Is that a coincidence? ”: Exploring health perceptions and the causal attributions of physical 2 health conditions in British nuclear test veterans. 3

Since the British nuclear testing programme, there have been several claims in the media and 2 from the veterans themselves that their health (and descendants’ health) has been adversely 3 affected by ionizing radiation exposure. Many health conditions associated with ionizing 4 radiation exposure are also age-related. Therefore, the purpose of this study was to explore how 5 British nuclear test veterans, with varying health conditions, perceive their health and attribute 6 causes to health conditions in themselves and in their family members, in the light of being an 7 aged cohort and their previous involvement in nuclear weapons testing. Semi-structured 8 interviews were conducted with 19 British nuclear test veterans and were analysed using 9 thematic analysis to generate broad themes describing the data. Four themes were generated: 10 (i) Sources of health risk information over the life course, (ii) Luck, (iii) What is ‘normal’?, 11 and (iv) Experience with healthcare professionals. Health conditions perceived as not ‘normal’ 12 considering one’s age, lifestyle, and hereditary risk, or perceived as incurable, appeared more 13 likely to be attributed to radiation exposure. Recommendations relating the transparency for 14 authorities dealing with exposure scenarios, and subsequent genetics and epidemiological 15 research are discussed. Healthcare professionals may benefit from understanding patients’ 16 narratives in healthcare consultations with individuals who perceive radiation exposure to have 17 impacted on their health. 18 19 20 25

involved in underground testing and atmospheric testing at the Nevada Test Site in co-operation 8 with the USA . Despite only 8% of the cohort receiving a non-zero dose 9 and only 37 individuals (predominantly Royal Air Force personnel) receiving a dose greater 10 than 100 mSv (the maximum dose recorded was 300 mSv according to available dose records; 11 Kendall et al., 2004), there have been claims their health and the health of their descendants 12 has been affected by ionizing radiation exposure. Previously, there had been no evidence to 13 suggest an increased risk of cancer incidence or mortality in this veteran population (Darby et 14 al., 1988(Darby et 14 al., , 1993Muirhead et al., 2003Muirhead et al., , 2004) but a recent updated analysis reported a 2% 15 increased risk of overall mortality compared to veteran controls, driven by 2% and 3% 16 increased risk of mortality from all cancers combined and non-cancer diseases (Gillies and 17 Haylock, 2022). Outside of the British testing programme, the incidence of certain cancers  24 The linear no-threshold model (the commonly used model by regulatory authorities for 25 radiation protection) implies there is no threshold for stochastic health effects, such as cancer, 26 to occur (i.e. there is no safe dose level of ionizing radiation) (National Research Council, 27 2005). Despite this, specific attribution of stochastic effects to ionizing radiation is difficult 28 because the effect manifests many years after exposure. Further, confounders such as exposure 29 to other mutagens or carcinogens through lifestyle, occupation or medically mean that 30 epidemiological studies often lack statistical power to support evidence for a radiation effect, 31 especially at very low doses (Shore et al., 2018). As such, veterans of nuclear testing 32 4 programmes may face considerable uncertainty in understanding any health conditions or 1 symptoms in the light of (perceived) exposure. This uncertainty could be further influenced by 2 allegations regarding inadequate protective equipment or inconsistencies in safety procedures, 3 and the limited proportion of British veterans with available dose data (generally limited to 4 external gamma radiation) (for further information please see Arnold, 1987;McClelland, 5 1985). 6 The term 'attribution' refers to inferences that people make about the causes of events or of 7 states of being (Michaela and Wood, 1986). Causal attributions of medically explained and   shares the experience of a singular profound event (e.g. nuclear weapons testing). Therefore, 27 the purpose of this article is to report qualitative findings exploring how BNTVs, with varying 28 health conditions, perceive their health and attribute causes to health conditions in themselves 29 and in their family members, in the light of being an aged cohort and their previous involvement 30 in nuclear weapons testing. 31  Research packs were distributed to 246 British nuclear test veterans listed on a mailing list 2 (compliant with data protection principals) provided by the Nuclear Community Charity Fund 3 (http://thenccf.org/). An advert with information about the study and contact details was also 4 placed in the quarterly charity fund magazine. Fifty-nine British nuclear test veterans consented 5 to being contacted regarding in-depth qualitative interviews. Of these, 29 veterans were invited 6 to take part based on geographical clusters with the aim of conducting between 15 and 20 7 interviews. Fifteen veterans agreed to participate in a face-to-face interview and a further four 8 agreed to a telephone interview. Thus, a total of 19 BNTVs (aged between 75 and 89) were 9 interviewed. 10 Thirteen participants were stationed at Christmas Island (Kiritimati), four at Maralinga, one at 11 the Monte Bello Islands, and one at Malden Island. Fourteen participants witnessed one or 12 more weapons tests, and the participants who did not witness any tests were primarily involved 13 in the clean-up operations. Most, but not all, participants had various health conditions. The 14 specific diagnoses are not reported here to protect participant anonymity. Participants received 15 study information prior to the study, and on the day of the interview they were briefed, assured 16 full anonymity, and were made aware that they were free to withdraw from the study at any 17 time with no consequences. Participants provided informed consent. 18

19
The interviews were biographical in nature. Each interview began by asking the participant to 20 describe what life was like in the few years leading up to their service in the nuclear testing 21 programme. Participants then described their nuclear testing experience and continued until the 22 present day in a chronological fashion. The interviews were semi-structured and loosely 23 followed a schedule covering a broad range of topics: identity, uncertainty, risk perception, 24 health, subsequent life events, and cognitive function. The topic of health came up naturally in 25 most interviews, but otherwise to elicit discussions on the topic of attributions of health 26 conditions, participants were asked "do you have any health conditions?", followed by "what 27 do you think caused these?" depending on the response. Responses were then probed to elicit 28 in-depth data. 29 Since the interviews were biographical, and relied heavily on the recall of past events, face-to-1 face participants were encouraged to have ready photographs significant to their testing 2 experience, drawing on aspects of photo-elicitation (Silver, 2013). Some participants presented 3 photographs taken during the testing programme and photographs taken from medical imaging 4 of descendants, while some participants also presented documents such as health reports, 5 newspaper clippings, brochure-type documents from their service, and other relevant 6 documents (e.g. safety protocols). Individuals can find significance and meaning in objects or 7 documents aside from photographs which facilitates memory and discussion regarding 8 potentially sensitive topics (Barton, 2015). These photographs and documents were used solely 9 to facilitate discussion and were not used in any analysis. 10 The face-to-face interviews took place in the participants homes, while the telephone 11 interviews were conducted in a private, secure room in a university department building. 12 Veteran family members were present in three of the face-to-face interviews and were aware 13 that the purpose of the research was to interview the veteran, specifically. In these interviews, 14 the family members would engage in discussion with the veteran, and these discussions were 15 encouraged to continue to gain further perspectives on the topics. All interviews were 16 conducted before the Covid-19 pandemic. The interviews were audio recorded and ranged from 17 45 minutes to 2 hours 17 minutes in duration. 18

19
The interview recordings were transcribed verbatim and analysed using thematic analysis 20 following the guidelines by Braun and Clarke (2006) to generate themes giving a rich 21 description of the verbal data. A reflexive diary was kept by the lead researcher which included 22 thoughts pertaining to potential themes and any key interactions between participants and 23 family members present during the interview. After all interviews had been recorded and 24 transcribed, the coding process was conducted by hand using highlighter pens and writing the 25 code labels in the margin of the transcript. Codes were generated for varying lengths of 26 transcript segments, ranging from single sentences to a full paragraph depending on the content.

27
By examining highlighted segments as a visual indicator of trends, relevant codes and quotes 28 (obtained from digital versions of the transcripts) comprising potential themes and subthemes 29 were inputted to digital tables and further refined. Notes to facilitate the writing of themes and 30 to identify quotes which may overlap multiple themes were included in the digital table. The 31 reflexive diary played a role in noting areas of importance. Themes were identified with the 1 intention of an inductive approach to analysis.

3
The topics elicited through the interviews were broad: the participants talked about a range of 4 physical, social, and psychological aspects relating to their time in service. While details  Four themes were generated: (i) Sources of health risk information over the life course, (ii) 8 Luck, (iii) What is 'normal'?, and (iv) Experience with healthcare professionals. These are 9 presented generally in chronological order for convenience, where the first theme includes 10 subthemes relating to the testing experience specifically, while the remaining themes relate to 11 experiences and perceptions after the tests.  14 Most participants, especially those who had witnessed weapons tests, were keen to describe 15 their experience of the British nuclear weapons testing programme. Generally, they described 16 the detonation to be an extraordinary and awesome experience, and recollections of the visual 17 aspect of the nuclear weapons tests were vivid. Some participants would recall the weather and 18 other details which would otherwise be considered as mundane. They often recalled the colours 19 of the fireball, the shape of the cloud, and the experience of seeing their bones through their 20 hands due to the strength of the flash. Six participants described the wind or the sensation of 21 heat from detonation. These sensations were sometimes used to demonstrate the perception of 22 being irradiated and, in a couple of interviews, were sometimes linked to specific health 23 problems: 24 Then the heat came and oh boy wasn't that hot. The next few seconds you were looking 25 up at a mushroom and on top of that was a great big ball of fire. I think that's how I 26 got all my problems because the blast came this way. I think that's how I got my knee 27 problems. (Veteran J) 28 8 In addition to the sensate experience, four participants would comment on the non-existent 1 sensate experience of radiation itself. While, with hindsight, they believe the area was 2 contaminated, they were unable to smell, touch or see it at the time, and as such are unable to 3 determine the extent they had been exposed to ionizing radiation. Despite this, six participants 4 described the impact of the tests on the environment around them, for example the wildlife and 5 the trees, and molten sand or rock from the heat of the blast, as an indicator of the risks of 6 nuclear weapons testing. The following quote illustrates this effectively, though it is unclear 7 whether these perceptions were current at the time or in hindsight: 8 I looked up to see, and quite often you can see the fins of sharks and that around the 9 island, but that morning there wasn't a fin to be seen. And there were quite a number 10 of birds but I couldn't hear a bird. And nature warned them but all was left was 11 basically barren rock and bits of sand. I thought, well, if they've done that to the island, 12 whats it doing to us? (Veteran K). 13

Availability of risk information.
14 Many participants described how their present-day perceptions contrasted with their 15 perceptions at the time of the nuclear testing programme. In general, the participants described 16 how they had no initial consideration for risk to their future health. For example, a couple of 17 participants said while at the time they knew what a nuclear bomb was because of their 18 awareness of the atomic bomb events of the second world war, this had limited significance to 19 them in terms of health risk. What appeared pertinent in the interviews was the relatively 20 limited availability of knowledge about radiation, or perhaps, limited access to the knowledge 21 existing at the time compared with the present day: 22 Information in those days was quite scarce. A lot of those leaflets published in the 1950s 23 weren't available in those days. Communication was entirely different so you couldn't 24 go with anything because that didn't exist. So as regards to us sitting over hundreds of 25 miles away in the middle of the pacific, the information that you got or was available 26 was very, very limited. Very limited. (Veteran Q) 27 The perceived limited risk during the tests may also be influenced by their young age at the 28 time. For example, nine participants mentioned this, suggesting that young adults are generally 29 unlikely to be concerned about health risks in this context. For most participants, the limited perceived health risk persisted throughout the next decade 2 following their service. The actual duration of this varied between veterans due to specific 3 events that followed, such as the birth of descendants, experiencing health conditions, 4 formation of the British Nuclear Test Veterans Association (BNTVA), and the emergence of 5 press reports about nuclear testing health effects. In one case Veteran N who witnessed 6 weapons tests at Christmas Island, had little concern over his lifetime for the risks of radiation 7 exposure on his health:  16 Even in the scenarios where first-generation descendants were born with serious health 17 conditions, or partners' miscarriages, these few veterans said they initially had no reason to  26 The 1980s appeared to be a central transition period, marked by the formation of the BNTVA, 27 and the emergence of media reports describing perceived health effects and mortality attributed 28 to ionizing radiation exposure. For several veterans, attending BNTVA and other veteran 29 meetings were sometimes described in the context of perceived health risk. Some told of their 30 experiences socialising at these meetings and learning of health issues in other veterans, or 1 learning through BNTVA magazines: 2 On top of that, my wife had two miscarriages, and nobody could think of the reason 3 why. Not long after that I went to Blackpool for a meeting with the nuclear veterans 4 and I found that quite a number of the chaps that were there had lost babies in the same 5 way and could only surmise that it was through radiation. (Veteran K) 6 To summarise the first theme, any perceived radiation risk to participants' health or family  3.2. Theme 2: Luck. 18 The notion of luck was especially salient across the interviews and was discussed in many 19 regards. For example, many participants described themselves as 'lucky' regarding their own 20 health, their survivorship, or their family's health. Eight of these expressions of luck were 21 described in the light of involvement in the weapons testing programme. Furthermore, one 22 participant (Veteran Q) described himself as lucky to have witnessed a nuclear weapons test. 23 As such, luck might be a central part of the nuclear test veteran identity, for example Veteran  12 any effects of me. (Veteran L) 13 As indicated above, some participants' (3) perception of luck stemmed from reading media 14 reports of nuclear test veterans elsewhere. Interestingly, Veteran D pointed out that it is 'bad 15 news that sells print' and stated that the press is unlikely to print stories about a nuclear test 16 veteran without health problems. This suggests that the perceived luck may be influenced by 17 media reports disseminating negative health outcomes in British nuclear test veterans. 18 The perception of luck may also be linked to the general ageing process. For example, 19 following a discussion regarding multiple transient ischaemic attacks, which he did not 20 attribute to radiation exposure, Veteran D notes: "You're very lucky to go through life without 21 getting something wrong with you". 22 Veteran M also alludes to luck in relation to general ageing:  29 It is interesting to note that Veteran M draws comparisons between himself and another veteran 1 who did national service, albeit not in the context of any nuclear weapons testing. He 2 demonstrates that arthritis can occur in veterans with no perceived ionizing radiation exposure. 3 These comparative processes form the next subtheme labelled 'What is normal?'. 4 3.3. Theme 3: What is 'normal'? 5 3.3.1. Social influence on health perceptions. 6 The social processes veterans use to understand which health conditions are attributed to 7 radiation exposure was a prominent feature.  16 Above, Veteran D compares his situation to non-veteran friends and asks himself whether it is 17 a coincidence or not. In answering this, veterans may identify ionizing radiation exposure as 18 the unique factor distinguishing themselves from others. Like Veteran D, Veteran F draws 19 comparisons between his immediate family and the families of his close relatives to understand 20 his wife's ectopic pregnancies: 21 Two had two children, one had three children, one had four children. Her brother had 22 three kids. So, all our close relatives had reasonably sized families and that made me 23 think "oh crikey, why me? Why me?". Or rather "why us?". That's me being a me, me, 24 me. Why us? That's when I started thinking about it and started thinking more about 25 things and looking more into it. Thinking, "hang on..my medical records weren't there. 26 Is there a connection?" (Veteran F) 27 Similarly, Veteran H perceived that a "normal father" is one who has not been exposed to 28 radiation, and this was used as his explanation for his descendant's health condition. 29

Perceived age norms and lifestyle influence.
1 Participants would also consider what the "normal" health conditions are that can be expected 2 for a male older adult while accounting for lifestyle factors. It appeared that the conditions 3 which are perceived as common or expected given one's age or lifestyle are less likely to be 4 attributed to radiation exposure. As such, most health conditions experienced by the veterans 5 in this study were generally not attributed to radiation exposure because they were not 6 experienced early in life, or they could be explained by lifestyle or heredity factors.

7
To illustrate, Veteran D described how heart attacks and knee pain, for example, are unlikely 8 to be caused by any exposure because they are also observed in older men and in active 9 individuals through wear-and-tear, respectively. Veteran B described how he doubts the 10 possibility that his heart condition could have been caused by exposure due to developing the 11 condition as an older adult. Veteran M described how his polyps could be explained by his 12 prior occupation which had scientifically been linked to polyps and cancer. Veteran R, who 13 was involved in the clean-up operation at Maralinga, explains why prostate cancer and lung 14 conditions could be expected due to his age and lifestyle: 15 I mean a lot of it is passing it on, you know, having children who were born...trying to 16 think of the correct expression um...to put it like thalidomide children, for instance, you 17 know you read instances and you can understand it. You know, they're probably a bit 18 23 This focus on age when understanding which health conditions are "normal" could be 24 speculated to apply to their descendants' health conditions. While it was not explicitly 25 discussed in relation to age, there appeared to be a tendency to attribute any health conditions 26 in descendants to the veteran's radiation exposure, relative to the veteran's own health 27 conditions. This could be because health conditions, generally, are less common in younger 28 adults than older adults. 29 To further illustrate, one topic explored was cognitive health. This was elicited by asking "do 1 you ever think about your memory?" Given that declining memory was often viewed as a 2 typical sign of ageing, the veterans did not perceive any declining memory to be related to 3 radiation exposure. This was further reinforced through instances where veterans described 4 their wives' cognitive health (and age-similar others) as being comparable to their own, like 5 the above processes of considering what conditions are normal for someone of that age.   The experience with healthcare professionals was also a prominent feature in the discussions. 9 Generally, the interactions between the veterans and healthcare professionals often described 10 uncertainty regarding attributing their conditions and symptoms to ionizing radiation: 11 12 is there any connection to what I've got?", you know, because I thought it might've 13 been something that comes with it, I don't know. And he did say, he said, typical 14 surgeon, he said "it could be, but I can't stand here and categorically say it did" 15 [laughs]. (Veteran P) 16 While most participants did not have any real resentment at the standard of healthcare that they 17 had received, three participants noted the reluctance of healthcare professionals to confirm that 18 a health condition was caused by ionizing radiation exposure. Some participants acknowledged 19 the medical uncertainty regarding many health conditions, even outside the context of ionizing 20 radiation exposure. But one veteran suggested that the ambiguity regarding causes for certain 21 conditions can lead to an easier explanation, namely age: 22 But whether any of that is attributed to radiation, who can say? Because 4 years ago, 23 of course was 76 so age-Always get away with it with age. Whatever you've got they 24  In a separate case, Veteran K described healthcare professionals detecting radioactivity in his 15 bone marrow when he was taken into hospital for an unspecified illness. Later in his life, 16 following an operation on his knee, radiation in his bone marrow was given as a suggestion for 17 his knee problem. This also appeared to influence Veteran K to attribute later health conditions 18 in himself and in his wife (miscarriages) to ionizing radiation, despite healthcare professionals 19 being unable to determine causality.  While most participants did not explicitly describe any real dissatisfaction with healthcare 22 providers, the notion of trust appeared important, where a couple of veterans expressed 23 disbelief in healthcare professionals' alternate explanations for health conditions. For example, 24 some veterans' skin cancers and keratosis had been attributed by healthcare professionals to 25 sun damage, which was treated with scepticism because they do not sit out in the sun often. 26 Despite Veteran F acknowledging that Maralinga is a location marked by strong sunlight, he 27 nevertheless maintains ionising radiation from weapons testing as a possible explanation for 28 his skin conditions. The scepticism towards healthcare providers suggestions is further 29 illustrated by Veteran I: 30 But as soon as I mention "do you think it is some sort of development from radioactive 1 stuff?", "Oh no, no, no, no. Nothing of the sort." How can they be so quick to say that 2 when they don't know? You're only asking the question, and never got an answer, so. 3 (Veteran I) 4 Veteran I's scepticism was explained by stating that most medical staff were not old enough to 5 be present at the nuclear testing programme and, like most of the public, they have limited 6 knowledge or awareness of the British testing programme. In two cases, medical staff were 7 perceived to avoid the topic of radiation exposure, possibly due to the political nature of the 8 nuclear weapons testing programme. For example, Veteran J described feeling let down, due 9 to the perceived reluctance of healthcare professionals to consider ionizing radiation as a 10 potential cause: 11 The only problem with the hospital, I used to say, "this being I've got bone problems, 12 anything to do with Christmas Island?" "Don't want to go there" he said. "Nothing to 13 do with us, not interested". And I thought "great, thank you very much." Another one 14 said, "we're not allowed to talk about it." I said, "talk about what?". "Well, your 15 problems and how you got them. Nothing to do with us". And I thought "well why can't 16 someone just say I'm not allowed to talk about it and do this that and the other but get 17 your doctor to write to Mr or Mrs so and so who deal with that sort of thing". But no, 18 no one would. 19

And when they said, "yeah we don't want anything to do about it". Can you describe
20 how that felt? 21 Well, I felt sort of let-down. Still no one. Nobody wants to know. (Veteran J).  23 Overall, we have presented how British nuclear test veterans understand the causes of their 24 health conditions, how they perceived health risk over the life course, and we have highlighted 25 their experiences with healthcare providers. These findings can inform general practice as they 26 are exemplars of when understanding the context of health beliefs might improve satisfaction 27 with health service. It must be noted that most participants in this study did not explicitly state 28 dissatisfaction with healthcare received. It is difficult to compare the level of dissatisfaction 29 with healthcare providers with similar recent studies with Gulf War veterans (Bloeser et al., 1 2021), but we do observe a parallel finding: some BNTVs indicated dissatisfaction at the 2 provider's attribution of some health conditions to ageing due to a perception that it was an 3 'easy' explanation. 4 The fact that some veterans may be convinced that some health conditions are caused by 5 ionizing radiation despite receiving alternative explanations from medical staff raises important 6 questions regarding the perceived validity of medical attributions, and suggests that 7 participants' health beliefs can be as strongly influenced by interactions with fellow nuclear 8 veterans and health risk information in the media. This is not to say that attributing physical 9 health conditions to radiation exposure in this context is irrational, because ionizing radiation  (Kasperson, 2012). 26 We argue that narrative-based medicine could have brought benefits to nuclear test veterans' one has been exposed may increase one's perceived risk for genetic health conditions. This

17
Our findings also shed light on how this unique population attributes causality to their health 18 conditions. It appeared that health conditions considered as incurable were generally attributed 19 to ionizing radiation exposure (in the absence of an alternate explanation), demonstrating that 20 individuals may turn to profound environmental events when interpreting the causes of such 21 health conditions. Generally, it is the causes which are thought to vary in their perceived 22 controllability/curability rather than the health condition (Shiloh et al., 2002). The present study 23 indicated that, in a few participants, perceiving a health condition as uncontrollable/incurable 24 preceded attributing to ionizing radiation. However, it could be that attributing a condition to 25 ionizing radiation precedes perceiving a health condition as uncontrollable/incurable and the 26 veterans point to the characteristics of their health conditions as evidence for their attribution, 27 but this particular sequence cannot be verified in this study. 28 Another factor was whether the condition was considered 'normal' for a non-irradiated older 29 adult while accounting for lifestyle factors (e.g. smoking). Health conditions in themselves or 30 family members perceived as being not 'normal' relative to the age at which a condition was 31 realised appeared more likely to be attributed to ionizing radiation exposure. Thus, social 32 comparison processes appeared central to participants understanding the causes of their health 1 conditions: observing similar health conditions in other nuclear test veterans may lead one to 2 attributing the conditions to ionizing radiation. Likewise, observing similar health conditions 3 in similarly-aged others (e.g. a spouse) may serve as reassurance that the health condition is 4 caused by ageing. This was evident particularly when considering the perceived impact of 5 ionizing radiation on cognitive decline. 6 While age was not explicitly discussed when considering descendants' health conditions, we 7 can speculate that it plays a similar role in determining the causes of health conditions in family 8 members. The veterans tended to attribute any health conditions in descendants to paternal 9 ionizing radiation exposure (involving genetics), presumably because serious health conditions 10 are generally uncommon in younger/middle-aged adults. This is an interesting scenario since 11 attributing one's own health conditions to uncontrollable causes (e.g. genetics) appears to 12 improve mental health outcomes by shifting responsibility, in comparison to behavioural . This complicated scenario must be addressed which may be a 18 fruitful avenue for promoting psychological well-being in exposed older populations. 19

20
Overall, this article explores how British nuclear test veterans' perceive their health and how 21 these veterans understand the causal nature of any existing health conditions. Unlike most 22 research examining perceived causal attributions, it is difficult to provide a comprehensive 23 framework for exposed populations because the participants vary greatly in their health 24 conditions. Due to relatively small numbers of nuclear test veterans, it is not feasible to have a 25 sample where health conditions are all standardised. None the less the findings highlight causal 26 attribution processes in a unique older population which has not yet been examined in this 27 context. We can offer a set of implications which one can speculate are rooted in the issue of 28 transparency and the marked ambiguity regarding dose levels and related health effects: i) 29 authorities dealing with exposure scenarios must emphasise transparency to avoid amplifying