British Nuclear Test Veterans Association

	Join Us		Timeline		Press		Useful Links
	Mailings		Roll of Honour		Events		Discussion Board
	Search the Site		Recollections		Contact Us		Medical Research

Mortality and Morbidity of Members of the British Nuclear Tests Veterans Association and the New Zealand Nuclear Tests Veterans Association and their Families

SUE RABBITT ROFF
University of Dundee Centre for Medical Education

Published in Medicine, Conflict and Survival (Volume 15, Supplement 1, July-September 1999)

Part Five

General Discussion

Sample construction

The NRPB report claims to have identified 85% of the men who were sent to the UK nuclear weapons tests through data linkage of various sources. These included government records and those of ex-servicemen's organizations such as the Royal British Legion and the British Nuclear Tests Veterans Association. The present research, using the BNTVA records constituting 10% of the NRPB survey, acts as a cross-check on the NRPB study.

According to the NRPB:
For the study to be definitive, either all the participants needed to he included or the authors needed to be able to show that those who were included were fully representative of all those who had participated. At the time of the test programme, no comprehensive list of participants was compiled, and this made it unlikely that after so many years the authors could be certain to identify every person eligible for the study. Therefore, names and identifying details of participants were sought from many sources other than MOD. People identified in this way include individuals who notified themselves as test participants, or who were notified by a third party. The identification may have been made by contacting a government department or some other public body (for example, NRPB, the BBC or the University of Birmingham) or through one of the organisations of test veterans who have assisted NRPB.

Assumptions about exposure

The NRPB researchers had access to exposure records for some men but they were a small minority of those who were present at the tests. There are many anecdotal reports of the dosimetry badges that were issued not being properly collected or recorded. These badges in any case only monitored small areas of the body; they were not capable of recording alpha, internal or committed doses. Caution should therefore be exercised in deducing health hazards from such scanty reconstructed dosimetry; it is more important to look at the outcome endpoints of potentially radiogenic illness in the men and then consider the role of long-term chronic exposure to residual, induced and resuspended radionuclides.

On page 4 of the NRPB report, it is stated that 'Men were not included if they had been involved only with peripheral activities associated with the test programme, such as weather forecasting or the handling of non-radioactive stores and supplies, at other locations' (that is, excluding RAAF Pearce in Western Australia and RAAF Edinburgh Field in South Australia).

Later, sub-groups regarded as at some risk of radiation exposure are listed (pp.8979):

At the start of the study NRPB was informed by MOD that only a small proportion of test participants were likely to have been exposed to radiation as a consequence of their test participation. The relevant groups of personnel were:

(i) the members of the crew of HMS Diana which sailed through the fallout plumes in Operation Mosaic

(ii) the members of the Buffalo Indoctrinee Force, a group of volunteer officers assembled to observe at first-hand the effects of a nuclear explosion;

(iii) RAF aircrews involved in radioactive sampling from the clouds of the explosions;

(iv) the RAF active handling flight, who decontaminated aircraft used in cloud sampling;

and

(v) individuals not in groups (i)-(iv) but who had recorded radiation doses greater than zero.

This resulted in the identification of 22,347 'test participants' who formed the sample for the NRPB studies. 'Of 22,347 test participants included in the study, only 1804 (8%) are believed by the MOD to have been liable to exposure to radiation. The proportion was much higher for AWRE personnel, 409 (50%) of whom were included in a special group' (p.9).

The NRPB report² (p.76) concludes that:
There was no increase in risk with measured external dose and no special accumulation of cases in men identified by MOD as liable to be exposed to radiation, in men employed by AWRE or involved with the minor trials at Maralinga, or in men present at one or other of the tests, or specifically at the tests at Christmas and Malden Islands, who include any men known to have been exposed to neutrons or thought by MOD to be the ones likely to have ingested or inhaled any radionuclides that would have escaped measurement on the dosemeters. Indeed, the greatest (or equal greatest) RR, the most highly significant difference from the controls, and the highest SMRs for both leukaemia and multiple myeloma were all found in the group of 'other test participants' after excluding the small number who, on any assumption, were unlikely to have been exposed to more radiation than the general public. These 'other' men had been involved in the test programme in a variety of ways: just under 60% of them had visited Christmas Island, but not during one of the operations listed in Table 3.1, and just over 30% had visited Maralinga, but were not known to have been involved in the programme of minor trials or to have been present during one of the major tests. Most of the remaining visits had been to the Monte Bello Islands either before or after tests in the Mosaic series. According to MOD, the experience of men in this group ... is, on all counts, likely to be less than for groups A and B in the same table. A comparison of the 11 men who developed leukaemia (other than chronic lymphatic leukaemia) or multiple myeloma with unaffected participants in the same group, failed, however, to highlight any characteristics that distinguished them.

Reservations about completeness of cohort

The authors of this cohort study⁷⁷ of mortality in test veterans, using the UK National Health Service (NHS) Central Register, discuss the completeness of follow-up. The details of men for whom follow up was achieved but unsatisfactory using the NHS Central Registers were submitted to the Department of Social Security (DSS) for further tracing. Information on men born before 1916 and reported by the NHS Central Registers as currently registered with a general practitioner, and a 1% sample of remaining men born in or after 1916, were also submitted. The additional follow up increased the number of deaths fully identified in the cohort by 6.5%. Mortality among those untraced on the NHS Central Registers was substantially greater than in the cohort as a whole (10.2% compared with 6.9%). Among those reported by the NHS Central Registers as not currently registered with a general practitioner, 2.7% were found to have died, as were 1.1% of men born before 1916 and currently reported to be registered with a general practitioner. There was clear evidence that information about emigrations supplied by both the NHS Central Registers and DSS is far from complete. The authors concluded that standardised mortality ratios based on follow up via the NHS Central Registers alone are likely to be somewhat low, and this should be borne in mind when interpreting the data.

The report itself notes (p.72) that there were particular problems with locating the records of precisely those individuals who made claims for ill-health arising out of their participation in the nuclear weapons programme.

Reservations about standardized mortality ratio methodology

In evaluating radiation hazards, analytic studies have utilized either the cohort type of investigation (where persons exposed and not exposed to radiation are followed forward in time for outcome in respect of disease), or the case-control approach (persons with and without a specific disease are evaluated for previous exposure to radiation). Most radiation studies have evaluated cohorts (such as radiologists), although important case-control studies have been conducted (for example, childhood leukaemia as related to prenatal X-ray). At its best, according to a senior US researcher,⁷⁸ epidemiology is capable of evaluating relative risks (RR) on the order of 1.4 (i.e. a 40% relative excess). However, the RRs of interest following low doses of radiation (0.Olgray) are on the order of 1.02-1.002. Thus, not much should be anticipated from direct observations at this level, and indirect approaches must be taken to estimate low-dose effects. Such indirect approaches include the study of:

• populations exposed to a range of doses, both low and high, where interpolation models can be reasonably applied to estimate low-dose effects; and
• populations exposed to fractionated doses over a long period of time where the resulting dose-effect relationship theoretically should be linear and the estimation of low-level health effects facilitated.

Proportionality mortality analyses are traditionally considered to be unreliable because they lack information on persons at risk?⁷⁹ Standardized mortality ratios based on follow-up through the United Kingdom National Health Service central registers alone are likely to be somewhat low,² and this should be borne in mind when interpreting the data. No reduction in proportionate mortality from cancer among members of the Vegetarian Society was found, even when allowing for the possibility that individuals joined that society because of ill-health, and those with cancer might be especially likely to do so.¹ As the authors point out:
The proportionate method of analysis has obvious limitations, for since the proportions must add up to unity, any 'real' deficiency of a major cause of death will tend to inflate the values for other diseases, and vice versa. If vegetarians differed from the general population in having, say, an appreciably reduced mortality from one disease (or one group of diseases), then this should be detectable by the present method. However, if vegetarians experience an altered mortality from several major diseases [for example the radiogenic cancers], this would probably not be evident using this method.

Correction for 'healthy soldier' effect

The NRPB report² acknowledged the strength of the 'healthy soldier' effect (p.74):
'Another reason for the low mortality rates observed in the study is that all ranks who served in the tropics and sub-tropics were selected for physical fitness. This might have had an effect throughout the study, but it would certainly have had a substantial effect on the mortality from neoplasms and from all non-violent causes of death in the early years. It provides an explanation for the fact that the SMRs for neoplasms rose (for participants and controls combined) from 65 in the first 5 years after the start of the observation, through 72 from 5-15 years after the start, to 86 for subsequent years and for all non-violent causes of death from 55 through 65 to 76.

But the conclusions precisely do not correct for this potential bias:
The difference between the two groups in the mortality from leukaemia and multiple myeloma (22 deaths from leukaemia and 6 from multiple myeloma in participants, against 6 from leukaemia and 0 from multiple myeloma in controls) was largely due to extraordinarily low rates from these diseases in the controls (SMRs, respectively, of 32 and 0), while the mortality in the participants was only slightly greater than expected from national rates (SMRs, respectively, of 113 and 111) and much of these differences seems likely to have been due to chance ... It is concluded that small hazards of leukaemia and multiple myeloma may well have been associated with participation in the nuclear weapons programme, but that such participation has not otherwise had a detectable effect on the particpanrs' expectation of life or on their total risk of developing cancer. [Abstract]

The four disease categories of Hodgkin's disease, non-Hodgkin's lymphoma, multiple myeloma, and leukaemia ... together constitute the broad group of 'cancers of lymphatic and haemotopoietic tissue'. Altogether there were 51 deaths from this group of cancers in the test participants and 28 in the controls, and the relative risk was estimated to be 1.65, which is significantly increased (p=O.02;90% confidence interval 1.08 to 2.51). The difference was not, however, due to a high mortality in the test participants in whom the number of deaths was equal to that expected from national rates (SMR100) but to a low mortality in the controls in whom the number was only just over half that expected (SMR 56, p>0.00l). (p.48)

In effect, instead of correcting for the 'healthy soldier' effect (which has been shown to be capable of halving the relative risks for some occupational groups) the interpretation of the findings of NRPB-R214 rests, in effect, on dividing the hazard by the healthy soldier effect rather than multiplying it. This was compounded by the initial comparison of two veterans groups in the construction of the sample and the control, then calculating the SMRs against the general population to reach the Relative Risks. It should be noted that even so, the RRs for leukaemia and multiple myeloma were markedly heightened for the sample group. At the same time, in another study,⁸⁰ an increased mortality in the younger age groups from multiple myeloma during the period 1950-79 was reported.

An account⁷⁵ of the Grapple test notes that HMNZS Pukaki and HMNZS Rotoiti, the New Zealand ships present, 'were stationed at varying distances of 20 to 150 miles from ground zero throughout the series of tests' and 'visited Christmas island after the tests, and it has been hypothesised that rainout into the lagoon and concentration in the food chain could have occurred ... Thus although currently available data indicate that the RNZN personnel in Operation Grapple probably received very low doses of gamma radiation, the possibility cannot be excluded that there could have been significant external exposure to neutron radiation or internal exposure due to inhalation or ingestion.' It is further commented that 'The strongest reason for concluding that the leukaemia findings may reflect a causal relation is that a similar excess risk was found in the previously published study of British participants in the same nuclear weapons testing programme'.

Techniques for data collection

Information for the New Zealand survey⁷⁶ was obtained from electoral rolls, driver's licence or car registration, and hospital discharges as well as mortality records and the Cancer Registry, and supplemented by a postal questionnaire. Identification of subjects in the earlier report's had utilized the resources of the Ministry of Defence, the Returned Services Association, and the general practitioner who had first raised the need for the study. This was supplemented by press advertisement and led to identification of 536 test participants. It is not clear how many subjects responded to the postal questionnaire. The authors do not seem to consider that the publicity surrounding the construction of the sample confounded the results or attracted a selective sample.

Exposure reconstruction

HMNZS Pukaki and HMNZS Rotoiti were under the command at the UK Royal Navy at the 'Grapple' hydrogen bomb tests at Malden and Christmas Island in 1957 and 1958. They served as weather ships, with as secondary tasks Air Sea Rescue, anti-submarine watch, thermal flash measuring and water sampling. Roy Sefron, Chairman of the NZNTVA, testified in September 1997:
The Test procedures ahoard RNZN frigates were basically the same as those aboard the UK Royal Navy ships. At each test explosion only an absolute minimum of men (6) were left internally to run the ship. As many men as possible were required to be on deck to observe the explosions during tbe line up procedures. Initially the procedure was quite rigid with men being appointed Blast Stations on the side of the ship facing detonation.
Protective clothing was worn at Grapple 1 and 2. No protective clothing was worn at Grapple 3. Only goggles were worn at Grapple X. At Grapple Y no protective clothing, but crew were instructed to face away from the blast. At Grapple Z1 Pukaki was 28 nautical miles from ground zero, no protective clothing was worn. At Grapple Z2 Conventional Anti Flash gear was worn (35 Naut miles from ground Zero). At Z3 Conventional Anti flash gear was worn (35 Naut miles from Ground Zero) At Z4 20 nautical miles from Ground Zero there is no record of protective clothing being worn.
Both Pukaki and Rotoiti were to alternate as 'close in observer ship'. The distances from ground zero ranged from 150 to 20 nautical miles.... On the 28th of April 1958 when Grapple 'Y' was exploded, ships' company wore NO protective clothing, the only precaution taken was to face away from the blast with eyes closed until 15 seconds after detonation. It is officially reported and remembered by crewmen, that the resulting bomb cloud had spread and drifted across the whole sky and was clearly distinguishable OVERHEAD at sunset.

Pukaki's captain, Lt CDR Elliot, is recorded to have remarked 'I hope it does not rain'. Elliot was clearly aware of the dangers posed by rain out. At 910 hour during her return to Christmas Island, Pukaki sailed through Ground Zero. No protective clothing was worn. Sea water monitored from the boiler room inlets detected levels of radiation. Lt Cdr Elliot described later in an Auckland newspaper that Pukaki had sailed under the detonation cloud for 23 hours before returning to Christmas Island, sailing through Ground Zero en route.
Because of a lack of fresh water on board both frigates, rain squalls on the horizon were often chased so that crewmen could shower in them and wash clothes. There is no record or recollection of such water being tested for radiation. This rainwater may have been purposely drunk, or accidentally ingested during showering. Seawater was distilled for domestic purposes, including drinking and cooking aboard the frigates. As such there was an obvious risk of irradiation through ingestion if this water was contaminated.
Generally after each test the frigates would return to Christmas Island. Every effort was made to afford leave to crewmen to go ashore. The general activities were recreational, swimming, catching and eating fish, competitive sport, socialising etc. There was also a fish and chip shop ashore. This shop was popular with our seamen. The fish were caught locally. Exploring the island, which had no restricted areas or warning signs of no entry or no fishing, was popular also.

Grant Howard⁸¹ quotes Commander Hale of the Pukaki describing the events of 15 May 1957:
I then opened my eyes, stood up and faced the burst. And then I removed my goggles. The fireball was just starting to grow in size, easily visible and well above the horizon. At plus two to three minutes the blast wave brought double pressure on the eardrums, and it was followed closely by the double rumble of the explosion.
For the first two or three minutes, the fire ball grew in size, shaped like a turbulent cauliflower and changing from an angry, deep red streaked with grey, to a larger smouldering ball of cloud with a glowing centre. Evidence of the intense heat remained visible for seven to eight minutes. Between the second and third minutes a terrific updraft of air and cloud soon became apparent by what seemed to be a strikingly white water spout being sucked up into the centre of the fireball. This rising mass increased in volume until the more familiar but equally fantastic shape of the mushroom was evident to everyone.
The mushroom did not stay for long. After six or seven minutes the upper atmosphere winds began to blow the top of the cloud flat. Then the stalk fell away and was lost to sight. Between the eleventh and twelfth minutes a speck of blue sky was seen in the midst of the nuclear cloud. Finally, the deadly cloud formed a huge smoke ring, which hung in the sky until about four o'clock in the afternoon, when it slowly dispersed

More bomb tests followed. The second was larger than the first but the weather not as good for observation.

Nothing was seen on board the Pukaki until about 11 minutes after the blast, and Commander Hale recorded that there was 'little character' to the mushroom cloud that followed. Conditions for watching this (third) bomb test were better than for the second, although on this occasion the Pukaki was 150 nautical miles east of the bomb zone. The flash of the exploding H-bomb was seen to cover the horizon 'like red-tinted sheet lightning', and approximately three minutes later the mushroom was seen towering above the other clouds. 'On leaving the bridge twenty or more minutes later, I clearly felt the pressure wave on my ears in my cabin', Commander Hale noted. 'It was followed by the noise of the explosion'.

Another veteran remembered:
Grapple Y, 28th April 1958 the Bomb in question went wrong. A megaton explosion, about 18 of them as far as I can make out, was supposed to be air burst. The RAF pilot of a Lysander assigned to duty that day pointed out the lower two to three hundred feet of the fireball was missing. Instead of detonating at 8000 feet she had gone off at about 1200, sucking sand water and general radioactive debris into the cloud core and throwing it into the atmosphere. On Christmas Island rain clouds gathered and a heavy downpour began. At sea the RNZN frigate Pukaki continued with the day's drill and scaled down its crew's ABCD status.

In December 1997 the NZ War Pensions Agency accepted the opinion of Dr John C. Probert, Associate Professor of Radiation Oncology at Auckland University, that the malignant meningioma presenting 26 years after a Pukaki veteran's presence at the Grapple tests and causing his death in 1995 was probably induced by exposure to ionising radiation at that time. In his opinion 'there is a strong probability that [the malignant tumour] was caused by ionising radiation from the nuclear detonations. This opinion is held because of the rarity of the tumour, its strong association with low dose ionising radiation, the two decades before it presented clinically, and the presumption that Mr X received low dose radiation while on the HMNZS Pukaki.' Dr Probert cited the report¹² of the high incidence of meningioma in Hiroshima and Nagasaki survivors.

Hiroshima and Nagasaki

Recent studies of the survivors of the bombings of Hiroshima and Nagasaki are revealing a similar pattern of morbidity and mortality to that reported in this study. One report shows an increase of malignant disease and also of anaemia, hypertension, heart and cerebro-vascular disease, gastrointestinal disorders, including diseases of the gallbladder and pancreas.⁸² The authors note that the Radiation Effects Research Council found, in the statistics of the confined survivor group in 1966-87, a higher excess relative risk of circulatory diseases (cerebro-vascular and cardiac diseases) and gastrointestinal diseases (especially hepatic cirrhosis) in those who were young (below 40 years) at the time of the atomic bombings and whose radiation dose was more than 200 rem. They postulate the existence of a Genbaku Bura-Bura syndrome, which is difficult to explain by a single mechanism. The syndrome is considered to result from complex interactions subsequent to multiple changes in the body after radiation exposure:

• Damage to stem cells: decreased immunological activity, anaemia, skin disorders and gastro-intestinal dysfunction
• Injury to the central and autonomic nervous system: gastro-intestinal dysfunction, circulatory symptoms, dysfunction of the autonomic nervous system
• Injuries to the bone: symptoms of damage to the motor system, such as lumbago
• Others.

A Japanese group^83-86 has pointed out that the pattern of appearance over time of radiation-induced cancer other than leukaemia differs from that of leukaemia. In general, radiation-induced solid cancer begins to appear after attaining the age at which the cancer is normally prone to develop (so-called cancer age), and continues to increase proportionately with the increase in mortality of the control group as it ages. Sensitivity to radiation, in terms of cancer induction, is in general higher for those who were young at the time of the bomb than for older persons.

There have been several reports of cardiovascular disease among the subjects of the Life Span Study of Hiroshima and Nagasakij^87-90 especially among those who were younger than 40 years old at the time of the bombing. Wong and co-workers⁹¹ have reported that the incidence of myocardial infarction has been increasing recently among the younger heavily exposed Japanese survivors. Other recent srudies^92-97 have also shown that irradiation can induce cardiovascular disease, including aortic calcification, and pulmonary fibrosis.

Spinal radiotherapy for ankylosing spondylitis

Review of the long-term mortality after a single treatment course with Xrays of patients with ankylosing spondylitis indicates an increased incidence of leukaemia and of colon cancer associated with ulcerative colitis. There appears to be about a 50% increase in mortality for a wide range of diseases, including cardiovascular, cerebrovascular and gastrointestinal conditions similar to those reported for the present sample of nuclear veterans.^98-105 ~ A review¹⁰⁶ commented that 'In the spondyliris patients multiple myeloma and cancers of the bladder and liver were among the few types of cancer for which higher relative risks were observed more than 25 years after exposure than in the earlier period', and noted that there is increasing evidence of 'different temporal patterns of radiogenic risk between the different cancer types as well as for leukemia'.

Soviet nuclear workers

A 'chronic radiation sickness/haematopoietic syndrome' reported^107-1152 for workers in the Mayak nuclear complex between 1948 and 1958 seems similar to the long-term sequelae experienced by the UK veterans: the Mayak region has been described as the most radiation-polluted area on the globe.¹¹³ The incidence of thyroid cancer has increased in the area irradiated after the Chernobyl accident,^114-115 and studies are in progress on the health of the clean-up workers.^17-19

Radiobiology

Newer radiobiological techniques can detect at the cellular level, effects of radiation which could predispose to malignancy,^116-117 perhaps some time after exposure. These techniques are applicable to individuals who have been exposed to radiation in the past.¹¹⁸

Read the Recommendations