Published by Oxford University Press on behalf of the International Epidemiological Association The Author 2006; all rights reserved. Advance Access publication 17 July 2006
International Journal of Epidemiology 2006;35:1278–1285 doi:10.1093/ije/dyl148
Heritability of cigarette smoking and alcohol use in Chinese male twins: the Qingdao twin registry Christina N Lessov-Schlaggar,1* Zengchang Pang,2 Gary E Swan,1 Qian Guo,3 Shaojie Wang,2 Weihua Cao,4 Jennifer B Unger,3 C Anderson Johnson3 and Liming Lee4
Accepted
8 June 2006
Background China has the world’s largest concentration of smokers (350 million) and rising alcohol consumption, yet little is known about tobacco and alcohol use aetiology. In 2000, the Chinese National Twin Registry was established to provide a genetically informative resource for investigation of health behaviour including tobacco and alcohol use. Methods
Using standard twin methodology, this study aimed to examine the relative contribution of genetic and environmental influences on cigarette smoking and alcohol drinking in a sample of adult Chinese twins (n 5 1010 individual twins). More than half of the male twins were smokers (58%), and 32.5% reported alcohol consumption. Among male smokers, 46.4% smoked 20 or more cigarettes per day (heavy smokers) and among drinkers, 32.8% consumed one or more drinks per day. Nearly all female twins were non-smokers (99.2%) and non-drinkers (98.7%); therefore, genetic analysis was limited to male data.
Results
In men, current smoking was significantly heritable [75.1%, 95% confidence interval (CI) 56.7–87.5] with no evidence for a significant contribution of shared environmental effects. Heavy smoking was more strongly influenced by genes (66.2%, 95% CI 0–88.4) than shared environment (8.7%, 95% CI 0–71.0). Similarly, current drinking was more strongly influenced by genetic effects (59.5%, 95% CI 0–87.8) than by shared environmental effects (15.3%, 95% CI 0–72.1). Amount of alcohol consumed was influenced to a similar degree by genetic (42.4%, 95% CI 0–91.8) and shared environmental factors (39.2%, 95% CI 0–82.7).
Conclusions These results support findings from twins of Western origin on the aetiology of tobacco and alcohol use and encourage further work in Chinese twins. Keywords
China, twins, cigarettes, alcohol, heritability, genetics
There are more than 350 million tobacco smokers in China.1 Cigarette smoking in China is a major preventable cause of morbidity and mortality associated with coronary heart disease, chronic obstructive pulmonary disease, and lung cancer.2–4 In 1 2 3 4
SRI International, Center for Health Sciences, Menlo Park, CA, USA. Qingdao Center for Disease Prevention and Control, Qingdao, China. University of Southern California, Institute for Health Promotion and Disease Prevention Research, Alhambra, CA, USA. Peking University School of Public Health, Peking, China.
* Corresponding author. E-mail:
[email protected]
a study of over 150 000 Chinese adults over 40 years of age, heart disease, cerebrovascular disease, and malignancies together accounted for about two-thirds of total mortality.2 The same study showed that total mortality would be reduced by 10% if smoking was eliminated.2 Despite evidence to the contrary, the majority of Chinese smokers believe that smoking is not harmful to their health,5 though there has been improvement in knowledge of health consequences in recent years.1 Only 14–16% of current Chinese smokers report wanting to quit smoking,1,6,7 in contrast to 60–75% of adult smokers of Western origin who report a desire to quit
1278
HERITABILITY OF CIGARETTE SMOKING AND ALCOHOL USE IN CHINESE MALE TWINS
smoking.8,9 Tobacco use is an epidemic that will have, unless curbed, far-reaching effects on Chinese society for the foreseeable future.10–12 Alcohol use has not received much scientific investigation in China, particularly as a potential problem behaviour.13 Alcohol dependence has moved from the ninth to the third most prevalent mental illness over the past 40 years.13,14 Compared with individuals from Western countries, Chinese individuals consume less alcohol.15,16 Nevertheless, surveys of both younger and older Chinese reveal a substantial prevalence of past year alcohol use (60–70%).13,16–18 Research in adult twins of Western origin shows significant genetic influences on consumption of tobacco and alcohol19–29 as well as on nicotine and alcohol dependence.30–35 The generalizability of these results to twins of non-Western origin is largely unknown. One family study in a Chinese sample36 showed significant familial aggregation for nicotine dependence measured using the Fagerstro¨m Test for Nicotine Dependence (FTND37). A general population study of alcohol use and abuse showed an effect of maternal daily drinking on offspring alcohol abuse, though no such effect was seen for paternal daily drinking.13 Considering that daily drinking in Chinese women is a rare and extreme behaviour, it may reflect high genetic risk for alcoholism in these women, which was expressed as increased risk for alcohol abuse in their offspring. Such results support the role of familial influences on diagnostic tobacco and alcohol use phenotypes in Chinese individuals. In order to establish a resource for epidemiological research, Peking University School of Public Health initiated the Chinese National Twin Registry (CNTR) in 2000,38 which was intended to be a ‘long-term special population resource for genetic epidemiological studies of etiologically complex diseases and related phenotypes (p. 347)’. The primary focus of the scientific efforts will be on cardiovascular and cerebrovascular diseases and risk factors. The present study represents the first effort in characterizing the relative contribution of genetic and environmental influences on cigarette smoking and alcohol consumption in a sample of Chinese twins (n 5 505 twin pairs) from one of the four recruitment sites of the CNTR.
Methods Sample Qingdao, China, a city of 7 million is located on the Yellow Sea, at the southern tip of the Shandong Peninsula. In 2000, as part of the CNTR, the Qingdao Municipal Centers for Disease Control and Prevention (QDCDC) initiated recruitment of twins in the Shandong province as one of four recruitment areas. Qingdao Twin Registry twins were initially recruited through local disease control networks and through the mass media, and, in 2003, tracking of families with twins was begun using birth registry records. From the initial recruitment, in 2001, medical examination (including blood samples) and questionnaire/interview data were collected from 1010 individual female and male twins aged 24 years or older. Questionnaire assessment included a standard
1279
battery of twin similarity questions (from which a zygosity algorithm was derived for zygosity determination in the present investigation), demographic information, measures of personality, and cigarette smoking and alcohol use. Questionnaire-based zygosity assignment has 95% agreement with blood tests.39 A latent class analysis approach of zygosity assignment based on zygosity questionnaire items was nearly 100% accurate as compared with DNA-based zygosity assignment,40 lending further validity to the assignment of zygosity based on questionnaire items. Data were provided by the director of the QDCDC and one of the co-authors (Dr Pang). Study protocols were approved by the Ethics Committee for Human Subject Studies of the Peking University Health Science Center.38
Assessment Twins who responded to advertisements were scheduled for registration and assessment in pairs over the weekend days, to avoid conflict with the twins’ work schedules. Assessment was conducted for 20–25 twin pairs per day. Signed informed consent for study participation was obtained from the twins prior to data collection. Self-administered questionnaire assessment was conducted under study staff supervision. For the purpose of genetic analysis, cigarette smoking and alcohol use questionnaire data only from the male twins were used because of the low prevalence of cigarette smoking and alcohol drinking in the female twins. Cigarette and alcohol use questions were asked in an identical manner and were similarly defined. In total, four phenotypes were operationalized: a two-category current smoking measure with current smokers in one category defined as those who gave affirmative responses to ‘Do you smoke?’ (‘Yes, I smoke.’); and former smokers and non-smokers in the second category defined as those who responded ‘I have quit smoking for one month or more’ or ‘No, I don’t smoke.’ to the same question; a two-category heavy smoking measure was defined as smoking 20 or more cigarettes per day reported by current smokers and having smoked 20 or more cigarettes per day reported by former smokers (the daily smoking quantity item had five response categories: 11, 51, 101, 151, and 201 cigarettes per day); since alcohol use questions were asked in a parallel way to the cigarette smoking questions, a two-category current drinking measure was defined as current drinkers in one category who gave an affirmative response to ‘Do you drink alcohol’ and former and non-drinkers in a second category; and drinking quantity, defined as number of alcoholic drinks consumed per day reported by current drinkers for their current consumption levels and by former drinkers for their previous consumption level dichotomized by median split (the daily alcohol quantity item comprised of five response categories: 11, 21, 31, 41, and 51 drinks per day). Data for current smoking, current drinking, and drinking quantity were available from both twins of 120 monozygotic (MZ) and 72 dizygotic (DZ) male twin pairs. Data for heavy smoking were available from 103 MZ and 62 DZ male twin pairs.
1280
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Statistical analysis The twin method Using data from MZ and DZ twins reared together, individual variability in measured phenotypes can be attributed to at least three causal factors: (i) the additive effects of genes (A) passed on from parents to their offspring, which contribute to greater similarity between twin pair members since MZ co-twins share 100% of their genes and DZ co-twins share, on average, 50% of their genes; (ii) shared environmental influences common to twin pair members (C), such as factors in the family environment, school, or neighbourhood. Shared environmental influences also contribute to twin pair similarity. It is assumed that environmental factors are shared to the same extent in members of MZ and DZ twin pairs; and (iii) non-shared environmental influences (E), which are unique to each member of a twin pair and which also include measurement error. Non-shared environmental factors contribute to twin pair dissimilarity.41 The expected twin pair correlation for MZ twin pairs is computed as rMZ 5A 1 C, and that for DZ twin pairs is computed as rDZ 5 0.5A 1 C. Two times greater MZ relative to DZ twin pair correlation (rMZ/rDZ 5 2) indicates significant heritable influences and no significant influence of shared environmental effects (i.e. C 5 0). If shared environmental effects contribute to phenotypic twin pair similarity, the rMZ/rDZ ratio would be less than 2. Model fitting First, the degree of MZ and DZ twin pair similarity for each measure was computed using tetrachoric correlations for dichotomous measures. The computation of tetrachoric correlations assumes underlying normally distributed latent liability for each measure (e.g. smoking susceptibility), with theoretical thresholds along this continuum above which the phenotype (e.g. current smoking) is expressed. Tetrachoric correlations are computed using data from summary MZ and DZ twin pair 2 · 2 contingency tables. Next, univariate ACE models were fit to the same MZ and DZ twin pair contingency tables. The use of twin data permits estimation of the A, C, and E components from the observed within-twin and cross-twin variances and covariances in MZ and DZ twins. Solutions of the equations yield estimates for A, C, and E. Maximum likelihood estimates of these parameters were obtained using the structural equation modelling program Mx.42 Model fit was evaluated using chi-square goodness of fit and the Akaike Information Criterion (AIC43). A non-significant chi-square (P . 0.05) and a small AIC value (including negative estimates) indicate a good model fit. The significance of each of the A or C components was tested by comparing the full ACE model fit with the fit of submodels where A, or C, or both were equated to zero. A parameter was considered significant if the difference chi-square between two models was significant (P , 0.05) at the difference degrees of freedom between the same two models.
Results
Table 1 Demographic characteristics of male Qingdao Twin Registry twins aged 24 years and older Demographic characteristics
Male twins (n individual twins)
%
Sample size
486
MZ twins
241
49.6
DZ twins
245
46.8
Age 6 SD
38.9 6 10.3
Education (n 5 486) Illiterate/semiliterate Elementary school Middle school
8
1.7
71
14.6
240
49.4
High school
96
19.7
Undergraduate degree
68
14.0
3
0.6
46
9.5
438
90.1
2
0.4
Factory worker
152
33.6
Occupational/admin
Graduate degree or more Marital status (n 5 486) Single Married Widowed Occupation (n 5 453)
112
24.7
Businessman
63
13.9
Farmer
93
20.5
Student
0
0
Fisher
5
1.1
Other
28
6.2
twins are listed in Table 1. The average age was in the late 30s. Nearly half of the men had middle school education (49.4%) and another 19.7% had completed high school. The large majority (90.1%) were married at the time of survey. About a third reported being factory workers, 24.7% held occupational/ administrative positions, and 20.5% were farmers.
Prevalence of cigarette smoking and alcohol use Cigarette and alcohol use in the male twins are listed in Table 2. In this sample, virtually all the female twins were non-smokers (99.2%) and non-drinkers (98.7%). Nearly 58% of the men reported smoking at the time of survey, and an additional 8.4% reported having quit for 1 month or more. In comparison, the prevalence of smoking in the US is currently 24.1% among men and 19.2% among women.44 Among the male Chinese twin smokers, nearly half (46.4%) were heavy smokers reporting smoking 20 or more cigarettes per day. Nearly 33% of the men reported that they drink alcohol. Among these men, more than half reported drinking one or two drinks per day, and nearly a quarter (23.6%) reported drinking five or more drinks per day. In comparison, 62.8% of men and 45.9% of women in the US reported drinking alcohol in the past month.45
Demographic characteristics
Genetic analyses of smoking and alcohol use
There were slightly more female (51.9%) than male twins (48.1%) in this study. Demographic characteristics of the male
The number of twin pairs concordant and discordant for each category of each of the smoking and drinking measures are
HERITABILITY OF CIGARETTE SMOKING AND ALCOHOL USE IN CHINESE MALE TWINS
Table 2 Cigarette smoking and alcohol consumption in male Qingdao Twin Registry twins aged 24 and older Cigarette smoking and alcohol drinking
Table 4 Twin pair tetrachoric correlations (95% CI) for cigarette smoking and alcohol drinking phenotypes in male twins
Male twins (n individual twins)
%
282
58.0
41
8.4
163
33.5
Current smoker Former smoker (quit 11 month) Non-smoker Cigarettes/day (n 5 323) 1 or more
18
5 or more
29
9.0
10 or more
80
24.8
15 or more
46
14.2
150
46.4
20 or more Age started smoking 6 SD Years smoked 6 SD
19.4 6 10.3 158
32.5
16
3.3
312
64.2
1 or more
57
32.8
2 or more
35
20.1
3 or more
31
17.8
4 or more
10
5.7
5 or more
41
23.6
Formerly drank (11 month ago) Non-drinker Number of drinks/day (n 5 181)
Age started drinking 6 SD
24.4 6 6.3
Years drank 6 SD
16.7 6 10.1
The sample size is noted for measures where it is less than the total 486 individual twins.
Table 3 Summary MZ and DZ twin pair contingency tables for each measure
Measures
Smoking and drinking phenotypes
MZ twin pairs (n pairs 5 103–120)
DZ twin pairs (n pairs 5 62–72)
Current smoking
0.75 (0.56–0.88)
0.40 (0.04–0.68)
Heavy smoking
0.75 (0.53–0.89)
0.43 (0.04–0.73)
Current drinking
0.75 (0.55–0.88)
0.45 (0.09–0.72)
Amount drank
0.81 (0.62–0.92)
0.60 (0.26–0.82)
5.6
20.5 6 4.6
Currently drinks
1281
MZ Twin Pairs
DZ Twin Pairs
Twin 1
Twin 2
Twin 1
Twin 2
Current smoker
Twin 1
57
15
29
12
Non-smoker
Twin 2
12
36
14
17
Heavy smoker
Twin 1
56
11
29
13
Non-heavy smoker
Twin 2
11
25
8
12
Current drinker
Twin 1
27
11
13
11
Non-drinker
Twin 2
14
68
12
36
Amount drank (upper 50%)
Twin 1
83
9
42
9
Amount drank (lower 50%)
Twin 2
9
19
9
12
The frequencies shown are number of twin pairs.
shown in the summary MZ and DZ twin pair contingency tables in Table 3. MZ and DZ twin pair tetrachoric correlations and associated 95% confidence intervals (CIs) are shown in Table 4. All twin pair correlations were significantly different from zero. The nearly 100% increase in correlation in MZ relative to DZ twin pairs for current and heavy cigarette smoking and for current drinking suggests significant heritable influences on these phenotypes and small or non-significant
shared environmental effects. For alcohol quantity the DZ twin pair correlation was greater than half of the MZ twin pair correlation suggesting a significant contribution of shared environmental effects in addition to additive genetic influences. Parameter estimates, goodness of fit statistics, and likelihood ratio chi-square difference test statistics for the general ACE models and for the associated genetic (AE) or environmental (CE and E) submodels for each smoking and drinking measure are shown in Table 5. The best fitting model for each measure is shown in italics. Consistent with the pattern of twin pair correlations, current smoking was highly heritable with 75.1% of its phenotypic variance explained by additive genetic effects and no evidence for a significant contribution of shared environmental effects. As seen in the application of the likelihood ratio chi-square difference test, dropping C from the model did not result in significant deterioration of the AE model fit (P 5 0.898) relative to the full ACE model. In contrast, both environmental models (CE and E) fit the data significantly worse relative to the full ACE model as indicated by the significant P-values associated with the chi-square difference test (P , 0.05 for the CE model and P , 0.001 for the E model). Phenotypic variability of heavy smoking and current drinking was more strongly influenced by genetic effects relative to shared environmental effects (A 5 66.2% and C 5 8.7% for current smoking; A 5 59.5% and C 5 15.3% for current drinking). Drinking quantity was influenced by both genetic (42.4%) and shared environmental effects (39.2%). For these three measures, either A or C, but not both parameters, could be equated to zero without significant deterioration of model fit. These results suggest an important role of familial influences that combines the effects of genetics and shared environment. However, the relative contribution of genetic and shared environmental influences could not be distinguished, as evidenced by the zero lower bound of the CIs for both A and C parameter estimates in the full ACE models. Therefore, the three parameter ACE models were considered best fitting. For every measure, non-shared environmental influences were moderate (18.6–25.2%), further supporting the important contribution of genetic and shared environmental influences on these smoking and alcohol measures.
Discussion According to the fifth Chinese population census conducted in 2000, Shandong province has 90.8 million people, 50.6% male, with 3.3% college graduates, 11.0% high school graduates, and 36.6% who have completed junior high school (http://www. china.org.cn/english/features/51613.htm). In our sample, the sex distribution consisted of slightly more women than men
1282
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Table 5 Relative contribution of additive genetic (A), shared environmental (C), and non-shared environmental influences (E) on phenotypic variability of current smoking, heavy smoking (20 or more cigarettes per day), current drinking, and amount drank per day in Qingdao Twin Registry male twins Model parameters Measures Current smoking
Heavy smoking*
Current drinking*
Amount drank*
2
x difference test
Model fit statistics 2
c
2
Ddf
3.512
–
–
–
0.202
1.386
4.126
1
0.042
3
0.918
5.495
0.017
1
0.898 ,0.001
df
P
AIC
0.488
2
0.783
37.0 (23.3–54.7)
4.614
3
24.9 (12.5–43.3)
0.505
A (95% CI)
C (95% CI)
E (95% CI)
70.4 (2.4–87.3)
4.4 (0–63.6)
25.2 (12.6–44.5)
–
63.0 (45.3–76.7)
75.1 (56.7–87.5)
–
Dc
P
–
–
100
41.057
4
,0.001
33.057
40.569
2
66.2 (0–88.4)
8.7 (0–71.0)
25.2 (11.5–47.1)
1.202
2
0.548
2.798
–
–
–
–
63.6 (43.9–78.4)
36.4 (21.7–56.1)
4.242
3
0.236
1.758
3.040
1
0.081
75.3 (54.7–88.6)
–
24.7 (11.4–45.3)
1.254
3
0.740
4.746
0.052
1
0.819 ,0.001
–
–
100
34.754
4
,0.001
26.754
33.552
2
59.5 (0–87.8)
15.3 (0–72.1)
25.2 (12.1–45.4)
0.404
2
0.817
3.596
–
–
–
–
64.7 (46.6–78.4)
35.3 (21.6–53.4)
3.297
3
0.348
2.703
2.892
1
0.089
75.7 (56.9–88.2)
–
24.3 (11.9–43.2)
0.594
3
0.898
5.406
0.190
1
0.663
–
–
100
39.524
4
,0.001
31.524
39.120
2
,0.001
42.2 (0–91.8)
39.2 (0–82.7)
18.6 (7.5–38.0)
0
2
1.0
4.000
–
–
–
–
73.7 (56.8–85.6)
26.3 (14.4–43.2)
1.859
3
0.602
4.141
1.859
1
0.173
83.1 (66.0–93.1)
–
16.9 (7.0–34.0)
1.365
3
0.714
4.635
1.365
1
0.243
–
–
100
46.999
4
,0.001
38.999
46.999
2
,0.001
The most parsimonious models are shown in italics. * Either A or C, but not both, could be dropped from the model without significant deterioration of model fit.
but the deviation is not greater than the distribution of the general population of the Shandong province. This sample of twins was more highly educated than the general population of Shandong province. Rates of completion of high school and college in the twins are more similar to other general population surveys in China.16,46 The high proportion of married individuals in adult samples is consistent with other reports.16 Occupation categories differ across studies, but between 40 and 50% of working adults hold jobs in manufacturing (e.g. factory workers) and agriculture (e.g. farmers).16,46 In this sample of Qingdao twins, the proportions of MZ and DZ twins were similar to each other. Based on birth records and using a large sample of twins across several Chinese provinces, the overall MZ : DZ ratio was computed as 0.37; however, it ranged 0.22–1.42 across provinces.47 In the Qingdao twin sample the ratio is closer to 1, which falls within the reported range. Observed sex differences in the prevalence of current smoking are consistent with those reported in other studies of Chinese individuals.4,10,48,49 The prevalence of current smoking in the male twins (58.0%) was similar to that reported in other studies of Chinese men (61–63%),4,10,47,49 and the prevalence of male twins heavy smokers consuming 20 or more cigarettes per day (47.6%) also closely paralleled prevalence of heavy smoking reported elsewhere (46%).10 The greater proportion of male relative to female current drinkers in the twin sample is consistent with that reported elsewhere.13,16–18 Prevalence for current drinking in the Qingdao male twins is lower in this study than that reported in other studies (60–80%).13,16–18 The discrepancies in demographic and smoking and drinking characteristics between the Qingdao
twin sample on the one hand and census data and other studies, on the other, may be due to the relatively small size of the twin sample and as the twin registry grows, proportions will likely approach provincial and national averages. It is also possible that because the twins were recruited largely from the city of Qingdao, they were more representative of urban dwellers who have slightly different demographic characteristics (e.g. more education) than rural dwellers. Additionally, twins may differ from the general population by virtue of being considered a special population, an effect that may be accentuated in China, where single child homes predominate. Evidence for significant genetic and familial effects on smoking and alcohol use measures in this sample of adult Chinese male twins is consistent with that reported in adult twins of Western origin. Results for current smoking are particularly similar across studies, with evidence for significant heritability (39–70%) and no significant shared environmental effects.50–53 In twins of Western origin, quantity smoked is significantly heritable with estimates ranging 45–70% for number of cigarettes smoked per day19,21,35,54–56 and 49% for heavy smoking (301 cigarettes per day).57 Rarely have relatively modest shared environmental effects been reported (19%).54 Compared with cigarette smoking, a greater importance of shared environmental effects relative to genetic effects has been shown for alcohol consumption, consistent with results in this study. In studies of twins of Western origin, individual variability in current drinking and alcohol amount is both heritable (24–61%) and significantly influenced by shared environmental effects (27–45%), in some studies21,22,54,57 but not in all.23,55,56 Heavy drinking was significantly heritable
HERITABILITY OF CIGARETTE SMOKING AND ALCOHOL USE IN CHINESE MALE TWINS
(72%) with no evidence for significant influences from shared environmental effects,29 but we did not define a heavy drinking phenotype in this study. Overall, this study shows evidence for significant heritable influences on current smoking and significant familial influences on heavy smoking and alcohol consumption phenotypes. The results are consistent with those reported in studies of Western twins and suggest that the relative contribution of genetic and environmental factors on general cigarette smoking and alcohol use measures are similar across the Western and Chinese cultures, despite possible differences in the nature and variability of environmental influences across cultures. However, analysis incorporating data from both cultures is needed to test this observational inference directly. Similarity in heritable influences on smoking initiation and persistence has been shown for Australian, Swedish, and Finish twins,52,53 though the nature of shared environmental effects seems to differ across cultures particularly for men.53 Comparison between US and Australian twins found differences in genetic and environmental influences on smoking initiation.58 Incorporation of genetic polymorphisms and measured environmental factors in cross-cultural twin models will more clearly identify genetic and environmental causes of cultural differences of drug use behaviour. Several limitations of this study should be considered. First, the cigarette smoking and alcohol drinking measures were broadly and categorically defined. These definitions in combination with the available sample size may have contributed to limited power in our ability to discern the relative contribution of genetic and shared environmental sources of variance for smoking amount and the alcohol use phenotypes. However, we were limited to the available data and used those data to their fullest potential. The consistency of the genetic results in the Qingdao male twins and larger twin samples of Western origin suggest validity of these phenotypes. Second, combining former smokers (or former drinkers) in the same category with never smokers (or never drinkers) may bias the results since the smoking or drinking liability of never users is unknown. There is a small overlap in genetic risk between smoking initiation and smoking persistence, and heritability of current smoking remains high (42–70%) even after accounting for the genetic risk on smoking initiation,50–52 suggesting that bias of results in estimating genetic and environmental influences on current use that is not conditioned on exposure is probably small. In addition, in this investigation, it was of interest to capture all the existing variance, variance that determines liability to both smoking and non-smoking (or drinking and non-drinking), as well as optimize our statistical power. Further, the way the questionnaire items were phrased made it impossible to discern whether twins who reported ‘I don’t smoke’ or ‘I don’t drink’ were actually never-smokers or never-drinkers. In other words, we had no specific information on their exposure and it is possible that they did smoke or drink years prior to the survey. Third, biometric analysis was possible only in the male twins, therefore, these results may not be generalizable to females. Additionally, the sample may be more representative of urban dwellers. Finally, twin methodology has been criticized for the assumption that trait-relevant environmental influences
1283
contribute equally to the resemblance of MZ and DZ twin pairs.59,60 This assumption predicts that greater phenotypic similarity in MZ compared with DZ twin pairs is attributable to genetic factors. However, environmental influences that are more similar for MZ twin pairs may be associated with a trait of interest and could account for greater similarity between MZ than DZ twins inflating estimates of trait heritability. Studies that have examined the validity of the equal environment assumption have shown that zygosity differences in environmental factors such as self-perceived zygosity, parental rearing, physical and emotional closeness between co-twins, and shared social influences61–66 are uncorrelated with zygosity differences in a number of psychiatric disorders including nicotine and alcohol dependence, regular smoking, or amount smoked, supporting the validity of the equal environmental assumption in twin studies assessing these disorders. One violation of the equal environment assumption has shown that greater sharing of peer group in adolescence in MZ compared with DZ twins is associated with greater MZ compared with DZ twin similarity for smoking initiation in female twins.63 Initiation of substance use was not a phenotype that was analysed in the current study. In addition, the average age of our sample of male Qingdao twins (38.9 years of age) was greater than their reported ages of starting to smoke (20.5 years of age) and starting to drink (24.4 years of age), suggesting that, as a whole, the Qingdao twins were beyond risk for smoking or drinking initiation. Therefore, possible violation of the equal environment assumption, with respect to a correlation between zygosity differences in peer influences and zygosity differences in current use or amount of tobacco and alcohol use, would not be relevant in this study. The collaboration represented in this paper has resulted in the expansion of the questionnaire assessment survey for the Qingdao Twin Registry including more information about frequency and amount of tobacco and alcohol use, nicotine dependence (FTND),37 nicotine sensitivity, screening for alcoholism using the CAGE,67 peer influences, exposure to Western culture, family environment, and conditions often comorbid with tobacco use, such as depressive symptomatology and externalizing behaviour (e.g. aggression, hostility, hyperactivity). The twin registry will be expanded to include adolescent twins as well. As such data are collected and sample size increases, multivariate analyses will explore the genetic and environmental covariation of comorbid conditions (e.g. tobacco and alcohol use) and the mediational role of environmental influences (e.g. shared peers, SES) on expression of genetic risk for smoking and alcohol use.
Acknowledgements The authors are grateful to the twins for their valuable participation in this research. The authors also thank Lisa Jack for help with data assembly and analysis. Portions of this paper were presented as part of a Young Investigator paper session at the annual meeting of the Society for Research on Nicotine and Tobacco, March 2005, Prague, Czech Republic. This research was supported by the University of Southern California Transdisciplinary Tobacco Use Research Center (P50 CA84735); the Sidney Garfield Endowment; the China
1284
INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
Medical Board, New York (CMB01); and the School of Public Health, Beijing University.
22
23
References 1 2 3
4 5
6
7
8
9
10
11 12 13
14 15
16
17
18
19
20
21
Yang GH, Ma JM, Liu N, Zhou LN. Smoking and passive smoking in Chinese. Zhonghua Liu Xing Bing Xue Za Zhi 2005;26:77–83.
24
He J, Gu D, Wu X et al. Major causes of death among men and women in China. N Engl J Med 2005;353:1124–34.
25
Lam TH, Chung SF, Janus ED et al. Smoking, alcohol drinking and non-fatal coronary heart disease in Hong Kong Chinese. Ann Epidemiol 2002;12:560–67.
26
Zhang H, Cai B. The impact of tobacco on lung health in China. Respirology 2003;8:17–21.
27
Chinese Academy of Preventive Medicine. Smoking in China: 1996 National Prevalence Survey of Smoking Pattern. Beijing: China Science and Technology Press, 1997.
28
Gong YL, Koplan JP, Feng W, Chen CH, Zheng P, Harris JR. Cigarette smoking in China. Prevalence, characteristics, and attitudes in Minhang District. JAMA 1995;274:1232–34.
29
Yang G, Ma J, Chen A et al. Smoking cessation in China: findings from the 1996 national prevalence survey. Tob Control 2001;10: 170–74.
30
Jarvis MJ, Wardle J, Waller J, Owen L. Prevalence of hardcore smoking in England, and associated attitudes and beliefs: cross sectional study. BMJ 2003;326:1061.
31
Sheahan SL, Free TA, Rayens MK. Smoking behavior and desire to quit among low-income female caregivers. Appl Nurs Res 2003;16:156–63. Chen ZM, Xu Z, Collins R, Li WX, Peto R. Early health effects of the emerging tobacco epidemic in China. A 16-year prospective study. JAMA 1997;278:1500–04.
32
Mackay J. Battling upstream against the tobacco epidemic in China. Tob Control 1997;6:9–10.
33
O’Sullivan B, Chapman S. Eyes on the prize: transnational tobacco companies in China 1976–1997. Tob Control 2000;9:292–302.
34
Jiafang Z, Jiachun W, Yunxia L, Xiaoxia Q, Ya F. Alcohol abuse in a metropolitan city in China: a study of the prevalence and risk factors. Addiction 2004;99:1103–10.
35
Cochrane J, Chen H, Conigrave KM, Hao W. Alcohol use in China. Alcohol Alcohol 2003;38:537–42. White M, Harland JO, Bhopal RS, Unwin N, Alberti KG. Smoking and alcohol consumption in a UK Chinese population. Public Health 2001;115:62–69. Hao W, Su Z, Liu B et al. Drinking and drinking patterns and health status in the general population of five areas in China. Alcohol Alcohol 2004;39:43–52. Wei H, Derson Y, Xiao S, Li L, Zhang Y. Alcohol consumption and alcohol-related problems: Chinese experience from six area samples. Addiction 1999;94:1467–76. Abdullah AS, Fielding R, Hedley AJ. Patterns of cigarette smoking, alcohol use and other substance use among Chinese university students in Hong Kong. Am J Addict 2002;11:235–46. Kaprio J, Koskenvuo M, Sarna S. Cigarette smoking, use of alcohol, and leisure-time physical acitivity among same-sexed adult male twins. Prog Clin Biol Res 1981;69:37–46. Kaprio J, Koskenvuo M, Langinvainio H, Romanov K, Sarna S, Rose RJ. Genetic influences on use and abuse of alcohol: a study of 5638 adult Finnish twin brothers. Alcohol Clin Exp Res 1987;11:349–56. Carmelli D, Swan GE, Robinette D, Fabsitz RR. Heritability of substance use in the NAS-NRC Twin Registry. Acta Genet Med Gemellol 1990;39:91–98.
36
37
38
39
40
41
42
43
Heath AC, Meyer J, Jardine R, Martin NG. The inheritance of alcohol consumption patterns in a general population twin sample: II. Determinants of consumption frequency and quantity consumed. J Stud Alcohol 1991;52:425–33. Prescott CA, Hewitt JK, Heath AC, Truett KR, Neale MC, Eavles LJ. Environmental and genetic influences on alcohol use in a volunteer sample of older twins. J Stud Alcohol 1994;55:18–33. Heath AC, Madden PAF. Genetic influences on smoking behavior. In: Turner RJ, Cardon LR, Hewitt JD (eds). Behavior Genetic Approaches in Behavioral Medicine. New York: Plenum Press, 1995. Swan GE, Carmelli D. Behavior genetic investigations of cigarette smoking and related issues. In Noble EP, Blum K. (eds). Handbook of Psychiatric Genetics, CRC Press, 1997, pp. 379–98. Sullivan PF, Kendler KS. The genetic epidemiology of smoking. Nicotine Tob Res 1999;1 Suppl 2:S51–7. Hopfer CJ, Stallings MC, Hewitt JK. Common genetic and environmental vulnerability for alcohol and tobacco use in a volunteer sample of older female twins. J Stud Alcohol 2001;62:717–23. Li MD, Cheng R, Ma JZ, Swan GE. A meta-analysis of estimated genetic and environmental effects on smoking behavior in male and female adult twins. Addiction 2003;98:23–31. Whitfield JB, Zhu G, Madden PA, Neale MC, Heath AC, Martin NG. The genetics of alcohol intake and of alcohol dependence. Alcohol Clin Exp Res 2004;28:1153–60. Kendler KS, Heath AC, Neale MC, Kessler RC, Eaves LJ. A population-based twin study of alcoholism in women. JAMA 1992;268:1877–82. Heath AC, Bucholz KK, Madden PA et al. Genetic and environmental contributions to alcohol dependence risk in a national twin sample: consistency of findings in women and men. Psychol Med 1997;27:1381–96. van den Bree MB, Johnson EO, Neale MC, Svikis DS, McGue M, Pickens RW. Genetic analysis of diagnostic systems of alcoholism in males. Biol Psychiatry 1998;43:139–45. Prescott CA, Aggen SH, Kendler KS. Sex differences in the sources of genetic liability to alcohol abuse and dependence in a populationbased sample of U.S. twins. Alcohol Clin Exp Res 1999;23:1136–44. True WR, Xian H, Scherrer JF et al. Common genetic vulnerability for nicotine and alcohol dependence in men. Arch Gen Psychiatry 1999;56:655–61. Lessov CN, Martin NG, Statham DJ et al. Defining nicotine dependence for genetic research: evidence from Australian twins. Psychol Med 2004;34:865–79. Niu T, Chen C, Ni J et al. Nicotine dependence and its familial aggregation in Chinese. Int J Epidemiol 2000;29:248–52. Heatherton TF, Kozlowski LT, Frecker RC, Fagerstro¨m KO. The Fagerstro¨m Test for Nicotine Dependence: a revision of the Fagerstro¨m Tolerance Questionnaire. Br J Addict 1991;86:1119–27. Yang H, Li X, Cao W et al. Chinese National Twin Registry as a resource for genetic epidemiologic studies of common and complex diseases in China. Twin Res 2002;5:347–51. Eaves LJ, Eysenck HJ, Martin NG. Genes, Culture and Personality: An Empirical Approach. London: Academic Press, 1989. Heath AC, Nyholt DR, Neuman R et al. Zygosity diagnosis in the absence of genotypic data: an approach using latent class analysis. Twin Res 2003;6:22–26. Neale MC, Cardon LR. Methodology for Genetic Studies of Twins and Families. Dordrecht/Boston/London: Kluwer Academic Publishers, 1992. Neale MC, Boker SM, Xie G, Maes HH. Mx: Statistical Modeling. 6th edn. Department of Psychiatry, Virginia Commonwealth University, Box 980126, Richmond, VA, 2002. Akaike H. Factor analysis and AIC. Psychometrika 1987;52:317–32.
HERITABILITY OF CIGARETTE SMOKING AND ALCOHOL USE IN CHINESE MALE TWINS 44
45
46
47 48 49
50
51 52
53
54
55
Centers for Disease Control and Prevention. Cigarette smoking among adults–United States, 2003. MMWR Morb Mortal Wkly Rep 2005;54:509–13. SAMHSA, Office of Applied Studies. Results from the 2003 National Survey on Drug Use and Health: National Findings (DHHS Publication No. SMA 04–3964, NSDUH Series H-25). Rockville, MD: Substance Abuse and Mental Health Services Administration, 2004.
56
57 58
Zhang X, Shu XO, Yang G et al. Association of passive smoking by husbands with prevalence of stroke among Chinese women nonsmokers. Am J Epidemiol 2005;161:213–18.
59
Can JP, Zheng J. The influence of geological factor on birth rates of twins in China. Chinese Public Health 2001;18:283–85.
60
Yang G, Fan L, Tan J et al. Smoking in China: findings of the 1996 National Prevalence Survey. JAMA 1999;282:1247–53. Gu D, Wu X, Reynolds K et al. Cigarette smoking and exposure to environmental tobacco smoke in China: The International Collaborative Study of Cardiovascular Disease in Asia. Am J Public Health 2004;94:1972–76. Heath AC, Martin NG. Genetic models for the natural history of smoking: evidence for a genetic influence on smoking persistence. Addict Behav 1993;18:19–34.
61
62
63
True WR, Heath AC, Scherrer JF et al. Genetic and environmental contributions to smoking. Addiction 1997;92:1277–87. Madden PA, Heath AC, Pedersen NL, Kaprio J, Koskenvuo MJ, Martin NG. The genetics of smoking persistence in men and women: a multicultural study. Behav Genet 1999;29:423–31. Madden PA, Pedersen NL, Kaprio J, Koskenvuo MJ, Martin NG. The epidemiology and genetics of smoking initiation and persistence: crosscultural comparisons of twin study results. Twin Res 2004;7: 82–97. Kaprio J, Koskenvuo M, Langinvainio H. Finnish twins reared apart. IV: smoking and drinking habits. A preliminary analysis of the effect of heredity and environment. Acta Genet Med Gemellol 1984;33: 425–33. Swan GE, Carmelli D, Cardon LR. The consumption of tobacco, alcohol, and coffee in Caucasian male twins: a multivariate genetic analysis. J Subst Abuse 1996;8:19–31.
64
65
66
67
1285
Hettema JM, Corey LA, Kendler KS. A multivariate genetic analysis of the use of tobacco, alcohol, and caffeine in a population based sample of male and female twins. Drug Alcohol Depend 1999;57:69–78. Swan GE, Carmelli D, Cardon LR. Heavy consumption of cigarettes, alcohol and coffee in male twins. J Stud Alcohol 1997;58:182–90. Heath AC, Cates R, Martin NG et al. Genetic contribution to risk of smoking initiation: comparisons across birth cohorts and across cultures. J Subst Abuse 1993;5:221–46. Pam A, Kemker SS, Ross CA Golden R. The ‘‘equal environments assumption’’ in MZ-DZ twin comparisons: an untenable premise of psychiatric genetics? Acta Genet Med Gemellol (Roma) 1996;45:349–60. Guo SW. Does higher concordance in monozygotic twins than in dizygotic twins suggest a genetic component? Hum Hered 2001;51: 121–32. Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ. A test of the equal-environment assumption in twin studies of psychiatric illness. Behav Genet 1993;23:21–27. LaBuda MC, Svikis DS, Pickens RW. Twin closeness and co-twin risk for substance use disorders, assessing the impact of the equal environment assumption. Psychiatry Res 1997;70:155–64. Kendler KS, Gardner CO. Twin studies of adult psychiatric and substance dependence disorders: are they biased by differences in the environmental experiences of monozygotic and dizygotic twins in childhood and adolescence? Psychol Med 1998;28:625–33. Xian H, Scherrer JF, Eisen SA et al. Self-reported zygosity and the equal-environments assumption for psychiatric disorders in the Vietnam era twin registry. Behav Genet 2000;30:303–10. Cronk NJ, Slutske WS, Madden PAF, Bucholz KK, Reich W, Heath AC. Emotional and behavioral problems among female twins: an evaluation of the equal environments assumption. J Am Acad Child Adolescent Psychiatry 2002;41:829–37. Pergadia ML, Heath AC, Agrawal A, Bucholz KK, Martin NG, Madden PA. The implications of simultaneous smoking initiation for inferences about the genetics of smoking behavior from twin data. Behav Genet 2006:1–10. Mayfield D, McLeod G, Hall P. The CAGE questionnaire: validation of a new alcoholism screening instrument. Am J Psychiatry 1974;131:1121–23.
