This report is the last of four volumes published in this third series of studies on Vietnam veterans. The other volumes are: • Cancer Incidence in Australian Vietnam Veterans Study 2005; • The Third Australian Vietnam Veteran Mortality Study 2005; and • Australian National Service Vietnam Veterans: Mortality and Cancer Incidence.
Study objectives
The objective of this study was to analyse the effects of Dapsone, an anti-malarial drug used during the Vietnam War on the mortality and cancer incidence of Vietnam Veterans. This investigation adds to the study completed in 1992 by the Australian Institute of Health and Welfare entitled Dapsone exposure, Vietnam service and cancer incidence.1
Study design
This study was a retrospective cohort study of male Army personnel who served in Vietnam between 1962 and July 1973. The study examined all deaths identified from the end of service to 31 December 2001 and all cancers diagnosed from 1982 to 31 December 2000. Relative mortality and cancer incidence rates were calculated for Army personnel who served in Vietnam and consumed Dapsone as part of the malarial prophylaxis treatment compared to Army personnel who did not consume Dapsone during their Vietnam service. Standardised ratios comparing the two treatment groups to the Australian population were also calculated. In addition, models were developed to assess the effect of the total cumulative Dapsone dose received during Vietnam service on mortality and cancer incidence.
Report structure
Chapter One of the report provides a brief background to Australia’s involvement in the Vietnam War and an overview of the drug Dapsone and its use in malaria prevention during that conflict. Chapter Two details the methods used for data matching and characteristics of the Army cohort. In brief, the study roll of Army personnel was matched to a number of databases to determine vital status, the number of deaths and their causes and the number and types of cancers diagnosed. Characteristics of the cohort include age and some service details by Dapsone exposure. Chapter Three describes the statistical methods used to analyse the mortality and cancer incidence data. Three methods were used. A direct comparison between the dapsone exposed veterans and the non-exposed veterans was used to calculate Relative Rate (RR). An indirect comparison was used in which the mortality and cancer incidence of the exposed veterans and the non-exposed veterans was compared to the male Australian population. This is presented as Standardised Mortality or Incidence Ratios (SMR/SIR). Lastly, regression analysis was used to assess whether a dose-response relationship between total cumulative dapsone dose and mortality or cancer incidence exists. The findings are presented and discussed in Chapter Four.
Features of the study
This was a large study that examined the long-term health effects amongst more than forty thousand veterans who consumed specific drug courses of therapy and who were followed for more than thirty years following service. This study compares mortality and cancer incidence among those who consumed a Dapsone plus Paludrine anti- malarial prophylactic treatment during their Vietnam service to those who consumed Paludrine only prophylaxis.
The consumption of the Dapsone was strictly monitored and recorded during the Vietnam era. Through these records, total cumulative Dapsone consumption was able to be reconstructed at an individual level. Dapsone consumption information was compiled for the previous study1 and no further refinement of this data was undertaken for the present study.
The temporal nature of Dapsone use needs to be considered. Dapsone was used from the last quarter of 1968 to 1972. Orders were given as to when Dapsone should be used and with few exceptions all who served during the time these orders were in effect (mainly during the wet seasons of these years) were given the Dapsone/Paludrine treatment. Over 80% of the Army personnel served one tour in Vietnam. The majority of those exposed to Dapsone served after 1968 and the majority of those who were not exposed to Dapsone served prior to 1969. Thus Dapsone exposure in general also correlates with the operational exposures occurring during the two time periods.
The study does not have information for many of the other exposures that would have occurred during service in Vietnam. Although this cohort was exposed to a range of stresses and toxic substances, little is known about the amount of exposure that an individual may have experienced. In addition, any adverse or beneficial health exposures an individual may have experienced following Vietnam service are also not known or quantified.
Statistical power must be considered when interpreting the results. Statistical power is the probability that a study will detect a statistically significant difference between two study groups if the groups truly differ. The size of the group being studied, the magnitude of the effect observed and the rate of occurrence of a health outcome influence statistical power. If a particular health outcome is rare then even a large study may not have sufficient power to detect a true difference, especially if this difference is small.
The role of chance and the concept of multiple comparisons must also be considered when interpreting the results. By convention statistical significance is at the 95% level, which means there is up to a one in twenty probability the result could be due to chance. Over 100 specific cancer diagnoses or causes of death are reported in this study. Thus by definition, apparently statistically significant associations could arise for up to 5 cancers or causes of death by chance alone.
Another feature of this study is that the cancer incidence could only be measured from 1982 (the first year of national cancer registration) to 2000 (the last year that all States and Territories had submitted their registry data to the National Cancer Clearing House at the time when the data matching was completed). Information on mortality from cancer was obtained from the end of Vietnam service to 2001. Any cancers diagnosed prior to 1982 and not resulting in death were not able to be captured in this study.
The healthy worker effect should also be considered. This is a phenomenon observed in occupational health studies in which those who are employed exhibit a lower mortality rate than the general population. This phenomenon is often referred to as the healthy soldier effect for occupational studies of military cohorts. This distinction denotes the fact that military populations are healthier than other employed populations, which in turn are healthier than the general population consisting of those employed and unemployed.
Findings
Indirect comparison
The indirect comparison showed that amongst Dapsone exposed Army veterans overall mortality was 6% lower than in the Australian male population, SMR = 0.94 (95% CI 0.90, 0.98). Two specific causes of death were more common in this group of veterans, alcoholic liver disease and suicide by gas, although suicide rates in total were not different from the Australian population. Mortality from infectious diseases, diabetes, nervous and circulatory system diseases was significantly less common.
Mortality was also lower amongst the Dapsone non-exposed group of Army veterans, SMR = 0.96 (95% CI 0.91, 1.00) compared to the Australian male population. No specific causes of death were significantly more common in this group compared to the Australian population whereas death from diabetes, diseases of the circulatory system and respiratory system were less common than expected.
The overall incidence of cancer was significantly higher among the Army personnel compared to the Australian population for both Dapsone exposed and non-exposed veterans, 7% and 20% higher than expected, respectively. Specifically, both groups had higher than expected incidence for Hodgkin’s disease, melanoma and cancer of the head and neck region. Dapsone exposed veterans also had higher than expected incidence for eye cancer whereas those veterans not exposed to Dapsone had higher than expected incidence for cancers of the lung and connective soft tissue.
However, overall mortality due to cancer was not significantly different from the Australian population for either exposure group. Mortality for specific types of cancer was higher than expected for lung and oesophageal cancer amongst the non- exposed veterans. Dapsone exposed veterans had higher than expected mortality due to cancer of the eye and oral cavity and lower than expected mortality from connective soft tissue cancer.
Direct comparison
The all cause mortality did not differ between the two exposure groups, RR = 1.00 (95% CI 0.94, 1.07). Nor were there statistically significant differences in mortality between the groups for any of the causes of death analysed.
When direct comparison between the exposure groups was calculated, overall cancer incidence was 10% lower amongst the Dapsone exposed group compared to the non- exposed veterans, RR = 0.90 (95% CI 0.83, 0.97). All calculations were adjusted for age. There were no cancer types for which the incidence was significantly elevated amongst the Dapsone exposed veterans. The incidence of connective soft tissue cancer, lung cancer and genitourinary cancer was significantly lower amongst those Army veterans exposed to Dapsone.
Dose response
There was no significant relationship between increasing Dapsone dose and all cause mortality or cancer mortality. There was a borderline significant inverse relationship between Dapsone dose and cancer incidence. That is, with increasing total cumulative Dapsone dose there was a small decrease in the incidence of cancer.
Summary and conclusion
In this study assessing long term health effects from exposure to Dapsone during Vietnam service, mortality was assessed for over 30 years since exposure and cancer incidence follow-up was for 19 years. There was no statistically significant difference in non-cancer mortality between those who consumed Dapsone and those who did not. There was a modest but statistically significant lower than expected overall cancer incidence, and for some specific cancers, a lower than expected mortality for the Dapsone exposed group.
This study concludes that those who took the anti-malarial Dapsone/Paludrine prophylaxis have not experienced adverse health, as measured by mortality and cancer incidence, compared to those veterans who took anti-malarial treatment without Dapsone.
References
1 AIHW. Dapsone exposure, Vietnam service and cancer incidence. Canberra: Australian Institute of Health and Welfare, 1992:149.
xv
