Long-term effects of lifestyle on multiple risk factors in male workers
Environ Health Prev Med
Long-term effects of lifestyle on multiple risk factors in male workers
Hanayo Koetaka 0 1
Yuko Ohno 0 1
Kanehisa Morimoto 0 1
0 K. Morimoto (&) Department of Social and Environmental Medicine, Osaka University Graduate School of Medicine , 2-2 Yamadaoka, Suita, Osaka 565-0871 , Japan
1 H. Koetaka Y. Ohno Department of Mathematical Health Science, Osaka University Graduate School of Medicine , Suita, Osaka , Japan
Objectives To examine the long-term effects of lifestyle on the recovery from risk factors of cardiovascular disease and to discuss the difference in the effects of lifestyle modification in subjects with a single risk factor and those with multiple risk factors. Methods We used checkup data compiled for 6477 male workers, aged 20-59 years in 1995, with risk factors of cardiovascular disease. The relation between the recovery from risk factors 9 years later and baseline lifestyles was examined by logistic regression according to the initial number of risk factors. Results Nine years following the baseline measurements, 1907 subjects had recovered from at least one risk factor. When there was initially a single risk factor, a good overall lifestyle was effective in the recovery [odds ratio (OR) 1.27; 95% confidence interval (CI) 1.03, 1.57], with maintaining good dietary habits (OR 1.24; 95% CI 1.07, 1.45) and moderate stress levels (OR 1.19; 95% CI: 1.03, 1.38) both found to be especially effective in the recovery. When there were multiple risk factors, although the effect of a good overall lifestyle on the recovery was less than that when there was only a single risk factor, non-smoking (OR 1.27; 95% CI 1.07, 1.51) and limiting working hours (OR 1.25; 95% CI 1.05, 1.49) were found to be effective. Conclusions Our results provide evidence that good lifestyles are effective in the recovery from multiple risk factors. Effects of lifestyle on recovery from multiple risk factors are different from effects on the recovery from a single risk factor, with the difference depending on the initial number of risk factors.
Health practice; Lifestyle; Long-term effects; Longitudinal studies; Metabolic syndrome; Workers
Metabolic syndrome is defined as a group of risk factors,
such as obesity, hypertension, dyslipidemia,
hyperglycemia, and other metabolic abnormalities, that can lead to the
development of serious diseases [
]. The etiology of
metabolic syndrome is complex, as it is determined by the
interplay of both genetic and environmental factors, such as
lifestyle . The long-term effects of lifestyle on the
prevention of a single risk factor are known [
], and it is
believed that a good lifestyle can prevent metabolic
syndrome. However, how lifestyles affect metabolic
syndrome would differ from how they affect a single risk
factor because metabolic syndrome is defined as a group of
multiple risk factors. Morimoto et al. examined the
relationship of life style to sister chromatid exchange [
natural killer cell activities [
], mutagen levels excreted
in the urine [
], and immunoglobulin E [
researchers provided evidence that a good lifestyle is
influential in the prognosis of lifestyle-related diseases,
such as cancer and allergic disorders. Within the context of
preventing metabolic syndrome, it is important to examine
the long-term effects of lifestyle on a group of multiple risk
factors and to clarify the effects of a good lifestyle. Many
studies of metabolic syndrome have been reported recently
on populations worldwide. In terms of the influences of
lifestyle on metabolic syndrome, however, few studies
have followed a large cohort over a long-term period [
], although several cross-sectional studies have
examined the relationship between lifestyle and metabolic
], and another study has investigated the
effects of lifestyle intervention on metabolic syndrome
improvement in patients with impaired plasma glucose
tolerance . In addition, almost all of these earlier
studies primarily focused on individual health practices,
such as smoking. Considering the complexity of human
lifestyles, the effect of overall lifestyle on a group of
multiple risk factors should be evaluated.
We have examined the long-term influences of overall
lifestyle and individual health practices on the recovery
from a group of multiple risk factors and discuss the
different effects of these influences on a single risk factor and
multiple risk factors, respectively.
Materials and methods
In Japan, employees generally undergo regular annual
checkups at their workplace or at workplace-designated
clinics. This study used checkup data collected and
managed by a Japanese foundation that performs medical
checkups under contract from various employers. The
checkups consisted of a physical examination, blood
pressure measurements, laboratory analyses of fasting
blood samples, and a questionnaire.
In 1995, 106,334 male workers aged 20–59 years
underwent checkups. Of these 106,334 subjects, 27,601
also underwent a checkup in 2004. We excluded subjects
with missing data on body mass index (BMI), blood
pressure, triglycerides or high-density
lipoprotein-cholesterol, fasting blood glucose, and eight lifestyles and those
who were in treatment for hypertension, dyslipidemia, or
hyperglycemia in 1995. The data from 13,394 subjects
were therefore analyzed.
The risk factors studied were obesity, hypertension,
dyslipidemia, and hyperglycemia. In choosing each risk factor,
we followed the criteria of the Japan Society for the Study
of Obesity, Japanese Society of Hypertension, Japan
Atherosclerosis Society, and Japan Diabetes Society
(Table 1). The number of risk factors each subject had in
1995 and 2004, respectively, and the differences in the
number of risk factors between these years were calculated.
Subjects were classified into five categories according to
the number of risk factors they had in 1995: zero risk
factors (no risk factor), one risk factor (a single risk factor),
and two, three, or four risk factors (multiple risk factors).
To examine the influences of lifestyle in terms of the
initial number of risk factors, relations between the
decrement in the number of risk factors after 9 years and initial
lifestyle were analyzed in the categories of one (single risk
factor) to four (multiple risk factors) risk factors. The
effects of lifestyle were also analyzed in groups of
categories of two to four risk factors (C2) and three to four risk
factors (C3) to examine their influences on a group of
multiple risk factors and severe multiple risk factors.
Eight health practices and health practice index
Lifestyle factors were assessed from a self-administered
questionnaire that asked about eight health practices that
had been used by Morimoto et al. in their earlier studies on
7–9, 11, 19
]. Subjects answered the eight
questions shown in Table 2 by either ‘‘true’’ or ‘‘false’’. For
each factor, a score of 1 was given for true and 0 for false.
In addition, the health practice index (HPI) was calculated
by summing scores for the eight health practices to give a
total score ranging from 0 to 8 points. Subjects were
classified into three categories by the HPI: good health
practices (6–8 points), moderate health practices (4–5
points), and poor health practices (0–3 points).
Each factor was scored 1 for true and 0 for false. The health practice
index (HPI) was calculated by summing the scores for the eight health
practices. HPI categories were good health practice (6–8 points),
moderate health practice (4–5 points), and poor health practice (0–3
Two logistic regression models were used in six groups of
risk factors (1, 2, 3, 4, C2, C3, respectively) to evaluate
relations between the recovery from risk factors and
lifestyles. We defined recovery from risk factors as a decrease
of at least one risk factor in the years between 1995 and
2004 and non-recovery as an increase/no change in the
number of risk factors. Using Model 1, we examined the
influence of overall lifestyle on the recovery from risk
factors. The HPI, age, and the presence of risk factors
(obesity, hypertension, dyslipidemia, and hyperglycemia)
in 1995 were included. Age was treated as a categorical
variable classified in 10-year cohorts. Risk factors were
treated as categorical variables (normal or abnormal for
each risk factor according to the criteria in Table 1). The
HPI was treated with three categorical variables. Odds
ratios (ORs) and 95% confidence intervals (CIs) were
computed for the good health practice group and the
moderate health practice group of the HPI, and the poor
health practice group was used as the reference group. A
linear trend across categories was tested with the HPI score
treated as a continuous variable. Using Model 2, we
examined the influences of the eight health practices on the
recovery from risk factors. A stepwise logistic regression
analysis with P \ 0.20 for inclusion as a determining
factor was performed. The eight health practices (good
health practice or poor health practice), age, and the
presence of risk factors in 1995 were included. Variables
other than health practices were treated as in Model 1. The
ORs and CIs were computed for the good health practice
group of each health practice, and the poor health practice
group was used as the reference group. All analyses were
performed using SAS statistical software, ver. 8 (SAS
Institute, Cary, NC).
This study was approved by the ethical committee of the
Course of Health Science, Osaka University Graduate
School of Medicine.
Table 3 shows the baseline characteristics of subjects
according to the number of risk factors in 1995. The mean
age of all subjects was 40.6 years. The proportion of
subjects without risk factors was 51.6%, and the proportions of
subjects with one or more risk factor, two or more risk
factors, three or more risk factors, and four risk factors
were 48.4, 17.5, 3.8, and 0.2%, respectively. In the
singlerisk-factor group, dyslipidemia had the highest prevalence
among the four risk factors. In the 2-risk-factor group, the
prevalence of the combination of dyslipidemia and obesity
was about 50%, and in the 3-risk-factor group, the
prevalence of the combination of dyslipidemia and obesity and
hypertension was about 80%.
Among 6477 subjects (excluding subjects in the
0-riskfactor group), 1907 subjects had recovered from at least
one risk factor 9 years later. Table 4 shows the ORs for the
recovery from risk factors according to the baseline HPI
and eight health practices. The result for the 4-risk-factor
group is not shown because there were so few subjects and
the reliability of the analysis was low. In the single- and
2risk-factor groups, the ORs of recovery increased as the
HPI increased (P for trend 0.01 and 0.02, respectively). In
the single-risk-factor group, the OR of recovery was higher
in the good health practice group than in the poor health
practice group (OR 1.27; 95% CI 1.03, 1.57). In the 2- and
C2-risk-factor groups, the ORs of recovery were higher in
the moderate health practice group (OR 1.29; 95% CI 1.04,
1.60 and OR 1.21; 95% CI 1.00, 1.45, respectively). In
Model 2, we found variables with a significant relation to
the recovery in some groups. In the single-risk-factor
group, those who ate nutritionally balanced diets or those
who maintained moderate levels of mental stress had
higher ORs of recovery than each of those groups with
poor health practices (OR 1.24; 95% CI 1.07, 1.45 and OR
1.19; 95% CI 1.03, 1.38, respectively). In the 2- and
C2risk-factor groups, non-smokers (OR 1.26; 95% CI 1.03,
1.53, and OR 1.27; 95% CI 1.07, 1.51, respectively) or
those who worked fewer than 9 h per day (OR 1.25; 95%
CI 1.02, 1.54 and OR 1.25; 95% CI 1.05, 1.49,
respectively) had higher ORs of recovery. In the 3- and
C3-riskfactor groups, we found no variables with a significant
relation to the recovery.
This study was initiated with the aim of clarifying how
lifestyle at baseline influences the recovery from risk
factors after 9 years. The results show that the relation
between the recovery from risk factors and overall lifestyle
differs according to the initial number of risk factors. In the
a Data are given as the mean with the standard deviation in parenthesis
b Data are given as frequency with the percentage in parenthesis
single-risk-factor group, the good health practice group had
a significantly higher OR of recovery than the poor health
practice group, and there was a significant linear trend,
indicating that a good overall lifestyle is effective in terms
of recovery from a single risk factor. In the 2- and
C2-riskfactor groups, the moderate health practice group had
significantly higher ORs of recovery than the poor health
practice group. However, a significant linear trend was
obtained only in the 2-risk-factor group. These results
suggest that the effect of a good overall lifestyle on the
recovery decreases for subjects with severe multiple risk
factors, although a good overall lifestyle is effective for the
recovery for subjects with two risk factors.
In almost all multiple-risk-factor groups, non-smoking
was positively associated with recovery from risk factors.
Non-smokers often show about a 1.2-fold higher recovery
than smokers. Some studies have reported that
non-smokers have a lower incidence of metabolic syndrome than
those with other smoking statuses [
13–15, 17, 20
results of our study agree with these findings, suggesting
that non-smoking is important in the recovery from
multiple risk factors.
Working 9 h or less per day was positively associated
with recovery from risk factors in the 2- and C2-risk-factor
groups. However, associations were not obtained for
subjects with severe multiple risk factors, suggesting that
working B9 h per day is important for subjects with two
risk factors, although the effect of working B9 h per day
decreases for subjects with severe multiple risk factors.
In the single-risk-factor group, the incidence of recovery
in subjects with good health practices who eat a
nutritionally balanced diet or maintain a moderate level mental
stress was higher than that in subjects with poor health
practices. However, these lifestyles were not associated
with recovery from risk factors in the multiple-risk-factor
groups. This result suggests that, although it is possible for
subjects with a single risk factor to recover from risk
factors by maintaining good eating habits and moderate
fe .4 .8
re ec il 9 8
s M P r 2 1 6 l 3 3
c e *
a v 7 7
f I o .5 .2
C f 1 1
y r R
r to O
e c e
ov fa fo ) in
trceeh ilrsek lecean (reyv% 9 taexom
ro ing rev eco ,12 26 le .55 .32
f S P r 1 4 9 d 2 2
N N N M
mental stress levels, it is difficult for subjects with multiple
risk factors to recover from risk factors solely by
maintaining these health practices.
Exercising at least twice a week was negatively
associated with recovery from the risk factor in the
single-riskfactor group. There were also no associations with the
recovery in multiple-risk-factor groups. Generally, regular
physical exercise is recommended for the prevention of
metabolic diseases and metabolic syndrome, and positive
effects of physical exercise on metabolic syndrome have
been reported in previous studies [
]. We believe
that discontinuous exercise habits can explain our results.
The number of subjects who changed their health practices
was not small because the follow-up period was long in this
study. We examined the continuousness of the lifestyle by
using longitudinal cohort data, which form the basis of the
data used in this study, and found that exercise habits are
not continued easily compared with the other seven health
practices. For example, the continuance rate of good health
practice over the 9 years was 13.7% for exercise habits
compared to 30.2 and 19.5% for eating a nutritionally
balanced diet and working time, respectively. To clarify the
effect of exercise habits, it would be necessary to consider
changes in lifestyle over the 9 years.
We found that the relation between the recovery from risk
factors and lifestyle differed depending on the initial number
of risk factors each subject had. For subjects who initially
had a single risk factor, a good overall lifestyle was found to
promote the recovery from the risk factors. In this context,
maintaining good dietary habits and managing stress were
found to be especially effective in the recovery. For subjects
who initially had multiple risk factors, a good overall
lifestyle was found to promote the recovery, with non-smoking
and limiting working hours especially effective. However,
the relation between recovery from multiple risk factors and
maintaining good dietary habits and managing stress was
found to be less in these subjects than in those with a single
risk factor. This result suggests that the effects of dietary
modification and mental stress reduction were less in
subjects who initially had multiple risk factors than in those with
a single risk factor. For subjects with severe multiple risk
factors, no significant relations were obtained, possibly due
to methodological issues. Only the age and risk factors were
adjusted in this study; consequently, other factors, such as
family history, may have influenced the results. The number
of subjects who had treatments for risk factors cannot be
considered to be too small as the follow-up period was quite
long. A consideration of data during the follow-up period and
other various factors would be necessary to examine the
long-term influences of lifestyle on severe multiple risk
factors. However, our results do show that the mechanism of
a group of multiple risk factors is different from the
mechanism of a single risk factor, and they have clarified one of the
methods for recovery from the group of the multiple risk
The large study sample and the long-term follow-up
are the strengths of our study; however, there are several
limitations. Firstly, much of the data were excluded due
to missing values, subjects dropping out of the follow-up
due to resignation or job changes, subjects not going for
their medical checkup, among others. Accordingly,
subjects who underwent checkups in 2004 represented 26%
of all subjects who underwent medical checkups in 1995.
Compared with non-participants who had missing values
or dropped out, there was a trend for participants to be in
the 30- to 40-year age group and to have a better lifestyle.
We did not adjust for any bias of lifestyle although age
was adjusted. A non-participant’s ratio of poor lifestyle
was higher than that of a participant. As such a
nonparticipant has a higher chance of dropping out due to
deterioration in metabolic syndrome caused by poor
lifestyle. It is believed that the recovery rate in the poor
lifestyle group would be lower if it could be followed for
9 years and included non-participants. Therefore, in this
study there is a possibility that the estimated difference of
the effect between a good lifestyle and poor lifestyle is
smaller than the actual difference.
Only data from 1995 and 2004 were examined.
Therefore, changes in lifestyle during this period were not
considered. In addition, the causal relation between a
change in health status and changes in lifestyle was not
clarified. Although lifestyle intervention was not given at
the checkup, changes in lifestyle during this period could
have occurred. Such changing patterns in health status and
lifestyles should be examined, and this causal relation
should be clarified.
A number of definitions of metabolic syndrome have
been reported in studies worldwide [
]. In Japan, the
National Metabolic Syndrome Criteria study Group has
proposed new criteria for metabolic syndrome in Japan
. However, the criteria are still being discussed because
evidence in the Japanese population is at yet insufficient.
Therefore, we used the definitions of abnormality in the
criteria of each Japan Society as definitions of risk factors
of metabolic syndrome. The National Cholesterol
Education Program Adult Treatment Panel III [
] and the
Japanese criteria emphasize the presence of visceral
obesity and include measurements of waist circumference (WC)
instead of BMI. Visceral obesity has been found to have a
higher correlation with cardiovascular risk factors, such an
insulin resistance, hyperinsulinemia, glucose intolerance,
type-2 diabetes, dyslipidemia, and hypertension, when
compared to obesity defined by BMI [
]. In this study,
BMI was used instead of WC because WC data were not
collected. In Japanese checkups, WC measurements are not
regularly carried out. Therefore, existing data cannot be
easily used to estimate metabolic syndrome prevalence by
criteria including visceral obesity.
Recent studies have reported an association between
genetic type and metabolic syndrome [
effect of lifestyle on hypertension according to genetic type
has been previously clarified . Future studies should
clarify the effect of lifestyle on the recovery from
metabolic syndrome taking genetic type into consideration.
For the prevention of metabolic syndrome, it is
important to clarify factors that influence not only the recovery
but also deterioration. We are now examining factors that
influence deterioration and plan to report our findings in the
In conclusion, we have shown that there is relation
between the recovery from a group of multiple risk factors
and lifestyle. This relation is different from that for the
recovery from a single risk factor, and it differs depending
on the number of risk factors. Our results provide evidence
that effective modifications in lifestyles for people with a
group of multiple risk factors are different from effective
modifications for people with a single risk factor.
1. Gami AS , Witt BJ , Howard DE , Erwin PJ , Gami LA , Somers VK , et al. Metabolic syndrome and risk of incident cardiovascular events and death: a systematic review and meta-analysis of longitudinal studies . J Am Coll Cardiol . 2007 ; 49 : 403 - 14 .
2. Takeuchi H , Saitoh S , Takagi S , Ohnishi H , Ohhata J , Isobe T , et al. Metabolic syndrome and cardiac disease in Japanese men: applicability of the concept of metabolic syndrome defined by the National Cholesterol Education Program-Adult Treatment Panel III to Japanese men-the Tanno and Sobetsu Study . Hypertens Res . 2005 ; 28 : 203 - 8 .
3. Yamada Y , Kato K , Hibino T , Yokoi K , Matsuo H , Segawa T , et al. Prediction of genetic for metabolic syndrome . Atherosclerosis . 2007 ; 191 : 298 - 304 .
4. Iso H , Date C , Wakai K , Fukui M , Tamakoshi A , Fukui M , et al. The relationship between green tea and total caffeine intake and risk for self-reported type 2 diabetes among Japanese adults . Ann Intern Med . 2006 ; 144 : 554 - 62 .
5. Nakanishi N , Okamoto M , Nakamura K , Suzuki K , Tatara K. Cigarette smoking and risk for hearing impairment: a longitudinal study in Japanese male office workers . J Occup Environ Med . 2000 ; 42 : 1045 - 9 .
6. Waki K , Noda M , Sasaki S , Matsumura Y , Takahashi Y , Isogawa A , et al. Alcohol consumption and other risk factors for selfreported diabetes among middle-aged Japanese: a populationbased prospective study in the JPHC study cohort I . Diabet Med . 2005 ; 22 : 323 - 31 .
7. Morimoto K , Takeshita T , Take-uchi T , Maruyama S , Ezoe S , Mure K , et al. Chromosome alterations in peripheral lymphocytes as indices of lifestyle and genotoxicity . Int Arch Occup Environ Health. 1993;65[Suppl 1 ]: S37 - 41 .
8. Kusaka Y , Kondou H , Morimoto K. Healthy lifestyles are associated with higher natural killer cell activity . Prev Med . 1992 ; 21 : 602 - 15 .
9. Morimoto K , Takeshita T , Inoue-Sakurai C , Maruyama S. Lifestyles and mental health status are associated with natural killer cell and lymphokine-activated killer cell activities . Sci Total Environ . 2001 ; 270 : 3 - 11 .
10. Morimoto K , Takeshita T , Nanno M , Tokudome S , Nakayama K. Modulation of natural killer cell activity by supplementation of fermented milk containing Lactobacillus casei in habitual smokers . Prev Med . 2005 ; 40 : 589 - 94 .
11. Mure K , Takeshita T , Takeuchi T , Morimoto K. Urinary mutagens and lifestyle factors . Prev Med . 1996 ; 25 : 569 - 74 .
12. Shirakawa T , Morimoto K. Lifestyle effect on total IgE. Lifestyles have a cumulative impact on controlling total IgE levels . Allergy . 1991 ; 46 : 561 - 9 .
13. Wannamethee SG , Shaper AG , Whincup PH . Modifiable lifestyle factors and the metabolic syndrome in older men: effects of lifestyle changes . J Am Geriatr Soc . 2006 ; 54 : 1909 - 14 .
14. Holme I , Tonstad S , Sogaard AJ , Larsen PGL , Haheim LL . Leisure time physical activity in middle age predicts the metabolic syndrome in old age: results of a 28-year follow-up of men in the Oslo study . BMC Public Health . 2007 ; 7 : 154 . doi: 10 .1186/ 1471 -2458-7-154.
15. Park HS , Oh SW , Cho SI , Choi WH , Kim YS . The metabolic syndrome and associated lifestyle factors among South Korean adults . Int J Epidemiol . 2004 ; 33 : 328 - 36 .
16. Villegas R , Creagh D , Hinchion R , O'Halloran D , Perry IJ . Prevalence and lifestyle determinants of the metabolic syndrome . Ir Med J. 2004 ; 97 : 300 - 3 .
17. Zhu S , St-Onge MP , Heshka S , Heymsfield SB . Lifestyle behaviors associated with lower risk of having the metabolic syndrome . Metabolism . 2004 ; 53 : 1503 - 11 .
18. Orchard TJ , Temprosa M , Goldberg R , Haffner S , Ratner R , Marcovina S , et al. The effect of metformin and intensive lifestyle intervention on the metabolic syndrome: the Diabetes Prevention Program randomized trial . Ann Intern Med . 2005 ; 142 : 611 - 9 .
19. Morimoto K. Lifestyle and health (in Japanese) . Jpn J Hyg . 2000 ; 54 : 572 - 91 .
20. Ishizaka N , Ishizaka Y , Toda E , Hashimoto H , Nagai R , Yamakado M. Association between cigarette smoking, metabolic syndrome, and carotid arteriosclerosis in Japanese individuals . Atherosclerosis . 2005 ; 181 : 381 - 8 .
21. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (2001) Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) . JAMA 285 : 2486 - 97 .
22. Alberti KG , Zimmet PZ . Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation . Diabet Med . 1998 ; 15 : 539 - 53 .
23. Matsuzawa Y. Metabolic syndrome: definition and diagnostic criteria in Japan (in Japanese) . J Jpn Soc Int Med . 2005 ; 94 : 188 - 203 .
24. Sonmez K , Akcakoyun M , Akcay A , Demir D , Duran NE , Genc¸bay M, et al. Which method should be used to determine the obesity, in patients with coronary artery disease? (body mass index, waist circumference or waist-hip ratio) . Int J Obes . 2003 ; 27 : 341 - 6 .
25. Wang Y , Rimm EB , Stampfer MJ , Willett WC , Hu FB . Comparison of abdominal adiposity and overall obesity in predicting risk of type 2 diabetes among men . Am J Clin Nutr . 2005 ; 81 : 555 - 63 .
26. Dalton M , Cameron AJ , Zimmet PZ , Shaw JE , Jolley D , Dunstan DW , et al. Waist circumference, waist-hip ratio and body mass index and their correlation with cardiovascular disease risk factors in Australian adults . J Intern Med . 2003 ; 254 : 555 - 63 .
27. Tanigawa T , Iso H , Yamagishi K , Muraki I , Kawamura N , Nakata A , et al. Association of lymphocyte sub-populations with clustered features of metabolic syndrome in middle-aged Japanese men . Atherosclerosis . 2004 ; 173 : 295 - 300 .
28. Gable DR , Hurel SJ , Humphries SE . Adiponectin and its gene variants as risk factors for insulin resistance, the metabolic syndrome and cardiovascular disease . Atherosclerosis . 2006 ; 188 : 231 - 44 .
29. Semple RK , Chatterjee VK , O'Rahilly S. PPAR gamma and human metabolic disease . J Clin Invest . 2006 ; 116 : 581 - 9 .
30. Itoh T , Matsumoto M , Nakamura M , Okada A , Shirahashi N , Hougaku H , et al. Effects of daily alcohol intake on the blood pressure differ depending on an individual's sensitivity to alcohol: oriental flushing as a sign to stop drinking for health reasons . J Hypertens . 1997 ; 15 : 1211 - 7 .