Tai Hing Lam

The University of Hong Kong, Hong Kong, Hong Kong

Are you Tai Hing Lam?

Claim your profile

Publications (299)1359.82 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: To examine the relationship between smoking, quitting, and mortality in older Chinese men. A cohort analytic study was carried out in Xi'an, China. A total of 1268 retired male military cadres aged 60 or older were examined in 1987 and followed for 12 years. At baseline, 388 men were never-smokers, 461 were former smokers, and 419 were current smokers. Through May 1999, a total of 299 had died. The relative risks [95% confidence intervals (CI)] for ever-smoking, after adjusting for age, blood pressure, body mass index, total cholesterol, triglycerides, alcohol drinking, exercise and existing diseases, for deaths resulting from all causes, chronic obstructive pulmonary disease (COPD), lung cancer, and coronary heart disease (CHD) were, respectively, 1.34 (1.02-1.76), 3.23 (0.95-10.91), 2.31 (0.95-5.61), and 1.60 (0.81-3.19). The risks increased significantly with increasing amount and duration of smoking. Compared with current smokers, former smokers had lower risks of total mortality (excess risk reduction of 56%) and from CHD death, but had higher risks for COPD death. Smoking is a major cause of death in older Chinese and quitting can save lives. Early recognition of the significance of COPD symptoms followed by prompt quitting should be emphasized in the control of the growing tobacco epidemic.
    Annals of Epidemiology 08/2002; 12(5):316-20. DOI:10.1016/S1047-2797(01)00258-7 · 2.15 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To study the main causes of death in workers from different industries in Guangzhou. The industrial systems in Guangzhou were classified into six categories according to the Chinese Public Health Inspection Statistical Report. Baseline data on 79,547 workers, age > or = 35, were retrieved from individual health records under the Guangzhou Occupational Health Surveillance Record System established in 1989-1992. In this prospective cohort study, the workers' vital status and causes of death were followed up until 31 December 1998. Using SPSS 10.0, crude death rates were calculated and relative risks(RR) (95% CI) were estimated using Cox proportional-hazard models. (1) Among 79,547 workers were 49,355 men and 30,192 women, with mean age of (43.8 +/- 6.5) years, and 64% were aged 35-44. The mean age difference of workers in different industries was +/- 1 year. At baseline, 41% had been exposed to occupational hazards. Exposure was most prevalent in petroleum chemical industry(53.9%), followed by metallurgical industry, and the lowest exposure was in mechanical industry (30.2%). (2) The mean follow-up period was (8.0 +/- 1.3) years with 633,510 person-years. 1,577 workers had died with a total death rate of 248.9 per 100,000 person-years. The death rate in rubber industry was the highest, followed by metallurgical and petroleum chemical industries. Malignant neoplasms, vascular and respiratory diseases were most common, constituting 80.3% of all causes of death. (3) For all causes of death and the three main categories, the top three crude death rates were observed in rubber and petroleum chemical industries, and in the metallurgical industry except for vascular causes. The rubber industry had the highest crude death rate for all and vascular causes, and the metallurgical industry for malignant neoplasms and respiratory causes. (4) With the light industry as a reference (RR = 1.0), the metallurgical industry had the highest adjusted relative risks (RR) (95% CI) of 1.62(1.20-2.18) and 7.42(3.38-16.31) for malignant neoplasms and respiratory causes, respectively. The rubber industry had the highest RR of 1.69(1.40-2.04) for all causes. A significant RR of 2.03(1.43-2.88) for vascular causes was observed only in the rubber industry. The main causes of death varied from industry. Malignant neoplasms and respiratory diseases were the leading causes in metallurgical and petroleum chemical industries. In rubber industry, vascular and respiratory causes were most common. Preventive measures for disease control should be tailor-made for each industry.
    Zhonghua lao dong wei sheng zhi ye bing za zhi = Zhonghua laodong weisheng zhiyebing zazhi = Chinese journal of industrial hygiene and occupational diseases 08/2002; 20(4):289-92.
  • Gabriel M Leung · Tai Hing Lam · Lai Ming Ho
    [Show abstract] [Hide abstract]
    ABSTRACT: To examine the effects of cesarean and forceps or vacuum delivery and parental smoking habits on the initiation and duration of breast-feeding. We conducted a prospective, population-based birth cohort study in 1997. Data were collected on breast-feeding history, household smoking habits, method of delivery, and other demographic, obstetric, behavioral, and potential confounding variables via a standardized self-administered questionnaire. Multivariable logistic regression was used to examine the association between method of birth (cesarean versus forceps or vacuum delivery versus normal vaginal birth) and either not initiating breast-feeding or doing so for less than 1 month. Among women who breast-fed for 1 month or more, multivariable survival analysis was employed to study the relationship between method of delivery and breast-feeding duration. We repeated these analyses to examine the link between parental smoking habits and breast-feeding initiation and duration. A total of 7825 mother-infant pairs were followed up for 9 months. Cesarean delivery was a risk factor for not initiating breast-feeding, for breast-feeding less than 1 month, and remained a significant hazard against breast-feeding duration. Assisted delivery with forceps or vacuum, although not associated with breast-feeding initiation, was a significant risk against breast-feeding duration. Conversely, current parental smoking habits only affected breast-feeding initiation but were unrelated to breast-feeding duration. This study indicates a possible effect of forceps or vacuum delivery on breast-feeding and of cesarean on long-term breast-feeding duration. The findings provide additional evidence in support of the avoidance of unnecessary obstetric interventions.
    Obstetrics and Gynecology 06/2002; 99(5 Pt 1):785-94. DOI:10.1016/S0029-7844(02)01940-3 · 4.37 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The hypothesis that protection of infants from exposure to infectious agents with delayed first exposure to one or more specific agents together contribute to the aetiology of childhood leukaemia, especially common acute lymphoblastic leukaemia (cALL), has substantial indirect support from descriptive epidemiology and case–control studies in developed Western countries. A case–control study of childhood leukaemia diagnosed at ages 2–14 years has now been conducted in Hong Kong. Cases (n = 98) formed a consecutive series of Chinese children diagnosed with acute leukaemia; controls (n = 228) were identified following a survey using random digit dialling and required to attend for medical examination by a paediatrician. Interviews with mothers were conducted in hospital by one trained interviewer using a structured questionnaire. Odds ratios (OR) and 95% confidence intervals (CI) are reported for exposure variables capable of serving as proxies for exposure to infection in two critical time periods: first year of life, year before reference date (diagnosis for cases, corresponding date for controls). Analyses used logistic regression with adjustment for appropriate confounders. Change of area of residence reduced risk if during the first time period (OR = 0.47 [95% CI 0.23, 0.98]) and increased risk if during the second (OR = 3.92, [95% CI 1.47, 10.46]). Reported roseola and/or fever and rash in the first year of life reduced risk (OR = 0.33 [95% CI 0.16, 0.68]) whereas tonsillitis in the period 3–12 months before reference date increased risk (OR = 2.56 [95% CI 1.22, 5.38]). Some other proxies for exposure to infection at the critical times were associated with predicted patterns of risk but day-care attendance failed to show predicted associations. These results provide support for the delayed exposure hypothesis in an affluent geographical setting in which population exposure to infectious agents is quite distinct from the settings of previous case–control studies.
    Paediatric and Perinatal Epidemiology 05/2002; 16(2):154-65. DOI:10.1046/j.1365-3016.2002.00406.x · 2.81 Impact Factor
  • Tai Hing Lam · Chang Qi Zhu · Chao Qiang Jiang
    [Show abstract] [Hide abstract]
    ABSTRACT: To study the effect of smoking, passive smoking, alcohol drinking, and occupational exposure to low level of benzene on DNA strand breaks in elevator manufacturing workers in Guangzhou, China. Three hundred and fifty-nine workers (252 men and 107 women) of a modern elevator manufacturing factory, 205 were from production departments and 154 from managerial department. Information on the workers' health conditions, smoking, passive smoking, alcohol consumption and occupational exposure history was collected by personal interview. Lymphocyte DNA damage was measured by the Comet assay. None of the women smoked and 20.6% of the men were daily smokers. In non-smokers, the prevalence of passive smoking at work was 25% for men and 11.2% for women, and at home, 37.8 and 48.6%, respectively. Smoking significantly increased tail moment (P<0.001). Daily smokers had the largest tail moment (geometric mean, 95% CI) (0.93 microm (0.81-0.94)), followed by occasional smokers (0.76 microm (0.59-0.95)), ex-smokers (0.70 microm (0.58-0.85)), and never smokers (0.56 microm (0.53-0.60)). Tail moment increased significantly with daily tobacco consumption (cigarettes per day) (r=0.26, P<0.001) after adjusting for age, gender, occupational exposure, passive smoking, and drinking. Analysis of covariance (ANCOVA) showed that smoking (P<0.001), passive smoking at home (P=0.026), occupational exposure (P<0.001), male gender (P<0.001), and age (P=0.001) had independent effects on tail moment, whereas passive smoking at work and alcohol drinking had no significant effect. Smoking, passive smoking at home, male gender, age and occupational exposure independently increased lymphocyte DNA strand breaks. The presence of excess DNA damage under low level of occupational exposure to benzene or other solvents suggest that the current allowance concentrations may not be safe to prevent genotoxicity.
    Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 03/2002; 515(1-2):147-57. DOI:10.1016/S1383-5718(02)00010-4 · 4.44 Impact Factor
  • Tze Yuen Lee · Tai Hing Lam
    Contact Dermatitis 11/2001; 45(4):237. DOI:10.1034/j.1600-0536.2001.450411.x · 3.62 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The objective was to describe the extent of passive smoking exposure and preventive actions against passive smoking in non-smoking pregnant women and to analyze factors associated with preventive actions. This was a cross-sectional survey on 1449 never-smoking pregnant women who made their first prenatal visit to the Women and Children's Hospital of Guangzhou, China during 1996-97. 60.2% (95% Confidence Interval 57.7%-62.7%) of the never-smoking pregnant women had a husband who currently smoked. Women with smoking husbands (n=872) were more exposed to ETS than those with non-smoking husbands (n=577) at home (71% vs. 33%), in public places (77% vs. 66%) and at work (60% vs. 50% of working women), and they took less actions against passive smoking in public places. Women with better education and knowledge on smoking and passive smoking had stronger preventive actions. Non-smoking pregnant women in China are often exposed to passive smoking and their preventive actions are weak. Passive smoking is an important obstetric and public health problem in developing countries and deserves urgent international attention.
    Acta Obstetricia Et Gynecologica Scandinavica 12/2000; 79(11):947-52. DOI:10.1080/00016340009169240 · 1.99 Impact Factor
  • AY Loke · TH Lam · SC Pan · YL Su · XJ Gao · YY Song
    [Show abstract] [Hide abstract]
    ABSTRACT: Background The objective was to describe the extent of passive smoking exposure and preventive actions against passive smoking in non-smoking pregnant women and to analyze factors associated with preventive actions. Methods. This was a cross-sectional survey on 1449 never-smoking pregnant women who made their first prenatal visit to the Women and Children's Hospital of Guangzhou, China during 1996-97. Results. 60.2% (95% Confidence Interval 57.7%-62.7%) of the never-smoking pregnant women had a husband who currently smoked. Women with smoking husbands (n = 872) were more exposed to ETS than those with non-smoking husbands (n = 577) at home (71% vs. 33%), in public places (77% vs. 66%) and at work (60% vs. 50% of working women), and they took less actions against passive smoking in public places. Women with better education and knowledge on smoking and passive smoking had stronger preventive actions. Conclusions. Non-smoking pregnant women in China are often exposed to passive smoking and their preventive actions are weak. Passive smoking is an important obstetric and public health problem in developing countries and deserves urgent international attention.
    Acta Obstetricia Et Gynecologica Scandinavica 11/2000; 79(11):947-952. DOI:10.1034/j.1600-0412.2000.079011947.x · 1.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To study the effect of occupational exposure to rubber processing, smoking, and alcohol drinking on lymphocyte DNA damage. Of 371 employees (197 men and 174 women) from a rubber factory in Guangzhou, 281 were rubber processing workers from five production sections and 90 were managerial workers. Information on occupational exposure, smoking, and drinking was collected by interviews. Blood samples were taken in the morning by venipuncture. DNA damages were measured by the Comet assay. Possible DNA-protein crosslinks were broken down by proteinase K. Tail moment, measured by Komet 4.0 image analysis software, was the measure of DNA damage. The rubber processing workers had larger tail moment than the managerial workers (Geometric mean, 95%CI) [1. 77microm (1.64-1.90) versus 1.52microm (1.36-1.71), P=0.04]. Both smoking [1.93microm (1.74-2.13) versus 1.59microm (1.47-1.71), P=0. 003] and alcohol drinking [2.21microm (1.87-2.62) versus 1.63microm (1.53-1.74), P<0.001] increased tail moment. Tail moment differed significantly among job categories (F=3.21, P=0.008), the largest was observed in mixers. In the non-smoking and non-drinking workers, rubber processing workers had larger tail moment than managerial workers after adjusting for age (P=0.033). General linear model analysis showed that after adjusting for each other, occupational exposure (P=0.027), smoking (P=0.012), and alcohol drinking (P=0. 013) was associated with larger tail moment, whereas age and gender had no effect. Occupational exposure to rubber processing, smoking, and alcohol drinking can cause DNA damage.
    Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 11/2000; 470(2):201-9. DOI:10.1016/S1383-5718(00)00106-6 · 4.44 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Few epidemiological studies have examined the relationship between chronic respiratory symptoms and exposure to environmental tobacco smoke (ETS) at work in adults, and none have shown clear dose-response relationships. To examine the respiratory effects of ETS exposure at home and at work among never-smoking adults. Cross-sectional, self-administered questionnaire survey conducted in December 1995 and January 1996 among 4468 male and 728 female police officers in Hong Kong who were never-smokers. Respiratory symptoms and physician consultation in the previous 14 days for such symptoms by presence and amount of ETS exposure at work. Eighty percent of both men and women reported ETS exposure at work. Significant odds ratios (ORs) for respiratory symptoms were found among men with ETS exposure at work (for any respiratory symptoms, difference in absolute rate, 20.4%; OR, 2.33; 95% confidence interval [CI], 1.97-2.75; attributable risk, 57%) and physician consultation (difference in absolute rate, 4.5%; OR, 1.30; 95% CI, 1. 05-1.61; attributable risk, 23%). Trends were similar among women for any respiratory symptoms (difference in absolute rate, 15.4%; OR, 1.63; 95% CI, 1.04-2.56; attributable risk, 39%) and for physician consultation (difference in absolute rates, 2.8%; OR, 1.45; 95% CI, 0.87-2.41; attributable risk, 31%). Positive dose-response relationships with number of coworkers smoking nearby and amount of ETS exposure in the work place were found. This study provides further evidence of the serious health hazards associated with ETS exposure at work. The findings support a ban on smoking in the workplace to protect all workers in both developed and developing countries. JAMA. 2000;284:756-763
    JAMA The Journal of the American Medical Association 09/2000; 284(6):756-63. DOI:10.1001/jama.284.6.756 · 30.39 Impact Factor
  • Gabriel M Leung · Tai Hing Lam · Anthony J Hedley
    The Lancet 03/2000; 355(9205):750-1; author reply 752. DOI:10.1016/S0140-6736(05)72158-9 · 45.22 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Few studies have examined fibrinogen in Chinese populations in which the incidence of coronary heart disease (CHD) is lower than that in the West. This study aimed to examine the relationship between fibrinogen and other CHD risk factors in Hong Kong Chinese. Fibrinogen was measured by the Clauss method in 1359 men and 1405 women aged 25-74 years, randomly selected from the Hong Kong population. Mean fibrinogen level increased with age, from 2.22 g/l in those aged 25-34 years to 2.76 g/l in 65-74 years in men, and from 2.42 to 2.94 g/l respectively in women. The most important factors associated with fibrinogen were age, obesity and blood lipid levels in both genders. In men, smoking was associated with higher fibrinogen levels and cessation of smoking with lower levels. Prospective studies are needed to examine the role of fibrinogen in CHD in Chinese and other Asian populations.
    Atherosclerosis 05/1999; 143(2):405-13. DOI:10.1016/S0021-9150(98)00294-9 · 3.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Study objective:To describe the prevalence of smoking in junior secondary school students in Hong Kong and to analyse the relationship between a range of risk factors and ever-smoking experience, including tobacco advertisements.Methods:Cross-sectional survey using an anonymous standardised self-administered questionnaire.Setting:61 randomly selected secondary schools in Hong Kong.Participants:6,304 form 1 to 3 Chinese students from 172 classes (51% girls and 49% boys; 90% were aged 12 to 15 years).Results:The smoking status (95% confidence interval) was: (1) never smoker, 71.1% (70.0%–72.2%), (2) tried only, 15.5% (14.6%–16.4%), (3) used to smoke but not now, 4.2% (3.7%–4.7%), (4) smoked 6 per week, 4.5% (4.0%–5.0%). In a backward stepwise logistic regression model, ever-smoking (including categories 2 to 6 above) was independently associated with thirteen factors, including gender (boys), increasing age, place of birth (outside Hong Kong, mainly China), poor knowledge of the hazards of smoking, positive attitudes to smoking, smoking in family members, participation in tobacco promotional activities, and perception of cigarette advertisements as attractive. Among the strongest associations observed was the youth’s perception of cigarette advertisements as attractive, with the adjusted odds ratio (95% confidence interval) of 2.68 (2.33–3.07).Conclusions:Smoking among young people is an important public health problem. Although the causes are multifactorial, in Hong Kong environmental tobacco advertising is an important risk factor that can be removed by banning all forms of tobacco promotion to young people.
    American Journal of Preventive Medicine 04/1998; 14(3):217-223. DOI:10.1016/S0749-3797(97)00071-8 · 4.28 Impact Factor
  • Tai Hing Lam · Yao He
    [Show abstract] [Hide abstract]
    ABSTRACT: 1. Whether passive smoking or environmental tobacco smoke (ETS) can cause coronary heart disease (CHD) is controversial. We have undertaken a comprehensive review of the epidemiological studies regarding the relationship between ETS and CHD. 2. We searched for all original papers and reviews and calculated pooled odds ratio using the Mantel-Haenszel method. 3. Ten prospective studies, nine case-control studies and one cross-sectional study were found. Almost all studies showed positive associations between ETS and CHD and a few were statistically significant, with dose-response relationships being evident. Six review papers calculated pooled estimates of the relative risks, which ranged from 1.23 to 1.51. Other studies also showed that ETS could have other cardiovascular effects. 4. The association observed is likely to be real. The criteria for causation of time sequence, consistency, coherence and biological plausibility are satisfied. We conclude that ETS may cause CHD, particularly in women who have never smoked.
    Clinical and Experimental Pharmacology and Physiology 01/1998; 24(12):993-6. DOI:10.1111/j.1440-1681.1997.tb02738.x · 2.41 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To examine plasma total antioxidant capacity (TAOC) in a Chinese population with a lower incidence of cardiovascular disease compared with Caucasian populations, in relation to dietary intake, age, sex, and the presence of cardiovascular diseases. As part of a randomized territory-wide survey stratified by sex and 10-year age groups; 728 subjects (367 men, 361 women) were recruited. Dietary intake assessment was by a food frequency questionnaire; plasma TAOC was estimated by the ABTS method. The TAOC values were normally distributed, the mean +/- SD being 1.78 +/- 0.18 mmol/L. The mean value was higher in men compared with women, inspite of a lower dietary intake of vitamins A and C per 1000 kcal in the former. Subjects who consumed water spinach twice or more a week had higher mean levels. No difference in mean levels was observed between those with and without hypertension or cardiovascular disease. Measurement of plasma TAOC as a risk factor in epidemiological studies of cardiovascular diseases may have limited use, since TAOC include substances associated with a protective effect as well as increased risk.
    Clinical Biochemistry 11/1997; 30(7):553-7. DOI:10.1016/S0009-9120(97)00112-4 · 2.28 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The relationship between the number of stenotic coronary arteries and passive smoking exposure from the husband was examined in a retrospective study of 78 cases with stenotic coronary arteries and 83 controls with normal coronary arteries. All the 161 subjects were examined by coronary arteriography and were Chinese women who had never smoked cigarettes (i.e. lifelong non-smokers). The Pearson's and Spearman's correlation coefficients between the number of stenotic coronary arteries and passive smoking exposure indices were 0.27-0.29 and 0.25-0.28 (all with P < 0.01), respectively. The linear prediction equations from multiple regression analysis showed that passive smoking exposure, after adjustment for other major risk factors, was a significant predictive variable for the number of stenotic coronary arteries. The multiple response logistic regression model was used to estimate the odds ratio of stenotic coronary arteries for passive smoking exposure categories. The crude and adjusted odds ratios showed that the number of stenotic arteries increased with the amount of exposure to passive smoking from the husband. These results provide new evidence to support that passive smoking is causally associated with coronary heart disease.
    Atherosclerosis 12/1996; 127(2):229-38. DOI:10.1016/S0021-9150(96)05960-6 · 3.99 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: *Purpose:* To assess the frequency and accuracy of weight comments received from different sources among Chinese adolescents and the relation of weight comments to weight misperception. *Methods:* In the Hong Kong Student Obesity Surveillance (HKSOS) project 2006-07, 22612 form 1 to 7 students (41.5% boys) completed a questionnaire on obesity. Students were asked if they had received serious "too fat" or "too thin" comments over the past 30 days from family members, peers and professionals, and the accuracy of the comments was assessed against the actual weight status. Weight misperception was defined as discordance between the actual and perceived weight status. Logistic regression was used to calculate adjusted odd ratios (ORs) for weight misperception by weight comments received. *Results:* One in three students received weight comments, most commonly from mother for both "too fat" and "too thin" comments. Health professional was the most accurate source of weight comments, yet less than half the comments were correct. Receiving incorrect weight comments was significantly associated with weight misperception in all sex-weight status sub-groups, particularly among obese girls. In contrast, student who received correct weight comments were less likely to have weight misperception. *Conclusions:* Weight comments were commonly received by Chinese adolescents in Hong Kong, yet most of the comments were inaccurate, and such incorrect comments were associated with weight misperception. Family members, peers and professionals should realize the potential adverse effects of their weight comments and adolescents should be taught how to correctly assess their weight status to reduce misperceptions.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Few prospective data from the Asia Pacific region are available relating body mass index to the risk of diabetes. Our objective was to provide reliable age, sex and region specific estimates of the associations between body mass index and diabetes. Twenty-seven cohort studies from Asia, New Zealand and Australia, including 154,989 participants, contributed 1,244,793 person-years of follow-up. Outcome data included a combination of incidence of diabetes (based on blood glucose measurements) and fatal diabetes events. Hazard ratios were calculated from Cox models, stratified by sex and cohort, and adjusted for age at risk and smoking. During follow-up (mean = 8 years), 75 fatal diabetes events and 242 new cases of diabetes were documented. There were continuous positive associations between baseline body mass index and risk of diabetes with each 2 kg/m(2) lower body mass index associated with a 27% (23-30%) lower risk of diabetes. The associations were stronger in younger age groups, and regional comparisons demonstrated slightly stronger associations in Asian than in Australasian cohorts (P = 0.04). This overview provides evidence of a strong continuous association between body mass index and diabetes in the Asia Pacific region. The results indicate considerable potential for reduction in incidence of diabetes with population-wide lowering of body mass index in this region.

Publication Stats

4k Citations
1,359.82 Total Impact Points

Institutions

  • 1996–2015
    • The University of Hong Kong
      • • School of Public Health
      • • Department of Community Medicine
      Hong Kong, Hong Kong
  • 2014
    • University of Queensland
      • School of Population Health
      Brisbane, Queensland, Australia
  • 2013
    • City University of Hong Kong
      Chiu-lung, Kowloon City, Hong Kong
  • 2006–2012
    • University of Birmingham
      • Department of Public Health, Epidemiology and Biostatistics
      Birmingham, ENG, United Kingdom
  • 2010
    • Queen Mary Hospital
      Hong Kong, Hong Kong
  • 2008
    • Hong Kong SAR Government
      Hong Kong, Hong Kong
  • 2007
    • The Chinese University of Hong Kong
      • Department of Medicine and Therapeutics
      Hong Kong, Hong Kong
    • Yonsei University
      Sŏul, Seoul, South Korea
    • The George Institute for Global Health
      • Renal and Metabolic Division
      Sydney, New South Wales, Australia
  • 2003
    • Isfahan University of Medical Sciences
      • Epidemiology and Biostatistics Department
      Isfahan, Ostān-e Eşfahān, Iran
  • 2000–2002
    • The University of Hong Kong - Shenzen Hospital
      Hong Kong, Hong Kong