC CT TTCC CT TT
−2.5
−2
−1.5
−1
−0.5
0
0.5
1
1.5
Change
−0.6
−0.4
−0.2
1E-15
0.2
0.4
0.6
0.8
1
Change, %
CC CT TT CC CT TT
−6
−4
−2
0
2
−2.5
−1.5
−0.5
0.5
1.5
−0.6
−0.4
−0.2
1E-15
0.2
0.4
0.6
0.8
1
CC CT TT CC CT TT CC CT TT CC CT TT CC CT TT
CC CT TT
Systolic Diastolic
CC CT TT
−20
−15
−10
−5
0
5
10
−15
−10
−5
0
5
10
Change, mm Hg
Change, mm Hg
Change
Change, %
Change, mm Hg
Change, mm Hg
−20
−15
−10
−5
0
5
10
−15
−10
−5
0
5
10
Blood Pressure Glycemic Control Biomarkers
P = 0.043
P = 0.231 P = 0.802 P = 0.637 P = 0.373 P = 0.994 P = 0.195
P = 0.085 P = 0.012 P = 0.024 P = 0.059 P = 0.006
The P values are for the comparison of the combined genotypes CC (ADH1B*1 homozygotes; slow alcohol metabolism) and CT (heterozygotes)
group versus the TT (ADH1B*2 homozygotes; fast alcohol metabolism) genotype group. Variables are mean changes; bars indicate 95% CIs, and
the between-group analyses for differences are for 2 y. A total of 173 participants with available DNA samples completed the 2-y trial—103 in the
combined wine group and 70 in the water group. To convert FPG values to mmol/L, multiply by 0.0555. FPG = fasting plasma glucose; HbA
1c
=
hemoglobin A
1c
; HOMA-IR = homeostatic model assessment of insulin resistance.
ORIGINAL RESEARCH Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes
576 Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
in Stata, version 12 [StataCorp]) adjusted for age, sex,
and the use of specific medications (lipid-lowering
medications for lipid biomarkers, oral medications for
glycemic control biomarkers, and antihypertensive
medications for BP). The main results present the
changes from baseline in the wine groups compared
with the change from baseline in the water group and
the corresponding 95% CIs. We did a sensitivity analy-
sis to compare participants who dropped out with
those who completed the study; we also analyzed only
participants who completed the study. We assessed the
changes in sex-specific criteria values of the metabolic
syndrome (31) over 24 months.
We evaluated the ADH1B polymorphism for Hardy–
Weinberg equilibrium and tested its effect on bio-
marker levels. We calculated mean changes in bio-
markers within each of the 3 ADH1B genotypes for the
combined red and white wine groups and the water
group, and we tested for interaction between genotype
and wine or water groups. For the analysis involving the
glycemic variables, BP, and lipid and liver function bio-
markers, we pooled data from carriers of either 1 or 2
wild-type alleles (CC and CT) and compared the data
with participants who were homozygous for the TT
(rs1229984) allele based on the biological effect and to
enhance statistical power.
On the basis of the results of our pilot study (28),
which compared the effects of wine with nonalcoholic
beer on glycemic control, we did not perform a formal
power calculation for this study. All P values were
2-sided. We used Stata software, version 12, and SPSS
software, version 19 (IBM), for statistical analyses.
Role of the Funding Source
The European Foundation for the Study of Diabe-
tes provided funding for the study. Beverages were
provided by Mey Eden and Golan Heights Winery.
These sources were not involved in the design of the
study, collection of data, statistical analysis, manuscript
preparation or interpretation, or decisions about sub-
mission for publication.
RESULTS
Baseline Characteristics
Of the 224 randomly assigned participants, 94%
and 87% completed the 12- and 24-month assess-
ments, respectively (mineral water, 94%; white wine,
77%; red wine, 88%; P = 0.007 among groups) (Appen-
dix Figure 1). Baseline characteristics were distributed
similarly across the groups (Table 1). The participants
(aged 59 years; 69% men; HbA
1c
level, 6.9%) had an
average of 3.1 of 5 criteria of the metabolic syndrome.
Most participants were receiving medications for dia-
betes, hypertension, and hypercholesterolemia. Base-
line alcohol intake was 2.3 grams per day (approxi-
mately 1 drink per week). The BGU–SMC participants
(Appendix Table 1, available at www.annals.org) were
slightly older, had a smaller proportion of men, and
had lower total cholesterol–HDL-C ratio than NRCN
participants. All other baseline biomarkers and ethanol
consumption were similar.
Adherence
During the trial, participants in the wine groups in-
creased their intake of the specifically assigned wines
(approximately 80% consumed daily), and the mineral
water group participants remained alcohol abstainers.
Beverage adherence, as reflected in the self-reported
scale (1% to 100%), was 82%, 85%, and 80% at 6
months and 87%, 84%, and 84% at 2 years for mineral
water, white wine, and red wine groups, respectively
(Appendix Figure 2, available at www.annals.org).
Figure 1.
Changes in key lipid biomarkers.
−1.5
−0.5
0.5
1.5
2.5
3.5
4.5
0 6 12 18 24
Change, mg/dL
Intervention, mo
HDL-C
−0.03
−0.01
0.01
0.03
0.05
0.07
0 6 12 18 24
Change, mg/dL
Intervention, mo
Apolipoprotein(a)
1
−0.6
−0.4
−0.2
0
0.2
0.4
061218
24
Change
Intervention, mo
Water White wine Red wine
Total Cholesterol–HDL-C Ratio
P < 0.001
P = 0.050
P = 0.039
The P values represent the comparison of 2-y differences in the red
wine group versus the water group. Variables are mean changes; bars
indicate 95% CIs, and the between-group analyses for differences are
for 2 y. At 6 mo, the participants who completed the study were as
follows: mineral water, 81; white wine, 62; red wine, 73. After 2 y, 30
participants dropped out (incomplete set of observations). The partic-
ipants who completed the study were as follows: mineral water, 78;
white wine, 52; red wine, 64. To convert HDL-C values to mmol/L,
multiply by 0.0259. To convert apolipoprotein(a)
1
values to g/L, mul
-
tiply by 0.01. HDL-C = high-density lipoprotein cholesterol.
Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes ORIGINAL RESEARCH
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 573
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Two-Year Changes
All of the findings are presented as the change
from baseline in the wine groups compared with the
change from baseline in the water group.
Primary Outcomes
After 2 years (Figure 1 and Table 2), HDL-C levels
significantly increased in the red wine group by 0.05
mmol/L (2.0 mg/dL) (95% CI, 0.04 to 0.06 mmol/L [1.6
to 2.2 mg/dL]; P < 0.001) compared with the water
group. Changes in apolipoprotein(a)
1
levels in the red
wine group had a similar pattern (0.03 g/L [CI, 0 to 0.06
g/L]; P = 0.050) compared with the water group. Bene-
ficial 2-year changes include that the total cholesterol–
HDL-C ratio further decreased in the red wine group by
0.27 (CI, 0.52 to 0.01; P = 0.039) compared with the
water group. The corresponding lipid changes in the
white wine group were not significantly different from
those in the water group.
Although both wines tended to improve some glu-
cose metabolism components after 2 years, only white
wine significantly decreased fasting plasma glucose
level by 1.0 mmol/L (17.2 mg/dL) (CI, 1.60 to 0.3
mmol/L [28.9 to 5.5 mg/dL]; P = 0.004) and
HOMA-IR score by 1.2 (CI, 2.1 to 0.2; P = 0.019)
compared with the water group (Figure 2). Primary out-
comes did not vary across the recruitment sites (Ap-
pendix Figure 3, available at www.annals.org). The re-
sults did not materially differ among participants who
completed the trial compared with the entire group
(Appendix Figure 1).
Secondary Outcomes
The apolipoprotein(b)
100
–apolipoprotein(a)
1
ratio
decreased only in the red wine group by 0.03 (CI,
0.06 to 0.00; P = 0.058) compared with the water
group. Changes in triglyceride levels were more favor-
able in the white wine group (0.09 mmol/L [7.9 mg/
dL] [CI, 0.2 to 0.03 mmol/L {13.3 to 2.5 mg/dL}];
P = 0.004) and red wine group (0.1 mmol/L [12.0
mg/dL] [CI, 0.3 to 0.02 mmol/L {22.4 to 1.6 mg/
dL}]; P = 0.023] than in the water group.
Adiposity and BP
All 3 groups had modest and similar reductions in
waist circumference (mean, 1.48 cm) and body
weight (mean, 1.4 kg) from baseline. After 2 years of
intervention, we found no significant differences in BP
among the 3 groups (Table 2).
Table 2. Mean 2-y Changes From Baseline in Cardiometabolic Variables in the CASCADE Trial*
Variable Mineral Water (n 83) White Wine (n 68)
Mean Change
(95% CI)
Mean Change
(95% CI)
Differences of the Mean
Changes vs. Water
(95% CI)†
P Value
Primary outcomes
HDL-C level 0.30
mmol/L 0.04 (0.0008 to 0.09) 0.02 (−0.03 to 0.07) −0.03 (−0.08 to 0.02)
mg/dL 1.70 (0.03 to 3.30) 0.66 (−1.30 to 2.60) −1.00 (−3.00 to 0.90)
Apolipoprotein(a)
1
level, g/L
0.03 (−0.01 to 0.06) 0.02 (−0.03 to 0.06) −0.01 (−0.07 to 0.04) 0.65
Total cholesterol–HDL-C ratio −0.08 (−0.44 to 0.27) −0.003 (−0.300 to 0.290) 0.08 (−0.08 to 0.23) 0.35
FPG level 0.004
mmol/L 0.57 (−0.03 to 1.18) −0.4 (−0.9 to 0.1) −1.0 (−1.6 to −0.3)
mg/dL 10.30 (−0.63 to 21.20) −7.1 (−15.7 to 1.6) −17.2 (−28.9 to −5.5)
HOMA-IR score −0.19 (−0.87 to 0.49) −1.36 (−2.00 to −0.74) −1.20 (−2.10 to −0.20) 0.019
Secondary outcomes
Apolipoprotein (b)
100
–apolipoprotein(a)
1
ratio
0.01 (−0.03 to 0.04) 0.02 (−0.03 to 0.07) 0.01 (−0.02 to 0.05) 0.47
Triglyceride level 0.004
mmol/L 0.1 (−0.2 to 0.4) 0.02 (−0.1 to 0.2) −0.09 (−0.2 to −0.03)
mg/dL 10.4 (−17.8 to 38.7) 1.6 (−13.0 to 16.2) −7.9 (−13.3 to −2.5)
LDL-C level 0.59
mmol/L 0.05 (−0.1 to 0.2) 0.1 (−0.1 to 0.4) 0.06 (−0.2 to 0.3)
mg/dL 2.1 (−5.1 to 9.4) 4.5 (−4.9 to 14.0) 2.2 (−5.9 to 10.4)
Apolipoprotein(b)
100
level, g/L
0.03 (−0.02 to 0.09) 0.04 (−0.02 to 0.11) 0.01 (−0.05 to 0.07) 0.77
Fasting insulin level, pmol/L −10.42 (−20.14 to −0.49) −20.1 (−29.9 to −10.4) −9.72 (−23.61 to 3.75) 0.155
HbA
1c
level, %
0.34 (0.08 to 0.60) 0.27 (0.07 to 0.47) −0.06 (−0.56 to 0.44) 0.82
Systolic BP, mm Hg −4.80 (−9.70 to 0.14) 1.7 (−3.9 to 7.3) 6.40 (−0.98 to 13.80) 0.089
Diastolic BP, mm Hg −0.9 (−3.8 to 2.1) −1.3 (−4.8 to 2.1) −0.40 (−2.50 to 1.70) 0.71
Waist circumference, cm −1.80 (−3.00 to −0.54) −1.400 (−2.800 to 0.002) 0.39 (−1.40 to 2.20) 0.67
Weight, kg −1.30 (−2.00 to −0.57) −1.50 (−2.30 to −0.70) −0.20 (−1.40 to 0.97) 0.74
Number of positive metabolic syndrome criteria −0.01 (−0.25 to 0.22) −0.20 (−0.52 to 0.11) −0.16 (−0.51 to 0.19) 0.38
BP = blood pressure; CASCADE = CArdiovaSCulAr Diabetes & Ethanol; FPG = fasting plasma glucose; HbA
1c
= hemoglobin A
1c
; HDL-C =
high-density lipoprotein cholesterol; HOMA-IR = homeostatic model assessment of insulin resistance; LDL-C = low-density lipoprotein cholesterol.
* After2yoftheintervention, 30 participants dropped out and had incomplete sets of observations. The following participants completed the
intervention: mineral water (n = 78), white wine (n = 52), and red wine (n = 64).
Between-group analyses for 2-y differences.
ORIGINAL RESEARCH Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes
574 Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Genetic Interaction of ADH1B and Wine
Of the 203 participants with available DNA sam-
ples, 35.6% were homozygous for the wild-type C allele
ADH1B*1 (CC: “slow ethanol metabolism”), 21.3% were
homozygous for ADH1B*2 (Arg48His; rs1229984; TT:
“fast ethanol metabolism”), and 43.1% were heterozy-
gous (CT), which was consistent with the Hardy–Wein-
berg equilibrium. We analyzed 2-year genetic data
from 173 participants who had DNA samples and had
completed the trial. As expected in the water group,
changes in biomarkers did not differ across genetic
ADH1B variants. When the red and white wine groups
were combined (Figure 3), however, the improvements
in glycemic control were mostly achieved among carri-
ers of ADH1B*1. These carriers had favorable signifi-
cant changes compared with carriers of ADH1B*2 ho-
mozygotes; FPG level was 0.21 mmol/L (3.8 mg/dL)
versus 0.82 mmol/L (14.8 mg/dL; P = 0.043), HOMA-IR
score was 1.4 versus 0.3 (P = 0.012), and HbA
1c
level
was 0.1% versus 0.6% (P = 0.024). The test for interac-
tion between genotype and wine or water had a P value
of less than 0.05.
In contrast to the favorable effect of ADH1B*1 on
glycemic control variables, we found a statistically sig-
nificant improvement in BP only among the fast me-
tabolizers; homozygotes for ADH1B*2 (TT; rs1229984)
exhibited stronger BP-lowering effects of wine on dia-
stolic (P = 0.006) and systolic (P = 0.059) BP compared
with ADH1B*1 wild-type C allele carriers. The ADH1B
polymorphism had no significant effect on changes in
lipids induced by wine (mean change in HDL-C level,
0.06 mmol/L [SD, 0.18] [2.2 mg/dL {SD, 7.0}] vs. 0.05
mmol/L [SD, 0.20] [2.0 mg/dL {SD, 7.7}]; change in apo-
lipoprotein(a)
1
level, 0.04 g/L [SD, 0. 17] vs. 0.03 g/L
[SD, 0.15] of CC vs. TT groups; P > 0.05]).
Diet, Quality of Life, Medication Use, and Liver
Function Biomarkers
During the trial, the participants had no material
changes in energy intake (36 kcal/d) or energy ex-
penditure (0.78 metabolic equivalents/wk). However,
all 3 groups improved their nutrition similarly as ex-
pected from the Mediterranean diet (Appendix Figure
4, available at www.annals.org).
We saw no significant wine-related adverse events,
symptoms (Appendix Table 2, available at www.annals
.org), or changes in quality of life, except that reported
sleep quality was significantly improved in both wine
groups compared with water (P = 0.040).
In all 3 groups, there was no material change in
drug therapy according to the use of oral hypoglyce-
mic agents; insulin; or antihypertensive, lipid-lowering,
or antiplatelet agents (Appendix Figure 5, available at
www.annals.org).
No material changes were seen in liver function
tests (alanine aminotransferase, aspartate aminotrans-
Figure 2.
Changes in key glycemic control biomarkers.
−16
−12
−8
−4
0
4
8
12
16
20
0 6 12 18 24
Change, mg/dL
Intervention, mo
FPG Level
−2
−1.5
−1
−0.5
0
0.5
1
0 6 12 18 24
Change
Intervention, mo
HOMA-IR Score
P = 0.004
P = 0.019
Water White wine Red wine
The P values represent the comparison of 2-y differences in the white
wine group versus the water group. Variables are mean change; bars
indicate 95% CIs, and the between-group analyses for differences are
for 2 y. At 6 mo, the participants who completed the study were as
follows: mineral water, 81; white wine, 62; red wine, 73. After 2 y, 30
participants dropped out (incomplete set of observations). The partic-
ipants who completed the study were as follows: mineral water, 78,
white wine, 52; red wine, 64. To convert FPG values to mmol/L, mul-
tiply by 0.0555. FPG = fasting plasma glucose; HOMA-IR = homeo-
static model assessment of insulin resistance.
Table 2—Continued
Red Wine (n 73)
Mean Change
(95% CI)
Differences of the Mean
Changes vs. Water
(95% CI)†
P Value
<0.001
0.09 (0.05 to 0.14) 0.05 (0.04 to 0.06)
3.6 (1.9 to 5.3) 2.0 (1.6 to 2.2)
0.06 (0.02 to 0.10) 0.03 (0 to 0.06) 0.050
−0.36 (−0.62 to −0.09) −0.27 (−0.52 to −0.01) 0.039
0.62
0.2 (−0.3 to 0.8) −0.4 (−1.7 to 1.0)
4.0 (−5.9 to 13.8) −6.4 (−31.3 to 18.6)
−0.98 (−1.70 to −0.30) −0.77 (−1.70 to 0.17) 0.109
−0.02 (−0.06 to 0.02) −0.03 (−0.06 to 0) 0.058
0.023
−0.01 (−0.2 to 0.2) −0.1 (−0.3 to −0.02)
−1.3 (−17.5 to 14.9) −12.0 (−22.4 to −1.6)
0.42
0.005 (−0.19 to 0.19) −0.05 (−0.2 to 0.06)
0.18 (−7.20 to 7.50) −1.9 (−6.5 to 2.7)
0.01 (−0.04 to 0.06) −0.02 (−0.05 to 0.01) 0.165
−20.1 (−30.6 to −9.7) −9.7 (−29.2 to 10.4) 0.35
0.12 (−0.08 to 0.32) −0.22 (−0.64 to 0.20) 0.30
−4.30 (−9.00 to 0.27) 0.48 (−11.00 to 12.00) 0.94
−3.00 (−5.80 to −0.21) −2.1 (−8.7 to 4.5) 0.54
−1.20 (−2.50 to 0.08) 0.63 (−1.10 to 2.40) 0.48
−1.60 (−2.60 to −0.63) −0.27 (−1.40 to 0.84) 0.63
−0.40 (−0.60 to −0.17) −0.34 (−0.68 to 0) 0.049
Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes ORIGINAL RESEARCH
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 575
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
ferase, alkaline phosphatase, or bilirubin) in any of the
3 groups (Appendix Figure 6, available at www.annals
.org). The ADH1B polymorphism had no significant ef-
fect on changes in liver enzymes (Appendix Figure 7,
available at www.annals.org).
Overall Effect on the Metabolic Syndrome
Compared with water, only the red wine group had
an overall further significant decrease in the number of
variables of the metabolic syndrome by 0.34 (CI, 0.68
to 0.001; P = 0.049) (Table 2)(Appendix Figure 8,
available at www.annals.org).
DISCUSSION
There are several clinical implications of this 2-year
trial. First, among patients with well-controlled T2DM
and a low risk for alcohol abuse, initiating moderate
alcohol consumption in the context of a healthy diet is
apparently safe and may modestly reduce cardiometa-
bolic risk. Second, red wine may be somewhat superior
in improving lipid variables, which indicates the poten-
tial synergy of moderate alcohol intake with specific
nonalcoholic wine constituents. Third, differential ef-
fects on the glycemic control we saw were based on
ADH1B genetic variants, which indicate that ethanol
may play a role in the glycemic effects of the wine in-
tervention. Finally, ADH1B variants may assist in identi-
fying patients with T2DM for whom moderate wine con-
sumption may be clinically beneficial.
The study has several limitations. First, study partic-
ipants were not blinded to group assignment. Second,
the gold standard for assessing insulin resistance is the
hyperinsulinemic euglycemic clamp (37), but we used
fasting insulin and glucose levels to estimate insulin re-
sistance (32). The HOMA-IR score, however, correlates
reasonably well with results of clamp studies, including
in diabetic patients (38). In addition, change in HDL-C
level may not necessarily correlate with HDL functional-
ity (39). Despite the known differences in pharmacoki-
netics of alcohol by sex (40), we provided similar alco-
hol doses for both sexes for safety and simplicity. Thus,
a 2-glass-per-day regimen among men might have
yielded larger changes. We relied on self-reported al-
cohol intake to assess adherence, but we also moni-
tored use by reviewing returned empty bottles. Finally,
we cannot dissect the contribution of nonalcoholic red
wine constituents from their combined effect with eth-
anol.
Strengths of the study include its long duration, use
of comprehensive measurements, and high percentage
of participants who completed the 2-year follow-up.
The nutritional education sessions and free mineral wa-
ter supply to the control group allowed equal intensity
of intervention and enabled assessment of wine-
specific effects within the setting of a Mediterranean
diet. The CASCADE design benefited from initiating
the intervention in a population that generally has low
consumption of alcohol (41) and reducing the risk for
abuse by including participants older than 40 years
Figure 3.
Effect of long-term consumption of 150 mL of mineral water, white wine, or red wine per day on glycemic control
and BP variables in type 2 diabetes mellitus according to genetic variation in ADH1B.
FPG Level Fasting Insulin Level
−20
−10
0
10
20
30
40
Change, mg/dLChange, mg/dL
Wine
Water
CC CT TT
HOMA-IR Score HbA
1c
Level
−6
−5
−4
−3
−2
−1
0
1
2
Change, µU/mL
Change, µU/mL
−20
−10
0
10
20
30
40
CC CT TTCC CT TT
−2.5
−2
−1.5
−1
−0.5
0
0.5
1
1.5
Change
−0.6
−0.4
−0.2
1E-15
0.2
0.4
0.6
0.8
1
Change, %
CC CT TT CC CT TT
−6
−4
−2
0
2
−2.5
−1.5
−0.5
0.5
1.5
−0.6
−0.4
−0.2
1E-15
0.2
0.4
0.6
0.8
1
CC CT TT CC CT TT CC CT TT CC CT TT CC CT TT
CC CT TT
Systolic Diastolic
CC CT TT
−20
−15
−10
−5
0
5
10
−15
−10
−5
0
5
10
Change, mm Hg
Change, mm Hg
Change
Change, %
Change, mm Hg
Change, mm Hg
−20
−15
−10
−5
0
5
10
−15
−10
−5
0
5
10
Blood Pressure Glycemic Control Biomarkers
P = 0.043
P = 0.231 P = 0.802 P = 0.637 P = 0.373 P = 0.994 P = 0.195
P = 0.085 P = 0.012 P = 0.024 P = 0.059 P = 0.006
The P values are for the comparison of the combined genotypes CC (ADH1B*1 homozygotes; slow alcohol metabolism) and CT (heterozygotes)
group versus the TT (ADH1B*2 homozygotes; fast alcohol metabolism) genotype group. Variables are mean changes; bars indicate 95% CIs, and
the between-group analyses for differences are for 2 y. A total of 173 participants with available DNA samples completed the 2-y trial—103 in the
combined wine group and 70 in the water group. To convert FPG values to mmol/L, multiply by 0.0555. FPG = fasting plasma glucose; HbA
1c
=
hemoglobin A
1c
; HOMA-IR = homeostatic model assessment of insulin resistance.
ORIGINAL RESEARCH Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes
576 Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
with low addictive risk. Finally, the consistent findings
within the groups of glycemic, lipid, liver, BP, and an-
thropometric variables underscore the robustness of
the results.
The differences we saw between red and white
wine contrast with our original hypothesis that the ben-
eficial effects of wine are mediated predominantly by
alcohol (16, 23). Although our results suggest that the
effect of wine on glycemic control was mainly driven by
alcohol, a stronger effect of red wine was seen on lipid
levels and overall variables of the metabolic syndrome.
The provided wines were nearly equal in alcohol and
caloric content; however, the levels of total phenols in
red wine were 7 times higher. Whether red wine's phe-
nolic compounds (mostly resveratrol and quercetin)
(42) render it a uniquely cardioprotective alcoholic bev-
erage is still debated (43– 45) because the systemic bio-
availability of polyphenols (46) is argued to be low. Dif-
ferences between red and white wine should be further
considered in this context.
The beneficial effect on lipid profile was mani-
fested mostly by increased HDL-C (9.8% increase in the
red wine group) and apolipoprotein(a)
1
levels. In a re
-
cent meta-analysis (16), 30 to 40 g of ethanol per day
significantly increased HDL-C levels by 0.09 to 0.10
mmol/L (3.5 to 4 mg/dL) and apolipoprotein(a)
1
level
by 0.1 g/L; further, triglyceride or low-density lipopro-
tein cholesterol levels were not affected during short-
term (up to 3 months) trials. An 8-week trial in 20
insulin-resistant participants with an alcohol intake of
30 g increased HDL-C levels (47), but this did not occur
in a 30-day trial in 18 diabetic patients (48). The in-
creased transport rate of apolipoprotein(a)
1
and apoli
-
poprotein(a)
2
(49), elevated lipoprotein lipase activity
(50), increased cellular cholesterol efflux and its esteri-
fication (51), and decreased cholesteryl ester transfer
protein activity (52, 53) are some suggested mecha-
nisms for wine's ability to increase HDL-C levels.
The improvement of glycemic control was revealed
mainly by changes in FPG level and HOMA-IR score in
the white wine group compared with the water group.
When alcohol is administered on a short-term basis to
diabetic patients, inhibited gluconeogenesis is com-
pensated by increased glycogenolysis and leaves he-
patic glucose output unaltered (54). Our earlier
3-month pilot trial (28) showed that FPG level, but not
HbA
1c
or 2-hour postprandial glucose levels, was de
-
creased by wine intervention. In CASCADE, although
the hypoglycemic effect was similar within the first 6
months for both the red and white wine groups, white
wine had a modest advantage over red wine after 2
years for FPG level; however, this could be a chance
finding. The CASCADE participants had well-treated
T2DM at baseline, with a satisfactory baseline mean
HbA
1c
level (6.9%). One may speculate that the effect of
moderate wine consumption might manifest in de-
creased HbA
1c
level only in patients whose metabolic
levels are less well-controlled (55).
Class 1 alcohol dehydrogenase contributes to ap-
proximately 70% of total hepatic ethanol-oxidizing ca-
pacity (56), and a common polymorphism of the
ADH1B gene, Arg48His (rs1229984), is associated with
greatly enhanced enzymatic activity (56, 57). The gene–
alcohol and diabetes risk interaction is controversial in
observational studies (58, 59). We found that diabetic
patients who were slow alcohol metabolizers had im-
proved glycemic control by initiating moderate wine
consumption, which suggests that alcohol may play a
role in glucose metabolism. In contrast, diabetic pa-
tients who were fast ethanol metabolizers benefited the
most from the wine-induced BP-lowering effect, which
suggests a mediatory role for ethanol metabolites and
potentially explains the inconsistent reports about BP
effects on moderate alcohol intake (17). Our results are
in accordance with a recent Mendelian randomization
analysis of observation studies reporting that carriers of
the ADH1B (rs1229984) had lower BP (2).
Initiation of wine consumption did not alter the
number of medications used and did not adversely af-
fect liver function biomarkers or adiposity. The benefit
of wine on sleep quality was in accordance with our
previous wine trial (28). To our knowledge, CASCADE
is the first large, long-term RCT of alcohol, and the re-
sults suggest modest beneficial effects of initiating
moderate wine consumption among alcohol-abstaining
patients older than 40 years with T2DM. These benefits
should be weighed against potential risks when trans-
lated into clinical practice.
From Ben-Gurion University of the Negev and Soroka Medical
Center, Beer Sheva, Israel; Nuclear Research Center Negev,
Dimona, Israel; Hadassah Hebrew University Medical Center,
Jerusalem, Israel; Karolinska Institute, Solna, Sweden; Univer-
sity of Leipzig, Leipzig, Germany; and Brigham and Women's
Hospital and Harvard School of Public Health, Boston,
Massachusetts.
Acknowledgment: The authors thank the CASCADE partici-
pants for their consistent cooperation. They thank Harel Segal
from Nuclear Research Center Negev; Dr. Lena Novak, Dr.
Michael Friger, Dr. Arie Moran, Dr. Amos Katz, Noa Cohen,
Michal Rein, Nitzan Bril, and Dana Serfaty from Ben-Gurion
University of Negev; Dr. Tatiana Shuster, Sagit Saadon, Malka
Kaminsky, Yasmin Asuly, Roman Tsirkin, and David Shushan
from Soroka Medical Center; Eyal Goshen, Meir Aviv, Hassia
Krakauer, Haim Strasler, Dr. Ziva Schwartz, Dr. Einat Sheiner,
Dr. Dov Brickner, Dr. Rachel Marko, Esther Katorza, Ilanit Asu-
lin, and Tzvika Tzur from Nuclear Research Center Negev; and
Dr. Rosa M. Lamuela-Raventos, University of Barcelona.
Grant Support: By the European Foundation for the Study of
Diabetes of the European Association for the Study of
Diabetes.
Disclosures: The authors have no relationship with the com-
panies that make products relevant to the manuscript. Drs.
Shai and Bolotin had full access to all of the data in the study
and take responsibility for the integrity of the data and the
accuracy of the data analysis. Dr. Blu¨ her reports compensa-
tion as a board member of Novartis Pharmaceuticals, Boehr-
inger Ingelheim, and Sanofi; compensation as a consultant for
Novo Nordisk, Eli Lilly Pharmaceuticals, and AstraZeneca; and
Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes ORIGINAL RESEARCH
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 577
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
payment for lectures (including service on speakers bureaus)
for Sanofi, Eli Lilly Pharmaceuticals, Novo Nordisk, Bayer
HealthCare Pharmaceuticals, AstraZeneca, Novartis Pharma-
ceuticals, and Berlin-Chemie outside of the submitted work.
Authors not named here have disclosed no conflicts of inter-
est. Disclosures can also be viewed at www.acponline.org
/authors/icmje/ConflictOfInterestForms.do?msNum=M14
-1650.
Reproducible Research Statement: Study protocol: Available
from Dr. Shai (e-mail, irish@bgu.ac.il). Statistical code and data
set: Not available.
Requests for Single Reprints: Iris Shai, RD, PhD, Department
of Public Health, The S. Daniel Abraham International Center
for Health and Nutrition, Ben-Gurion University of the Negev,
PO Box 653, Beer Sheva, 8410501, Israel; e-mail, irish@
exchange.bgu.ac.il.
Current author addresses and author contributions are avail-
able at www.annals.org.
References
1. Grønbaek M. The positive and negative health effects of alcohol-
and the public health implications. J Intern Med. 2009;265:407-20.
[PMID: 19298457] doi:10.1111/j.1365-2796.2009.02082.x
2. Holmes MV, Dale CE, Zuccolo L, Silverwood RJ, Guo Y, Ye Z, et al;
InterAct Consortium. Association between alcohol and cardiovascu-
lar disease: Mendelian randomisation analysis based on individual
participant data. BMJ. 2014;349:g4164. [PMID: 25011450] doi:10
.1136/bmj.g4164
3. Baliunas DO, Taylor BJ, Irving H, Roerecke M, Patra J, Mohapatra
S, et al. Alcohol as a risk factor for type 2 diabetes: A systematic
review and meta-analysis. Diabetes Care. 2009;32:2123-32. [PMID:
19875607] doi:10.2337/dc09-0227
4. Carlsson S, Hammar N, Grill V. Alcohol consumption and type 2
diabetes Meta-analysis of epidemiological studies indicates a
U-shaped relationship. Diabetologia. 2005;48:1051-4. [PMID:
15864527]
5. Howard AA, Arnsten JH, Gourevitch MN. Effect of alcohol con-
sumption on diabetes mellitus: a systematic review. Ann Intern Med.
2004;140:211-9. [PMID: 14757619]
6. Koppes LL, Dekker JM, Hendriks HF, Bouter LM, Heine RJ. Mod-
erate alcohol consumption lowers the risk of type 2 diabetes: a meta-
analysis of prospective observational studies. Diabetes Care. 2005;
28:719-25. [PMID: 15735217]
7. Stampfer MJ, Colditz GA, Willett WC, Manson JE, Arky RA, Hen-
nekens CH, et al. A prospective study of moderate alcohol drinking
and risk of diabetes in women. Am J Epidemiol. 1988;128:549-58.
[PMID: 3414660]
8. Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ. Moderate
alcohol intake and lower risk of coronary heart disease: meta-analysis
of effects on lipids and haemostatic factors. BMJ. 1999;319:1523-8.
[PMID: 10591709]
9. Stampfer MJ, Colditz GA, Willett WC, Speizer FE, Hennekens CH.
A prospective study of moderate alcohol consumption and the risk
of coronary disease and stroke in women. N Engl J Med. 1988;319:
267-73. [PMID: 3393181]
10. Koppes LL, Dekker JM, Hendriks HF, Bouter LM, Heine RJ. Meta-
analysis of the relationship between alcohol consumption and coro-
nary heart disease and mortality in type 2 diabetic patients. Diabeto-
logia. 2006;49:648-52. [PMID: 16463045]
11. Solomon CG, Hu FB, Stampfer MJ, Colditz GA, Speizer FE, Rimm
EB, et al. Moderate alcohol consumption and risk of coronary heart
disease among women with type 2 diabetes mellitus. Circulation.
2000;102:494-9. [PMID: 10920059]
12. Tanasescu M, Hu FB, Willett WC, Stampfer MJ, Rimm EB. Alco-
hol consumption and risk of coronary heart disease among men with
type 2 diabetes mellitus. J Am Coll Cardiol. 2001;38:1836-42. [PMID:
11738282]
13. American Diabetes Association. Standards of medical care in
diabetes—2012. Diabetes Care. 2012;35 Suppl 1:S11-63. [PMID:
22187469] doi:10.2337/dc12-s011
14. Goldberg IJ, Mosca L, Piano MR, Fisher EA; Nutrition Commit-
tee, Council on Epidemiology and Prevention, and Council on Car-
diovascular Nursing of the American Heart Association. AHA Sci-
ence Advisory: Wine and your heart: a science advisory for
healthcare professionals from the Nutrition Committee, Council on
Epidemiology and Prevention, and Council on Cardiovascular Nurs-
ing of the American Heart Association. Circulation. 2001;103:472-5.
[PMID: 11157703]
15. Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA. Asso-
ciation of alcohol consumption with selected cardiovascular disease
outcomes: a systematic review and meta-analysis. BMJ. 2011;342:
d671. [PMID: 21343207] doi:10.1136/bmj.d671
16. Brien SE, Ronksley PE, Turner BJ, Mukamal KJ, Ghali WA. Effect
of alcohol consumption on biological markers associated with risk of
coronary heart disease: systematic review and meta-analysis of inter-
ventional studies. BMJ. 2011;342:d636. [PMID: 21343206] doi:10
.1136/bmj.d636
17. Briasoulis A, Agarwal V, Messerli FH. Alcohol consumption and
the risk of hypertension in men and women: a systematic review and
meta-analysis. J Clin Hypertens (Greenwich). 2012;14:792-8. [PMID:
23126352] doi:10.1111/jch.12008
18. Schu¨ tze M, Schulz M, Steffen A, Bergmann MM, Kroke A, Lissner
L, et al. Beer consumption and the ‘beer belly’: scientific basis or
common belief? Eur J Clin Nutr. 2009;63:1143-9. [PMID: 19550430]
doi:10.1038/ejcn.2009.39
19. Suter PM. Is alcohol consumption a risk factor for weight gain and
obesity? Crit Rev Clin Lab Sci. 2005;42:197-227. [PMID: 16047538]
20. Cleophas TJ. Wine, beer and spirits and the risk of myocardial
infarction: a systematic review. Biomed Pharmacother. 1999;53:417-
23. [PMID: 10554677]
21. Hansen AS, Marckmann P, Dragsted LO, Finne´ Nielsen IL,
Nielsen SE, Grønbaek M. Effect of red wine and red grape extract on
blood lipids, haemostatic factors, and other risk factors for cardiovas-
cular disease. Eur J Clin Nutr. 2005;59:449-55. [PMID: 15674304]
22. Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA Jr,
Stampfer MJ, Willett WC, et al. Roles of drinking pattern and type of
alcohol consumed in coronary heart disease in men. N Engl J Med.
2003;348:109-18. [PMID: 12519921]
23. Shai I, Rimm EB, Schulze MB, Rifai N, Stampfer MJ, Hu FB. Mod-
erate alcohol intake and markers of inflammation and endothelial
dysfunction among diabetic men. Diabetologia. 2004;47:1760-7.
[PMID: 15502925]
24. Black S. Clinical and pathologic reports. Newry, UK: Alex Wilkin-
son; 1819:1-47.
25. Chiva-Blanch G, Urpi-Sarda M, Llorach R, Rotches-Ribalta M,
Guille´ n M, Casas R, et al. Differential effects of polyphenols and al-
cohol of red wine on the expression of adhesion molecules and in-
flammatory cytokines related to atherosclerosis: a randomized clini-
cal trial. Am J Clin Nutr. 2012;95:326-34. [PMID: 22205309] doi:10
.3945/ajcn.111.022889
26. Estruch R, Sacanella E, Mota F, Chiva-Blanch G, Antu´ nez E,
Casals E, et al. Moderate consumption of red wine, but not gin, de-
creases erythrocyte superoxide dismutase activity: a randomised
cross-over trial. Nutr Metab Cardiovasc Dis. 2011;21:46-53. [PMID:
19819677] doi:10.1016/j.numecd.2009.07.006
27. Rimm EB, Klatsky A, Grobbee D, Stampfer MJ. Review of mod-
erate alcohol consumption and reduced risk of coronary heart dis-
ease: is the effect due to beer, wine, or spirits. BMJ. 1996;312:731-6.
[PMID: 8605457]
28. Shai I, Wainstein J, Harman-Boehm I, Raz I, Fraser D, Rudich A,
et al. Glycemic effects of moderate alcohol intake among patients
with type 2 diabetes: a multicenter, randomized, clinical intervention
trial. Diabetes Care. 2007;30:3011-6. [PMID: 17848609]
ORIGINAL RESEARCH Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes
578 Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
29. Report of the Expert Committee on the Diagnosis and Classifica-
tion of Diabetes Mellitus. Diabetes Care. 1997;20:1183-97. [PMID:
9203460]
30. Shai I, Schwarzfuchs D, Henkin Y, Shahar DR, Witkow S, Green-
berg I, et al; Dietary Intervention Randomized Controlled Trial
(DIRECT) Group. Weight loss with a low-carbohydrate, Mediterra-
nean, or low-fat diet. N Engl J Med. 2008;359:229-41. [PMID:
18635428] doi:10.1056/NEJMoa0708681
31. Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Do-
nato KA, et al; International Diabetes Federation Task Force on Ep-
idemiology and Prevention. Harmonizing the metabolic syndrome: a
joint interim statement of the International Diabetes Federation Task
Force on Epidemiology and Prevention; National Heart, Lung, and
Blood Institute; American Heart Association; World Heart Federa-
tion; International Atherosclerosis Society; and International Associ-
ation for the Study of Obesity. Circulation. 2009;120:1640-5. [PMID:
19805654] doi:10.1161/CIRCULATIONAHA.109.192644
32. Neumark YD, Friedlander Y, Durst R, Leitersdorf E, Jaffe D, Ram-
chandani VA, et al. Alcohol dehydrogenase polymorphisms influ-
ence alcohol-elimination rates in a male Jewish population. Alcohol
Clin Exp Res. 2004;28:10-4. [PMID: 14745297]
33. Shai I, Rosner BA, Shahar DR, Vardi H, Azrad AB, Kanfi A, et al;
DEARR study. Dietary evaluation and attenuation of relative risk: mul-
tiple comparisons between blood and urinary biomarkers, food fre-
quency, and 24-hour recall questionnaires: the DEARR study. J Nutr.
2005;135:573-9. [PMID: 15735096]
34. Chasan-Taber S, Rimm EB, Stampfer MJ, Spiegelman D, Colditz
GA, Giovannucci E, et al. Reproducibility and validity of a self-
administered physical activity questionnaire for male health profes-
sionals. Epidemiology. 1996;7:81-6. [PMID: 8664406]
35. Haffner SM, Miettinen H, Stern MP. The homeostasis model in
the San Antonio Heart Study. Diabetes Care. 1997;20:1087-92.
[PMID: 9203442]
36. Golan R, Shelef I, Rudich A, Gepner Y, Shemesh E, Chassidim Y,
et al. Abdominal superficial subcutaneous fat: a putative distinct pro-
tective fat subdepot in type 2 diabetes. Diabetes Care. 2012;35:
640-7. [PMID: 22344612] doi:10.2337/dc11-1583
37. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a
method for quantifying insulin secretion and resistance. Am J
Physiol. 1979;237:E214-23. [PMID: 382871]
38. Katsuki A, Sumida Y, Gabazza EC, Murashima S, Furuta M, Araki-
Sasaki R, et al. Homeostasis model assessment is a reliable indicator
of insulin resistance during follow-up of patients with type 2 diabe-
tes. Diabetes Care. 2001;24:362-5. [PMID: 11213893]
39. Karavia EA, Zvintzou E, Petropoulou PI, Xepapadaki E, Con-
stantinou C, Kypreos KE. HDL quality and functionality: what can
proteins and genes predict? Expert Rev Cardiovasc Ther. 2014;12:
521-32. [PMID: 24650316] doi:10.1586/14779072.2014.896741
40. Baraona E, Abittan CS, Dohmen K, Moretti M, Pozzato G, Chayes
ZW, et al. Gender differences in pharmacokinetics of alcohol. Alco-
hol Clin Exp Res. 2001;25:502-7. [PMID: 11329488]
41. Organisation for Economic Co-operation and Development.
Health at a Glance 2011. OCED indicators. OECD Publishing.
Accessed at www.oecd.org/els/health-systems/49105858.pdf on 24
August 2015.
42. Szmitko PE, Verma S. Antiatherogenic potential of red wine:
clinician update. Am J Physiol Heart Circ Physiol. 2005;288:H2023-
30. [PMID: 15653767]
43. Sahebkar A. Effects of resveratrol supplementation on plasma
lipids: a systematic review and meta-analysis of randomized con-
trolled trials. Nutr Rev. 2013;71:822-35. [PMID: 24111838] doi:10
.1111/nure.12081
44. Williamson G, Manach C. Bioavailability and bioefficacy of poly-
phenols in humans. II. Review of 93 intervention studies. Am J Clin
Nutr. 2005;81:243S-255S. [PMID: 15640487]
45. Liu Y, Ma W, Zhang P, He S, Huang D. Effect of resveratrol on
blood pressure: a meta-analysis of randomized controlled trials. Clin
Nutr. 2015;34:27-34. [PMID: 24731650] doi:10.1016/j.clnu.2014.03
.009
46. Kroon PA, Clifford MN, Crozier A, Day AJ, Donovan JL, Manach
C, et al. How should we assess the effects of exposure to dietary
polyphenols in vitro? Am J Clin Nutr. 2004;80:15-21. [PMID:
15213022]
47. Kim SH, Abbasi F, Lamendola C, Reaven GM. Effect of moderate
alcoholic beverage consumption on insulin sensitivity in insulin-
resistant, nondiabetic individuals. Metabolism. 2009;58:387-92.
[PMID: 19217456] doi:10.1016/j.metabol.2008.10.013
48. Bantle AE, Thomas W, Bantle JP. Metabolic effects of alcohol in
the form of wine in persons with type 2 diabetes mellitus. Metabo-
lism. 2008;57:241-5. [PMID: 18191055] doi:10.1016/j.metabol.2007
.09.007
49. De Oliveira E Silva ER, Foster D, McGee Harper M, Seidman CE,
Smith JD, Breslow JL, et al. Alcohol consumption raises HDL choles-
terol levels by increasing the transport rate of apolipoproteins A-I
and A-II. Circulation. 2000;102:2347-52. [PMID: 11067787]
50. Nishiwaki M, Ishikawa T, Ito T, Shige H, Tomiyasu K, Nakajima K,
et al. Effects of alcohol on lipoprotein lipase, hepatic lipase, choles-
teryl ester transfer protein, and lecithin:cholesterol acyltransferase in
high-density lipoprotein cholesterol elevation. Atherosclerosis. 1994;
111:99-109. [PMID: 7840818]
51. van der Gaag MS, van Tol A, Vermunt SH, Scheek LM, Schaafsma
G, Hendriks HF. Alcohol consumption stimulates early steps in re-
verse cholesterol transport. J Lipid Res. 2001;42:2077-83. [PMID:
11734581]
52. Savolainen MJ, Hannuksela M, Seppa¨ nen S, Kervinen K, Kesa¨ ni-
emi YA. Increased high-density lipoprotein cholesterol concentra-
tion in alcoholics is related to low cholesteryl ester transfer protein
activity. Eur J Clin Invest. 1990;20:593-9. [PMID: 2127749]
53. Hannuksela ML, Rantala M, Kesa¨ niemi YA, Savolainen MJ.
Ethanol-induced redistribution of cholesteryl ester transfer protein
(CETP) between lipoproteins. Arterioscler Thromb Vasc Biol. 1996;
16:213-21. [PMID: 8620335]
54. Puhakainen I, Koivisto VA, Yki-Ja¨ rvinen H. No reduction in total
hepatic glucose output by inhibition of gluconeogenesis with etha-
nol in NIDDM patients. Diabetes. 1991;40:1319-27. [PMID: 1936594]
55. Monnier L, Lapinski H, Colette C. Contributions of fasting and
postprandial plasma glucose increments to the overall diurnal hyper-
glycemia of type 2 diabetic patients: variations with increasing levels
of HbA(1c). Diabetes Care. 2003;26:881-5. [PMID: 12610053]
56. Edenberg HJ. The genetics of alcohol metabolism: role of alco-
hol dehydrogenase and aldehyde dehydrogenase variants. Alcohol
Res Health. 2007;30:5-13. [PMID: 17718394]
57. Ehrig T, Bosron WF, Li TK.
Alcohol and aldehyde dehydroge-
nase. Alcohol Alcohol. 1990;25:105-16. [PMID: 2198030]
58. Beulens JW, Rimm EB, Hendriks HF, Hu FB, Manson JE, Hunter
DJ, et al. Alcohol consumption and type 2 diabetes: influence of
genetic variation in alcohol dehydrogenase. Diabetes. 2007;56:
2388-94. [PMID: 17563066]
59. Dakeishi M, Murata K, Sasaki M, Tamura A, Iwata T. Association
of alcohol dehydrogenase 2 and aldehyde dehydrogenase 2 geno-
types with fasting plasma glucose levels in Japanese male and fe-
male workers. Alcohol Alcohol. 2008;43:143-7. [PMID: 18216179]
doi:10.1093/alcalc/agm173
Two-Year Moderate Alcohol Intervention in Adults With Type 2 Diabetes ORIGINAL RESEARCH
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 579
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Current Author Addresses: Drs. Golan, Bolotin, Rudich, and
Shai; Mr. Gepner, Ms. Kovsan, Ms. Witkow, Ms. Tangi-
Rosental, and Ms. Ben-Avraham: Department of Public Health,
Ben-Gurion University of the Negev, PO Box 653, Beer Sheva,
8410501, Israel.
Drs. Harman-Boehm, Henkin, Shelef, Shemesh, Chassidim,
and Liberty: Soroka Medical Center, Rager Boulevard, PO Box
151, Beer Sheva, 85025, Israel.
Dr. Schwarzfuchs and Mr. Sarusi: Nuclear Research Center Ne-
gev, 16th Beth Lethem Street, Dimona, 8477605, Israel.
Drs. Durst, Leitersdorf, Balag; and Ms. Spitzen: Hadassah He-
brew University Medical Center, Kiryat Hadassah, PO Box
12000, Jerusalem, 91120, Israel.
Dr. Helander: Department of Laboratory Medicine, H5, Divi-
sion of Clinical Chemistry, CI:74, Karolinska Institute, Karolin-
ska University Laboratory Hudding, Stockholm, SE-14186,
Sweden.
Drs. Ceglarek, Stumvoll, Blu¨ her, and Thiery: Department of
Diagnostics, University of Leipzig, Paul List Street 13-15,
04103 Leipzig, Germany.
Dr. Stampfer: Channing Division of Network Medicine, De-
partment of Medicine, Brigham and Women's Hospital and
Harvard School of Public Health, 181 Longwood Avenue, Bos-
ton, MA 02115.
Author Contributions: Conception and design: Y. Gepner, I.
Harman-Boehm, Y. Henkin, D. Schwarzfuchs, I. Shelef, R.
Durst, E. Shemesh, S. Witkow, M. Stumvoll, A. Rudich, M.J.
Stampfer, I. Shai.
Analysis and interpretation of the data: Y. Gepner, R. Golan, I.
Harman-Boehm, I. Shelef, R. Durst, J. Kovsan, A. Bolotin, S.
Shpitzen, E. Shemesh, Y. Chassidim, A. Helander, U. Ceglarek,
M. Stumvoll, M. Blu¨ her, A. Rudich, M.J. Stampfer, I. Shai.
Drafting of the article: Y. Gepner, R. Golan, Y. Henkin, D.
Schwarzfuchs, I. Shelef, R. Durst, J. Kovsan, A. Helander, U.
Ceglarek, M. Stumvoll, A. Rudich, M.J. Stampfer, I. Shai.
Critical revision of the article for important intellectual con-
tent: Y. Gepner, I. Harman-Boehm, Y. Henkin, I. Shelef, R.
Durst, J. Kovsan, E. Shemesh, A. Helander, M. Stumvoll, M.
Blu¨ her, J. Thiery, A. Rudich, I. Shai.
Final approval of the article: Y. Gepner, R. Golan, I. Harman-
Boehm, Y. Henkin, D. Schwarzfuchs, I. Shelef, R. Durst, J.
Kovsan, A. Bolotin, E. Leitersdorf, E. Shemesh, I.F. Liberty, B.
Sarusi, A. Helander, U. Ceglarek, M. Stumvoll, M. Blu¨ her, J.
Thiery, A. Rudich, M.J. Stampfer, I. Shai.
Provision of study materials or patients: Y. Gepner, R. Golan, I.
Harman-Boehm, Y. Henkin, D. Schwarzfuchs, R. Durst, S. Wit-
kow, O. Tangi-Rosental, I.F. Liberty, I. Shai.
Statistical expertise: R. Golan, R. Durst, A. Bolotin.
Obtaining of funding: Y. Gepner, R. Durst, M. Stumvoll, I. Shai.
Administrative, technical, or logistic support: I. Shelef, J.
Kovsan, E. Leitersdorf, S. Balag, E. Shemesh, O. Tangi-
Rosental, B. Sarusi, M. Stumvoll.
Collection and assembly of data: Y. Gepner, R. Golan, Y. Hen-
kin, D. Schwarzfuchs, I. Shelef, R. Durst, J. Kovsan, E. Shemesh,
S. Witkow, O. Tangi-Rosental, I.F. Liberty, B. Sarusi, S. Ben-
Avraham, M. Blu¨ her, A. Rudich, I. Shai.
APPENDIX
Revisions of the Original Protocol
Modifications of the basic protocol were made be-
tween 2008 and 2009 before it was submitted for ap-
proval by the institutional review board and before re-
cruitment. The modifications were reported online to
the European Foundation for the Study of Diabetes and
were as follows:
1. We added dry white wine as an additional inter-
vention group (to address red–white wine differences)
and changed the control from dealcoholized red wine
to mineral water (as a better control group because it
avoids the caloric content of dealcoholized wine).
2. We further excluded smokers and women with
family history of breast cancer for safety considerations
(to reduce any potential adverse interaction of alcohol
with those factors).
3. We added genetic measurements (analysis of
ADH1B polymorphism) to assess a potential genetic
interaction.
4. To simplify the operation, we decided to enroll
the patients from 2 centers: BGU–SMC and NRCN. The
original plan was to enroll the patients from 3 centers in
Israel.
Screening
Participants were recruited by using advertise-
ments for a dietary trial at BGU–SMC and NRCN. Fur-
ther announcements were made in local press and on
radio stations. We intentionally did not emphasize the
alcohol component to correctly identify alcohol ab-
stainers. After obtaining informed consent, candidates
met with a certified physician (both in the SMC and
NRCN clinics) to be screened for inclusion and exclu-
sion criteria and to obtain medical information.
Randomization
The randomization was performed within strata of
patients by sites and types of analysis planned for each
site group. The 2 sites were NRCN and BGU–SMC. The
BGU–SMC site included a subgroup of patients who re-
ceived additional follow-ups.
Substudies included continuous glucose monitor-
ing, magnetic resonance imaging, and Holter studies.
This formed 3 strata: NRCN patients (n = 59); BGU–SMC
patients with additional substudy (n = 41); and BGU–
SMC patients without additional substudy (n = 124).
The treatments of the first 2 groups were limited to
water and red wine with a randomization ratio 1:1 to
enhance statistical power to compare these groups
within specific substudies. The third group was ran-
domly assigned to 3 treatments—water, red wine, and
white wine—with a 1:1:3 ratio to allow for the final ratio
of the 3 treatment groups to be 1:1:1. The study pop-
ulation of BGU–SMC included 2 married couples (1
from each stratum). For each couple, the participant
with the lower serial number was randomly assigned
and the spouses received the same treatment as the
randomly assigned spouse.
Annals of Internal Medicine
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
The randomization was performed by the statisti-
cians from BGU in SAS, version 9.2, using the proce-
dure PROC PLAN.
Electronic Questionnaires
Participants completed electronic questionnaires
(at 0, 6, and 24 months) to collect data on demograph-
ics, lifestyle patterns, specific medications and symp-
toms, and quality of life (28). We assessed changes in
quality of life by inquiring about the frequency of the
following feelings: active, nervous, calm, energetic, de-
pressed, sad, exhausted, and happy. We also assessed
changes in hypoglycemia, panic, euphoria, illusions,
headaches, bleeding, eye function, diarrhea, body
pains, sexual desire, and sleep quality. We assessed
adherence to diet by a validated food frequency ques-
tionnaire (35) and used a validated questionnaire to as-
sess physical activity (36).
Laboratory Blood Biomarker Methods
Blood samples were obtained by venipuncture at 8
a.m. after an 8-hour fast at baseline and at 6 and 24
months; samples were stored at 80 °C. Measure-
ments were performed in laboratories in Leipzig, Ger-
many. Fasting plasma glucose level was measured by
Roche Glucose Hexokinase, generation 3 (Roche). Gly-
cated hemoglobin (HbA
1c
) was measured with Tina-
quant hemoglobin A
1c
, generation 3 (Roche). Plasma
insulin was measured with the use of an enzyme immu-
nometric assay (Immulite automated analyzer [Diagnos-
tic Products]), with a coefficient of variation (CV) of
2.5%. Serum total cholesterol (CV, 1.3%), HDL-C, low-
density lipoprotein cholesterol, and triglycerides (CV,
2.1%) were measured enzymatically with a Cobas 6000
automatic analyzer (Roche). Serum apolipoprotein(a)
1
(CV, 1.0% to 4.7%) and apolipoprotein(b)
100
(CV, 1.1%
to 3.1%) were measured by immunoturbidimetric as-
says (Tina-quant apolipoprotein A-1 and B100, version
2 [Roche]) on a Cobas 6000 automatic analyzer. Liver
enzyme and bilirubin were measured with Roche chem-
icals on the Cobas 6000 (Alkaline Phosphatase acc. to
IFCC, generation 2; Alanine Aminotransferase acc. to
IFCC with pyridoxal phosphate activation; Aspartate
Aminotransferase acc. to IFCC with pyridoxal phos-
phate activation; Bilirubin Total DPD, generation 2).
Sensitivity Analyses to Evaluate Departures
From the Assumption That Data Were Missing at
Random
Among the participants who dropped out of the
study, baseline characteristics were similar across their
assigned intervention groups in demographics, clinical
presentation, blood biomarkers, and the use of medi-
cations. Comparing those who dropped out with those
who completed the study, baseline characteristics in
demographics, clinical presentation, blood biomarkers,
and the use of medications were similar except that
those who dropped out had higher baseline HOMA-IR
scores than those who completed the study.
Within each intervention group, baseline character-
istics in demographic, clinical presentation, blood bio-
markers, and the use of medications were similar be-
tween those who dropped out versus completed the
study with the exception of the white wine group, in
which those who dropped out were younger, had
lower baseline HDL-C levels, and used fewer oral gly-
cemic medications than those who completed the
study.
Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 1.
Study flow diagram.
Screened for participation in the trial (n = 571)
Mineral water (n = 83) White wine (n = 68)
Withdrew (n = 6)
Personal reasons (immediately
after randomization): 2
Medical reason (elevated insulin): 1
Lack of motivation: 3
Withdrew (n = 4)
Lack of motivation: 2
Medical reasons: 2
Recurrent pancreatitis: 1
Medication interaction: 1
Withdrew (n = 6)
Lack of motivation: 3
Dislikes the taste of the wine: 1
Symptoms: 2
Dizziness: 1
Headaches: 1
Withdrew (n = 2)
Symptoms (headaches): 1
Lack of motivation: 1
Withdrew (n = 7)
Lack of motivation: 3
Symptoms (headaches): 3
Medical reason
(medication interaction): 1
Withdrew (n = 3)
Lack of motivation: 2
Medical reason
(hospitalization): 1
Withdrew for
personal reasons
immediately after
randomization (n = 2)
Red wine (n = 73)
Completed 6 months (n = 81)
Completed 6 months (n = 62) Completed 6 months (n = 73)
Completed 12 months (n = 81) Completed 12 months (n = 58) Completed 12 months (n = 71)
Included in primary analysis (n = 83) Included in primary analysis (n = 68) Included in primary analysis (n = 73)
Completed 24 months (n = 78)
Adherence: 94.0%
Completed 24 months (n = 52)
Adherence: 76.5%
Completed 24 months (n = 64)
Adherence: 87.7%
Nov. 2009
June 2010
(Start of
intervention)
Dec. 2010
(Adherence:
96.4%)
May 2011
(Adherence:
93.8%)
May 2012
(Adherence:
86.6%)
Randomization and assignment to allocated intervention (n
= 224)
Excluded according to inclusion and exclusion criteria (n = 347)
Low HbA
1c
level: 36
Smokers: 42
Using insulin more than 2 times a day: 68
Moderate alcohol consumer: 102
Lack of motivation: 30
Clinically unstable: 17
Women with first-degree relative with breast cancer: 12
HbA
1c
= hemoglobin A
1c
.
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Table 1. Key Demographic and Baseline Characteristics of the CASCADE Study Population, by Center*
Variable BGU–SMC NRCN
Mineral Water
(n 53)
White Wine
(n 68)
Red Wine
(n 44)
Total
(n 165)
Mineral Water
(n 30)
Red Wine
(n 29)
Total
(n 59)
Age, y 60.0 (7.1) 60.6 (6.8) 61.1 (7.5) 60.6 (7.1) 57.4 (5.6) 56.5 (7.4) 57.0 (6.6)
Men, % 58 65 66 63 79 93 86
BMI, kg/m
2
30.3 (4.2) 30.4 (5.1) 29.7 (3.9) 30.2 (4.5) 28.5 (3.6) 30.5 (4.4) 29.5 (4.1)
Ethanol intake, g/d 2.2 (2.7) 2.5 (3.0) 1.9 (2.8) 2.3 (2.8) 1.9 (2.6) 3.3 (4.1) 2.6 (3.5)
HDL-C level, mg/dL 42.7 (11.8) 43.0 (10.6) 46.0 (12.7) 43.7 (11.6) 40.8 (12.7) 45.1 (13.4) 43.1 (13.1)
Triglyceride level, mg/dL 145.0 (130.6) 141.0 (65.8) 128.2 (60.8) 139.0 (91.1) 161.1 (90.8) 140.6 (72.7) 150.5 (81.8)
FPG level, mg/dL 147.5 (30.9) 153.3 (38.2) 153.2 (31.7) 151.3 (34.1) 153.8 (56.2) 142.5 (34.2) 147.8 (45.8)
BP, mm Hg
Systolic 138.9 (18.7) 136.3 (19.2) 145.4 (19.4) 139.6 (19.4) 131.0 (13.4) 130.9 (14.8) 131.0 (14.0)
Diastolic 76.9 (11.8) 77.4 (11.2) 80.0 (12.8) 77.9 (11.8) 78.0 (6.9) 78.6 (8.1) 78.3 (7.4)
HbA
1c
level, % 6.9 (0.88) 6.9 (0.96) 7.0 (0.76) 6.9 (0.88) 6.9 (1.4) 6.7 (1.1) 6.8 (1.2)
Fasting insulin level, μIU/mL 13.5 (7.7) 15.1 (9.8) 13.5 (7.3) 14.2 (8.5) 12.2 (5.7) 14.2 (8.3) 13.3 (7.2)
HOMA-IR score 5.0 (3.5) 5.8 (4.2) 5.1 (3.0) 5.3 (3.7) 4.6 (3.3) 5.0 (3.4) 4.8 (3.3)
Total cholesterol–HDL-C ratio 4.1 (1.2) 4.0 (1.1) 3.7 (1.0) 4.0 (1.1) 4.7 (1.5) 4.3 (1.4) 4.5 (1.5)
BGU–SMC = Ben-Gurion University of the Negev–Soroka Medical Center; BMI = body mass index; BP = blood pressure; CASCADE = CArdiova-
SCulAr Diabetes & Ethanol; FPG = fasting plasma glucose; HbA
1c
= hemoglobin A
1c
; HDL-C = high-density lipoprotein cholesterol; HOMA-IR =
homeostatic model assessment of insulin resistance; NRCN = Nuclear Research Center Negev.
* Values are means (SDs). To convert HDL-C values to mmol/L, multiply by 0.0259. To convert FPG values to mmol/L, multiply by 0.0555.
Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 2.
Adherence to the assigned beverage
type after the 2-y intervention.
Mineral Water Group
0
20
40
60
80
100
White Wine Red Wine
Frequency of Wine Consumption
After 2 y, %
Frequency of Wine Consumption
After 2 y, %
Frequency of Wine Consumption
After 2 y, %
White Wine Group
0
20
40
60
80
100
White Wine Red Wine
Red Wine Group
0
20
40
60
80
100
White Wine Red Wine
Never Once a week 2 to 5 times a week Once a day
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 3.
Baseline and 2-y absolute levels of key variables, by site.
1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5
HDL-C Level, mg/dL
FPG Level, mg/dL HOMA-IR Score
Total Cholesterol–HDL-C Ratio
24
0
NRCN
BGU-SMC
NRCN
BGU-SMC
18 28 38 48 58 68 78
24
0
NRCN
BGU-SMC
NRCN
BGU–SMC
100 150 200 250 300 350
24
0
NRCN
BGU-SMC
NRCN
BGU-SMC
Water White wine
Red wine
1 6 11 16
24
0
NRCN
BGU-SMC
NRCN
BGU-SMC
To convert HDL-C values to mmol/L, multiply by 0.0259. To convert FPG values to mmol/L, multiply by 0.0555. BGU–SMC = Ben-Gurion University
of the Negev–Soroka Medical Center; HDL-C = high-density lipoprotein cholesterol; HOMA-IR = homeostatic model assessment of insulin resis-
tance; NRCN = Nuclear Research Center Negev.
Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 4.
2-y changes in food group consumption.
0
1
−1
100%
Decrease
Water White Wine
Intervention Group
Red Wine
Cereals, pasta, and potatoes
Vegetable oils
Vegetables
Fish
Nuts and grains
Legumes
Poultry
Fruits
Sweets
Bread and pastry
Solid oils (trans fat)
Eggs and dairy
100%
Increase
No
Change
0.5
−0.5
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Table 2. Acute and Adverse Effects*
Variable Water
(n 83)
White
Wine
(n 68)
Red
Wine
(n 73)
Deaths 000
Discontinued due to adverse effects
Total 1 5 5
Headaches 1 4
Dizziness 1
Recurrent pancreatitis 1
Elevated insulin 1
Medication interaction 1 1
Hospitalization (pulmonary cancer) 1
Acute events reported to ethics
committee
Total 6 9 2
MI 2 1
Morning motorcycle accident 1
Breast cancer 1
Pacemaker implantation 1
Coronary catheterization 1 1
Infectious mononucleosis 1
Lung cancer 1
Acute eye inflammation 1
Arrhythmia 2
Orthopedic problem 1
Lung inflammation 1
Hernia 1
Bypass surgery 1
MI = myocardial infarction.
* During the trial, there were no significant differences across the
groups in reported symptoms (e.g., bleeding, diarrhea, headaches,
illusions, and feeling energetic/calm/exhausted/nervous). Values are
numbers.
Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 5.
Mean changes (95% CIs) in the number of medications during the trial (at 0, 6, 12, and 24 mo) across the
assigned intervention groups.
Intervention, mo Intervention, mo
Intervention, mo Intervention, mo
Intervention, mo
Medications, n
Medications, n
Medications, n
Medications, n
Medications, n
Medications, n
Intervention, mo
0.2
0.05
0.15
0.4
0.6
0.8
1
0
0.2
0.4
0.6
0.8
1
0
0
0.1
0.2
0.25
0 6 12 18 24
0 6 12 18 24
0 6 12 18 24
0 6 12 18 24
0 6 12 18 24
0 6 12 18 24
1.2
0.2
0.4
0.6
0.8
1
0
1.2
0.2
0.4
0.6
0.8
1
0
0.2
0.3
0.1
0.4
0.5
0.6
0.7
0.8
0
1.2
1.4
1.6
Oral Glycemic Control Insulin Treatment
Lipid-Lowering Therapy Antihypertensive Therapy
Antiplatelet Agents Other
Water White wine Red wine
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 6.
Changes in liver function biomarkers.
−6
−5
−4
−3
−2
−1
1
0
−2
−4
−0.04
−0.02
0.02
0
0.04
−0.06
−0.08
2
4
6
0
0 6 12 18
24
0 6 12 18
0 6 12 18
24
01218
24
2
−4
−3
−2
−1
1
0
2
ALT Level
Intervention, mo
ALP Level
Intervention, mo
AST Level
Intervention, mo
Bilirubin Level
Intervention, mo
Change, U/L
Change, U/L
Change, U/L
Change, U/L
24
Water White wine Red wine
Liver Function
Biomarkers
Mineral Water (n = 83) White Wine (n = 68) Red Wine (n = 73)
Mean Change (95% CI) Mean Change
(95% CI)
Differences of
the Mean
Changes vs.
Water (95% CI)
P Value P ValueMean Change
(95% CI)
Differences of
the Mean
Changes vs.
Water (95% CI)
ALT level, U/L
−2 (−5.2 to 1.1) −0.89 (−3.3 to
1.5)
1.1 (−4.1 to 6.3) 0.87 −1.8 (−5.2 to 1.6) 0.20 (−4.8 to 5.2) 1.00
AST level, U/L
−1.5 (−3.1 to 0.20) 0.93 (0.58 to
2.4)
2.4 (−0.62 to 5.4) 0.149 −0.8 (−2.9 to 1.3) 0.67 (−2.2 to 3.6) 0.85
ALP level, µkat/L
0.052 (0.002 to 1.000) −0.020 (−0.058
to 0.017
−0.072 (−0.148
to 0.006)
0.075 0.003 (−0.042 to
0.048)
−0.048 (−0.123 to
0.027)
0.28
Bilirubin level
µmol/L
−0.34 (−0.86 to 0.34) −0.17 (−0.86 to
0.51)
0.17 (−1.30 to
1.20)
0.96 −0.51 (−1.20 to
0.34)
−0.17 (−1.20 to
0.86)
0.92
mg/dL
−0.02 (−0.05 to 0.02) −0.01 (−0.05 to
0.03)
0.01 (−0.06 to
0.07)
−0.03 (−0.07 to
0.02)
−0.01 (−0.07 to
0.05)
The mean changes from baseline are plotted; bars indicate 95% CIs. At 6 mo, the participants who completed the study were as follows: mineral
water, 81; white wine, 62; red wine, 73. After 2 y, we had 30 participants who dropped out; the participants who completed the study were as
follows: mineral water, 78; white wine, 52; red wine, 64. ALP = alkaline phosphatase; ALT = alanine aminotransferase; AST = aspartate
aminotransferase.
Annals of Internal Medicine Vol. 163 No. 8 20 October 2015 www.annals.org
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
Appendix Figure 7.
2-y changes in liver function biomarkers according to genetic variation in ADH1B.
−10
−5
5
10
0
−10
−5
5
10
0
−2
−4
−6
2
4
6
0
−2
−4
−6
2
4
6
0
−6
6
12
0
−6
6
12
0
−0.05
0.05
−0.1
−0.15
0
−0.05
0.05
−0.1
−0.15
0
Change, U/L
Change, U/L
Change, U/L
Change, mg/dL
Change, U/L
Change, U/L
Change, U/L
Change, mg/dL
Wine
ALT Level
CC CT TT CC CT TT CC CT TT CC CT TT
CC CT TT CC CT TT CC CT TT CC CT TT
AST Level ALP Level Bilirubin Level
Water
No significant differences in mean changes from baseline at 2 y were seen between the combined genotypes CC (ADH1B*1 homozygotes; “slow
alcohol metabolism”) and CT (heterozygotes) group versus the TT (ADH1B*2 homozygotes; “fast alcohol metabolism”) genotype group for water
and wine groups. Bars indicate 95% CIs. A total of 173 participants with available DNA samples completed the 2-y trial—103 in the combined wine
group and 70 in the water group. ALP = alkaline phosphatase; ALT = alanine aminotransferase; AST = aspartate aminotransferase.
Appendix Figure 8.
Overall effect of moderate wine
consumption on changes in the number of positive criteria
of the metabolic syndrome in persons with type 2
diabetes mellitus.
Change
The Metabolic Syndrome Components
Intervention, mo
Water White wine Red wine
0.2
−0.2
−0.4
−0.6
−0.8
0
0 6 12 18 24
P = 0.049
Bars indicate 95% CIs, and the P value denotes comparison of 2-y
differences in the red wine group versus the water group.
www.annals.org Annals of Internal Medicine Vol. 163 No. 8 20 October 2015
Downloaded From: http://annals.org/ by a Ben Gurion University User on 11/08/2015
... Source: Au Yeung et al., [29], Gepner et al., [30] ...
... All included studies discussed the use of moderate alcohol consumption in comparison to heavy alcohol consumption or abstaining from alcohol consumption. To produce heterogeneity of results, the included studies belong to 8 different countries: 1 in Israel [4], 2 in China [22,29,30], 1 in the United Kingdom [23], 1 in Poland [24], 1 in Turkey [25] 1 in Australia [26], 1 in Norway [27], and 1 in Italy [28]. Table 2. ...
... Source: Au Yeung et al., [29], Bell et al., [23], Gepner et al., [30], Onat et al., [25] b) Diastolic blood pressure ...
Article
There remains a debate on the effects of alcohol use on cardiovascular health. This systematic review and meta-analysis examined the relationship between this link and blood pressure, lipid profiles, and the frequency of cardiovascular events to shed light on it. A comprehensive search strategy using PRISMA guidelines resulted in nine papers that met the inclusion requirements. Moderate alcohol usage was shown to have a U- or J-shaped connection with cardiovascular events; excessive intake or abstinence was associated with increased risk, but moderate consumption had preventive advantages. Moderate drinkers showed positive changes in their lipid profiles and lowered blood pressure in comparison to heavy or non-drinkers. The wide range of research highlights the need to consider individual characteristics and study techniques. Despite the potential cardiovascular benefits of moderate alcohol use shown by these studies, treatment recommendations should be cautious. To provide personalized recommendations to patients and policymakers, future research should focus on explaining mechanisms, examining modifiers, and assessing the effects of different types of alcohol and drinking habits on cardiovascular health.
... Confounding could be overcome with the use of a randomized controlled trial (RCT) in which participants are randomly assigned to different alcohol consumption levels and followed up on health outcomes. Short-term trials have shown small amounts of alcohol lead to benefits for several cardiometabolic biomarkers (Brien et al., 2011), including a 2-year RCT in people with diabetes (Gepner et al., 2015). But these studies cannot tell us whether benefits would accrue in the long term, a question that can only be assessed with long-term exposure and follow-up. ...
Article
Full-text available
Heavy and disordered alcohol consumption is a known risk factor for several health conditions and is associated with considerable disease burden. However, at low‐to‐moderate levels, evidence suggests that drinking is associated with reduced risk for certain health outcomes. Whether these findings represent genuine protective effects or mere methodological artifacts remains unclear, but has substantial consequences for policy and practice. This critical review introduces methodological advances capable of enhancing causal inference from observational research, focusing on the ‘G‐methods’ and Mendelian Randomization. We also present and evaluate recent research applying these methods and compare findings to the existing evidence base. Future directions are proposed for improving our causal understanding of the relationships between alcohol and long‐term health outcomes.
Chapter
Polyphenol components of red wine have notable antioxidant and anti-inflammatory properties. Additionally, they interact with many intracellular signaling pathways that mediate fundamental cellular processes like survival and death, differentiation, proliferation and senescence. Although polyphenols have poor solubility and low bioavailability, they may accumulate in tissues and may act synergistically, therefore can account for the beneficial biological effects that are attributed to them. One of the most efficient cytoprotective pathways, the Akt signaling pathway, was demonstrated to exert massive protective effect in cardiovascular diseases. This chapter attempts to provide a summary about polyphenol components of red wine, their effects on the cardiovascular system and their interaction with intracellular signaling mechanisms with special interest on the Akt pathway.
Article
Objective This study focused on the association of the Dietary Inflammatory Index (DII) with the risk of gestational diabetes mellitus (GDM) using data from the National Health and Nutrition Examination Survey (NHANES). Methods Data from two cycles of NHANES (2007–2010) were analyzed. Weighted logistic regression models were constructed to explore the relationship between DII and GDM. Stratified and subgroup analyses with adjustment for confounding factors were carried out to explore the association between DII and GDM. Mediation analysis was conducted to investigate potential mediating factors. Results A total of 1,421 samples were analyzed. Based on the weighted logistic regression model, a strong positive association was observed between DII score and GDM risk (OR: 1.14, 95%CI: 1.02–1.26, P<0.01), suggesting that a pro-inflammatory diet could considerably increase the likelihood of GDM occurrence. This positive association was particularly significant in populations who consume alcohol (OR:1.17, 95%CI: 1.03–1.34, P<0.01) and have a body mass index>30 kg/m² (OR: 1.18, 95% CI: 1.01–1.38, P<0.05). Mediation analysis indicated that the white blood cell count mediated the association between DII and GDM, with an indirect effect of 0.000828 (95% CI: 0.000226–0.001347, P=0.001) and a mediation proportion of 8.01%. Specifically, the neutrophil count as a component contributed to this mediation effect (0.000774, 95% CI: 0.00014–0.001063, P=0.001), with a mediation proportion of 7.59%. Conclusion A pro-inflammatory diet in women elevates the likelihood of GDM occurrence. An anti-inflammatory dietary pattern may be a protective intervention for GDM.
Article
Full-text available
For some classes of dietary polyphenols, there are now sufficient intervention studies to indicate the type and magnitude of effects among humans in vivo, on the basis of short-term changes in biomarkers. Isoflavones (genistein and daidzein, found in soy) have significant effects on bone health among postmenopausal women, together with some weak hormonal effects. Monomeric catechins (found at especially high concentrations in tea) have effects on plasma antioxidant biomarkers and energy metabolism. Procyanidins (oligomeric catechins found at high concentrations in red wine, grapes, cocoa, cranberries, apples, and some supplements such as Pycnogenol) have pronounced effects on the vascular system, including but not limited to plasma antioxidant activity. Quercetin (the main representative of the flavonol class, found at high concentrations in onions, apples, red wine, broccoli, tea, and Ginkgo biloba) influences some carcinogenesis markers and has small effects on plasma antioxidant biomarkers in vivo, although some studies failed to find this effect. Compared with the effects of polyphenols in vitro, the effects in vivo, although significant, are more limited. The reasons for this are 1) lack of validated in vivo biomarkers, especially in the area of carcinogenesis; 2) lack of long-term studies; and 3) lack of understanding or consideration of bioavailability in the in vitro studies, which are subsequently used for the design of in vivo experiments. It is time to rethink the design of in vitro and in vivo studies, so that these issues are carefully considered. The length of human intervention studies should be increased, to more closely reflect the long-term dietary consumption of polyphenols.
Article
Full-text available
Objective To use the rs1229984 variant in the alcohol dehydrogenase 1B gene (ADH1B) as an instrument to investigate the causal role of alcohol in cardiovascular disease. Design Mendelian randomisation meta-analysis of 56 epidemiological studies. Participants 261 991 individuals of European descent, including 20 259 coronary heart disease cases and 10 164 stroke events. Data were available on ADH1B rs1229984 variant, alcohol phenotypes, and cardiovascular biomarkers. Main outcome measures Odds ratio for coronary heart disease and stroke associated with the ADH1B variant in all individuals and by categories of alcohol consumption. Results Carriers of the A-allele of ADH1B rs1229984 consumed 17.2% fewer units of alcohol per week (95% confidence interval 15.6% to 18.9%), had a lower prevalence of binge drinking (odds ratio 0.78 (95% CI 0.73 to 0.84)), and had higher abstention (odds ratio 1.27 (1.21 to 1.34)) than non-carriers. Rs1229984 A-allele carriers had lower systolic blood pressure (−0.88 (−1.19 to −0.56) mm Hg), interleukin-6 levels (−5.2% (−7.8 to −2.4%)), waist circumference (−0.3 (−0.6 to −0.1) cm), and body mass index (−0.17 (−0.24 to −0.10) kg/m2). Rs1229984 A-allele carriers had lower odds of coronary heart disease (odds ratio 0.90 (0.84 to 0.96)). The protective association of the ADH1B rs1229984 A-allele variant remained the same across all categories of alcohol consumption (P=0.83 for heterogeneity). Although no association of rs1229984 was identified with the combined subtypes of stroke, carriers of the A-allele had lower odds of ischaemic stroke (odds ratio 0.83 (0.72 to 0.95)). Conclusions Individuals with a genetic variant associated with non-drinking and lower alcohol consumption had a more favourable cardiovascular profile and a reduced risk of coronary heart disease than those without the genetic variant. This suggests that reduction of alcohol consumption, even for light to moderate drinkers, is beneficial for cardiovascular health.
Article
Full-text available
Epidemiological and clinical studies have over the years established that dyslipidemia constitutes the main risk factor for atherosclerosis. The inverse correlation between HDL cholesterol (HDL-C) levels and coronary heart disease morbidity and mortality identified HDL-C as an alternative pharmacological target to LDL-C and a potential anti-atherosclerosis marker. However, more recent data reinforced the principle of 'HDL quality' in atherosclerosis that refers to the functionality of HDL particle, as defined by its protein and lipid content, rather than HDL-C levels in plasma. Since HDL functionality depends on the genes and proteins of the HDL metabolic pathway, its apoprotein composition may serve as a surrogate marker of atheroprotection. In this manuscript we review the atheroprotective properties of HDL in relation to the proteins of HDL metabolic pathway and discuss what HDL-associated genes and proteins may reveal about HDL functionality in the assessment of coronary risk.
Article
Full-text available
Background: The enhanced vulnerability of women to develop alcohol-related diseases may be due to their higher blood alcohol levels after drinking, but the mechanism for this effect is debated. Methods: Sixty-five healthy volunteers of both genders drank 0.3 g of ethanol/kg of body weight (as 5%, 10%, or 40% solutions) postprandially. Blood alcohol concentrations were monitored by breath analysis and compared with those after intravenous infusion of the same dose. First-pass metabolism was quantified (using Michaelis-Menten kinetics) as the route-dependent difference in the amount of ethanol reaching the systemic blood. Gastric emptying was assessed by nuclear scanning after intake of 300 μCurie of technetium-labeled diethylene triamine pentacetic acid in 10% ethanol. The activities of alcohol dehydrogenase isozymes were assessed in 58 gastric biopsies, using preferred substrates for γ-ADH (acetaldehyde) and for ς-ADH (m-nitrobenzaldehyde) and a specific reaction of χ-ADH (glutathione-dependent formaldehyde dehydrogenase). Results: Women had less first-pass metabolism than men when given 10% or 40%, but not 5%, alcohol. This was associated with lower gastric χ-ADH activity; its low affinity for ethanol could explain the greater gender difference in first-pass metabolism with high rather than with low concentrations of imbibed alcohol. Alcohol gastric emptying was 42% slower and hepatic oxidation was 10% higher in women. A 7.3% smaller volume of alcohol distribution contributed to the higher ethanol levels in women, but it did not account for the route-dependent effects. Conclusions: The gender difference in alcohol levels is due mainly to a smaller gastric metabolism in females (because of a significantly lesser activity of χ-ADH), rather than to differences in gastric emptying or in hepatic oxidation of ethanol. The concentration-dependency of these effects may explain earlier discrepancies. The combined pharmacokinetic differences may increase the vulnerability of women to the effects of ethanol.
Article
Polyphenols are abundant micronutrients in our diet, and evidence for their role in the prevention of degenerative diseases is emerging. Bioavailability differs greatly from one polyphenol to another, so that the most abundant polyphenols in our diet are not necessarily those leading to the highest concentrations of active metabolites in target tissues. Mean values for the maximal plasma concentration, the time to reach the maximal plasma concentration, the area under the plasma concentration-time curve, the elimination half-life, and the relative urinary excretion were calculated for 18 major polyphenols. We used data from 97 studies that investigated the kinetics and extent of polyphenol absorption among adults, after ingestion of a single dose of polyphenol provided as pure compound, plant extract, or whole food/beverage. The metabolites present in blood, resulting from digestive and hepatic activity, usually differ from the native compounds. The nature of the known metabolites is described when data are available. The plasma concentrations of total metabolites ranged from 0 to 4 mumol/L with an intake of 50 mg aglycone equivalents, and the relative urinary excretion ranged from 0.3% to 43% of the ingested dose, depending on the polyphenol. Gallic acid and isoflavones are the most well-absorbed polyphenols, followed by catechins, flavanones, and quercetin glucosides, but with different kinetics. The least well-absorbed polyphenols are the proanthocyanidins, the galloylated tea catechins, and the anthocyanins. Data are still too limited for assessment of hydroxycinnamic acids and other polyphenols. These data may be useful for the design and interpretation of intervention studies investigating the health effects of polyphenols.
Article
Methods for the quantification of beta-cell sensitivity to glucose (hyperglycemic clamp technique) and of tissue sensitivity to insulin (euglycemic insulin clamp technique) are described. Hyperglycemic clamp technique. The plasma glucose concentration is acutely raised to 125 mg/dl above basal levels by a priming infusion of glucose. The desired hyperglycemic plateau is subsequently maintained by adjustment of a variable glucose infusion, based on the negative feedback principle. Because the plasma glucose concentration is held constant, the glucose infusion rate is an index of glucose metabolism. Under these conditions of constant hyperglycemia, the plasma insulin response is biphasic with an early burst of insulin release during the first 6 min followed by a gradually progressive increase in plasma insulin concentration. Euglycemic insulin clamp technique. The plasma insulin concentration is acutely raised and maintained at approximately 100 muU/ml by a prime-continuous infusion of insulin. The plasma glucose concentration is held constant at basal levels by a variable glucose infusion using the negative feedback principle. Under these steady-state conditions of euglycemia, the glucose infusion rate equals glucose uptake by all the tissues in the body and is therefore a measure of tissue sensitivity to exogenous insulin.
Article
Background & aims The results of human clinical trials that have investigated the effects of resveratrol on blood pressure are inconsistent. We aimed to quantitatively evaluate the effects of resveratrol on systolic blood pressure (SBP) and diastolic blood pressure (DBP). Methods We conducted a strategic literature search of PubMed, EMBASE, MEDLINE, and the Cochrane Library (updated to January, 2014) for randomized controlled trials that evaluate the effects of resveratrol on SBP and DBP. Study quality was assessed using the Jadad scale. Weighted mean differences were calculated for net changes in SBP and DBP using fixed-effects or random-effects models. We performed pre-specified subgroup, sensitivity and meta-regression analyses to evaluate potential the heterogeneity. Dose effects of resveratrol on SBP and DBP were estimated using meta-regression analyses. Results Six studies comprising a total of 247 subjects were included in our meta-analysis. The overall outcome of the meta-analysis indicates that resveratrol consumption can not significantly reduce SBP and DBP. Subgroup analyses indicated that higher-dose of resveratrol consumption (≥ 150 mg/d) significantly reduces SBP of -11.90 mm Hg (95% CI: -20.99, -2.81 mmHg, P =0.01), whereas lower dose of resveratrol did not show a significant lowering effect on SBP. The meta-regression analyses did not indicate dose effects of resveratrol on SBP or DBP. Conclusions The present meta-analysis indicates that resveratrol consumption significantly decreases the SBP level at the higher dose, while resveratrol has no significant effects on DBP levels. Additional high-quality studies are needed to further evaluate the causal conclusions.
Article
A systematic review and meta-analysis of available evidence was conducted to obtain a conclusive result on the lipid-modulating effects of resveratrol. Seven randomized controlled trials with a total of 282 subjects (141 in each group) met the eligibility criteria. Overall, resveratrol supplementation had no significant effect on any of the lipid parameters assessed: total cholesterol (weighted mean difference [WMD] -8.70; 95% confidence interval [95%CI] -21.54-4.14; P = 0.18), low-density lipoprotein cholesterol (WMD -3.22; 95%CI -12.56-6.12); P = 0.50), high-density lipoprotein cholesterol (WMD -0.26; 95%CI -4.25-3.73; P = 0.90), and triglycerides (WMD -4.30; 95%CI -20.22-11.63; P = 0.60). These results were robust in sensitivity analysis and were not dependent on the resveratrol dose, the duration of supplementation, or the cardiovascular risk status of the population studied. While future large-scale, well-designed trials are warranted, the current evidence suggests that mechanisms other than hypolipidemic effects account for the established cardioprotective properties of resveratrol.
Article
J Clin Hypertens (Greenwich). 2012;14:792–798. ©2012 Wiley Periodicals, Inc. Heavy alcohol intake increases the risk of hypertension, but the relationship between light to moderate alcohol consumption and incident hypertension remains controversial. The authors sought to analyze the dose-response relationship between average daily alcohol consumption and the risk of hypertension via systematic review and meta-analysis. Electronic databases were searched for prospective control studies examining quantitative measurement of alcohol consumption and biological measurement of outcome. The primary endpoint was the risk of developing hypertension based on alcohol consumption. The level of alcohol consumption from each study was assigned to categorical groups based on the midpoint of their alcohol consumption classes to make possible the comparison of heterogeneous classification of alcohol intake. A total of 16 prospective studies (33,904 men and 193,752 women) were included in the analysis. Compared with nondrinkers, men with alcohol consumption with <10 g/d and 11 to 20 g/d had a trend toward increased risk of hypertension (relative risk [RR], 1.03; 95% confidence interval [CI], 0.94–1.13; P=.51) and (RR, 1.15; 95% CI, 0.99–1.33; P=.06), respectively, whereas a significantly increased risk of hypertension was found with heavy alcohol consumption of 31 to 40 g/d (RR, 1.77; 95% CI, 1.39–2.26; P<.001) and >50 g/d (RR, 1.61; 95% CI, 1.38–1.87; P<.001). Among women, the meta-analysis indicated protective effects at <10 g/d (RR, 0.87; 95% CI, 0.82–0.92; P<.001) and a trend toward decreased risk of hypertension with alcohol consumption 11 to 20 g/d (RR, 0.9; 95% CI, 0.87–1.04; P=.17), whereas a significantly increased risk of hypertension was indicated with heavy alcohol consumption of 21 to 30 g/d (RR, 1.16; 95% CI, 0.91–1.46; P=.23) and 31 to 40 g/d (RR, 1.19; 95% CI, 1.07–1.32; P=.002). In men, heavy alcohol consumption is associated with increased risk of hypertension, whereas there is a trend toward increased risk of hypertension with low and moderate alcohol consumption. The relationship between alcohol consumption and hypertension is J-shaped in women. Limiting alcohol intake should be advised for both men and women.