Article Text

Download PDFPDF

Mortality and cancer incidence among physicians of traditional Chinese medicine: a 20-year national follow-up study
  1. Shu-Hui Liu1,
  2. Yu-Feng Liu1,
  3. Saou-Hsing Liou2,
  4. Yun-Lian Lin3,
  5. Yuen-Chen Hsiao4,
  6. Chu-Chieh Chen5,
  7. Chung-Yi Li5,
  8. Trong-Neng Wu1,2,6
  1. 1Institute of Environmental and Occupational Health Sciences, National Yang Ming University, Taipei, Taiwan
  2. 2Division of Environmental Health and Occupational Medicine, National Health Research Institutes, Zhunan, Miaoli County, Taiwan
  3. 3National Research Institute of Chinese Medicine, Ministry of Education, Taipei, Taiwan
  4. 4Department of Industrial Education, National Taiwan Normal University, Taipei, Taiwan
  5. 5Department of Health Care Management, National Taipei College of Nursing, Taipei, Taiwan
  6. 6Department of Public Health and Institute of Environmental Health, China Medical University and Hospital, Taichung, Taiwan
  1. Correspondence to Dr Trong-Neng Wu, Institute of Environmental Health, College of Public Health, China Medical University and Hospital, 91 Hseuh-Shih Road, Taichung 404, Taiwan; tnwu{at}


Objective To study the risks of mortality and cancer incidence in physicians of traditional Chinese medicine (TCM) who had frequent exposure to herbal medicine.

Methods A population-based cohort design was conducted in which a total of 7675 certified physicians of TCM who ever practised between 1985 and 2005 were compared with the age-, sex- and calendar year-specific mortalities and cancer incidence rates of the general population of Taiwan. The age-, sex- and calendar year-standardised mortality ratio (SMR) and standardised cancer incidence ratio (SIR) were calculated to estimate the relative risks of all causes and site-specific mortality and cancer incidence.

Results Over an up to 20-year observational period, 796 (10.4%) physicians of TCM died, and 279 (3.6%) developed cancer. The study cohort showed a significantly reduced SMR for all-causes mortality (68, 95% CI 63 to 73), and for deaths from infectious (SMR=64), circulatory (SMR=68), respiratory (SMR=64) and digestive (SMR=56) disease. The study cohort also had a significantly reduced SIR (80, 95% CI 71 to 90) for all cancers, and for neoplasm of rectum, rectosigmoid junction, and anus (SIR=45), female breast (SIR=30) and cervix uteri (SIR=10). On the other hand, we noted that physicians of TCM suffered from a significantly increased SIR for neoplasm of liver and intrahepatic bile ducts (SIR=151, 95% CI 116 to 192) and of bladder cancer (SIR=259, 95% CI 167 to 382).

Conclusion Like other healthcare workers, we noted that physicians of TCM had significantly reduced risks of all-causes mortality and cancer incidence. Nonetheless, reasons truly responsible for significantly increased risks of liver and bladder neoplasm among physicians of TCM warrant further investigations.

  • Cancer
  • epidemiology
  • herbal medicine
  • liver
  • liver neoplasms
  • mortality studies
  • physicians
  • traditional Chinese medicine
  • urinary bladder neoplasms

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Increased utilisation of complementary alternative medicine (CAM) has been seen worldwide. It was estimated that 40% of the Chinese population and 65% of the rural Indian population seek treatment from alternative medicine regularly.1 Additionally, CAM has also been quite popular in developed nations. More than 65% of Germans, 70% Canadians, 48% Australians and 42% Americans ever utilised CAM in their lifetime.2 Traditional Chinese medicine (TCM) has been a main form of healthcare and a large proportion of people seek treatment from TCM in Taiwan, as well as in many parts of the world. Approximately 18.8% of the Taiwanese population older than 15 years ever received the care of TCM at least once in their life time.3

Given that a high prevalence of TCM utilisation and that little is known regarding the health status of TCM physicians whose health is central to the safety of patients, and to the quality of care received by patients, this study aims to elucidate, using a national cohort, the potential health problems of physicians of TCM in Taiwan.


Source of data and study cohort

Three national datasets including the Certified Medical Professional Registry (CMPR), Taiwan Cancer Registry Database (TCRD) and the Taiwan Death Registries (TDR) were used in this analysis.

On the first day of 1985, a total of 7675 certified physicians of TCM (6345 males and 1330 females) were registered in the CMPR, and their ages on the first day of 1985 were calculated. The mean age (±SD) for male and female subjects was 35.2 (15.8) and 27.9 (6.8) years, respectively. It is mandatory for all practice medical professional to be certified and registered, and the data of CMPR can be considered accurate and complete. The choice of 1985 was due to the fact that the electronic information on mortality was only available from 1985.

In Taiwan, the population-based TCRD was initiated by the Department of Health of the central government in 1979. The information reported from hospitals included diagnostic codes and treatment regime for all newly diagnosed patients with cancer. Duplicate checks and quality controls are run periodically to detect possible erroneous registrations and inconsistencies.

Data on mortality were obtained from the TDR. Due to mandatory registration of all deaths in Taiwan, the mortality registration is generally accurate and complete.4 We retrieved information on the date of death and underlying cause of death of each deceased individual.

Statistical analysis

The expected number of deaths for Chinese medical physicians was calculated from the person-year approach, using the age (5-year intervals) and sex-specific annual rates of death with reference to the general population. Standardised mortality ratios (SMRs) were calculated as risk estimates. The annual average size of the general population over the study period was 15 586 025. A similar approach was used to calculate the age-, sex- and calendar-year-standardised incidence ratios (SIRs) for all cancers as well as for site-specific cancers. The 95% CI for both SMR and SIR was estimated using the Byar approximation proposed by Breslow and Day.5 The analyses were performed using SAS (version 9.1; SAS Institute), and the level of significance was set at a p value of 0.05.


Over a 20-year study period (ie, 1985–2005), 796 study subjects (10.4%) died of all causes, and 279 study subjects (3.6%) developed cancer. The corresponding figures for male and female subjects were 775 (12.2%), 254 (4.0%) and 21 (1.6%), 15 (1.1%). The mean age of death and of cancer incidence was 73.8 (±14.4) and 65.86 (±17.63) years, respectively. The corresponding figures for male subjects were 74.1 (±14.2) and 67.2 (±17.1) years; for female subjects were 62.9 (±17.6) and 44.3 (±12.3) years.

Table 1 shows a significantly reduced SMR (68, 95% CI 63 to 73) of all-causes mortality in the study cohort. Significantly reduced SMRs were also noted for causes of infection (SMR=64), circulatory disease (SMR=68), respiratory disease (SMR=64), digestive disease (SMR=56), symptoms and ill-defined conditions (SMR=41) and injury (SMR=31). Physicians of TCM also experienced reduced risks of all cancers (SIR=80, 95% CI 71 to 90), and some other site-specific cancers including other and unspecified parts of mouth neoplasm (SIR=24), neoplasm of rectum, rectosigmoid junction and anus (SIR=45), female breast cancer (SIR=30) and cervical cancer (SIR=10). On the other hand, physicians of TCM suffered from significantly increased risks of both liver cancer (SIR=151, 95% CI 116 to 192) and bladder cancer (SIR=259, 95% CI 167 to 382) (table 2).

Table 1

Cause-specific number of death and age–sex–calendar-year-standardised mortality ratio of the study cohort

Table 2

Cause-specific number of incident cancer and age–sex–calendar-year-standardised incidence ratio of study cohort


This study demonstrated that physicians of TCM, like many other healthcare professionals in Taiwan3 and in other nations,6 had lower rates of all-causes mortality and overall cancer incidence than had the general population. It is likely that healthcare workers, especially physicians, usually had better health, probably due to a low rate of smoking and, to a lesser extent, other beneficial health-related behaviours; better access to healthcare also played a role.6 Although physicians of TCM had lower mortalities, all-causes cancer incidence rates and most site-specific cancer incidence rate, they were found to suffer from significantly elevated cancer risks of liver and bladder, which is worth noting.

Due to an increase in popularity of Chinese herbal medicine (CHM) regularly reimbursed under the National Health Insurance (NHI) system in Taiwan, Lee et al7 followed some 200 000 randomly selected insurers from 1997 to 2002 to investigate the association of CHM intake with non-viral, non-alcoholic hepatitis. After adjustment for conventional hepatotoxic drug utilisation, the odds ratio (3.4, 95% CI 1.1 to 9.8) of developing acute hepatitis was increased for taking certain prescribed formulae containing Radix Paeoniae and Radix Glycyrrhizae. The physicians of TCM might have a higher chance than the general population of being exposed to the CHM suspected with hepatotoxicity, which may be associated with hepatoma. Additionally, acupuncture is not uncommon in the practice by TCM physicians. It has been reported that needle-stick injury may increase the risk of blood transmitted disease.8 Given a high prevalence of chronic hepatitis B in Taiwanese population, physicians of TCM in Taiwan might have a greater occupational exposure to patients' blood and body fluids, leading to an increased risk of being infected by chronic hepatitis, which is also a known risk factor for liver cancer.

Previous studies reported an association of renal interstitial fibrosis and urothelial carcinoma with utilisation of CHM. A recent study conducted in Taiwan by Yang et al also reported a significantly increased risk of urological cancers in Chinese herbalists with an SMR of threefold, compared with the general population.9 Further site-specific analysis revealed a significantly increased SMR for kidney cancer (SMR=3.81), but the increased risk for bladder cancer was not statistically significant (SMR=2.26, 95% CI 0.47 to 6.59), which was mainly due to the limited number of kidney cancer cases. Our study noted a significantly increased risk of bladder cancer among physicians of TCM who might have been more likely than the general population of becoming CHM users. A link between CHM utilisation and neoplasm of bladder could be due to CHM's contamination by a number of heavy metals, including lead, mercury, cadmium and arsenic during the manufacture process.10 Further, comprehensive evidence has established the causal association between exposure to arsenic and bladder cancer.11

Nephropathy associated with aristolochic acid (AA), an ancient CHM used to reduce toxicity, has been documented by human and animal studies. In a nationally representative sample of some 200 000 people insured in Taiwan's NHI programme, Lai et al12 found increased risks of chronic kidney disease and end-stage renal disease in people treated with AA-related Chinese herbal products. Nonetheless, our study did not reveal any significantly elevated risk of neoplasm of kidney in physicians of TCM.

This study has several strengths. First, given the completeness of the national death and cancer incidence registry, and the successful linkage of the study cohort to these registries, this study has little room for selection bias caused by loss to follow-up of study subjects. Second, in Taiwan the underlying causes of death of each deceased individual must be certified by physicians, and all incident cancers registered must be pathologically confirmed, which provides reassurance that the validity of the information on death and cancer incidence is adequate, thus resulting in little likelihood of information bias. The major limitation of this study was a lack of information on study subjects' lifestyle, disease histories and work habits, which makes it difficult to provide specific interpretations of the study findings.

Findings of this study have important policy implications. First, ambulatory care visits to physicians of TCM are increasing in Taiwan as they are regularly reimbursed under Taiwan's NHI system, and the health of TCM physicians is highly related to the quality and safety of patient care, a mandatory periodical physical examination, especially for hepatological and urological disorders, for TCM physicians must seriously be considered. Second, because the popularity of CHM utilisation has increased, and empirical evidence has shown increased risks of certain cancer among herbalists, there is an urgent need for safety assessments of Chinese herbs.

What this paper adds

  • Physicians of traditional Chinese medicine are occupationally exposed to complementary alternative medicine for which the safety has been a concern.

  • Like many other healthcare workers, physicians of traditional Chinese medicine experienced significantly reduced risks of all causes and various causes-specific mortalities.

  • Although physicians of traditional Chinese medicine are at lower risks of overall cancer incidence, they are at significantly increased risks of neoplasm of liver and bladder.

  • Periodical health check-up and disease screening, especially for hepatological and urological disorders, in physicians of traditional Chinese medicine, are necessary.

  • Given the popularity of Chinese herbal medicine that has been shown to pose risks of certain cancer among herbalists, there is an urgent need for safety assessments of Chinese herbs which are also possible occupational hazards for the physicians of traditional Chinese medicine.



  • C-YL and T-NW contributed equally to this study.

  • Funding This study was supported by a grant from National Scientific Council, Taiwan (NSC 95-2314-B-030-002).

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the Taiwan Chinese Medical Association and the Department of Health, Taiwan.

  • Provenance and peer review Not commissioned; externally peer reviewed.