In general, musculoskeletal symptoms in divers and offshore workers had an association with observable physical abnormalities, but the relationship was not strong. The parent questionnaire study  found that musculoskeletal symptoms were 7% commoner in divers, and the observation that symptoms do reflect observable abnormalities to some degree might support the concept that musculoskeletal problems may be commoner in divers in a larger population than studied here. Musculoskeletal symptoms, however, were also related to the expression of other unrelated symptoms in divers. Multiple symptom reporting in the absence of identifiable disease can imply a tendency to somatise in a population [20, 21], in which case symptom reporting might be taken as an unreliable indicator of physical illness and would not be justified as part of any health screening process.
Musculoskeletal symptom reporting was also associated with poorer performance on several questionnaires designed to assess quality of life, mood, executive function and memory. Only SF36 PCS, however, fell below the normative value of 50 (standard deviation (SD) 10) . The population norm for SF36 MCS is also 50 (SD 10), and the values from this study are within this range. Values from normal populations are also available for the other measures used: HADSa: median 5, IQR 3 to 8; HADSd: median 3, IQR 1 to 5 ; PRMQ: mean 38.9, SD 9.2 ; CFQ: mean 43.5, SD 17.0 ; DEX: mean 20.8, SD 9.6 . Although associated with musculoskeletal complaint in divers, at group level, the scores for these measures did not depart from the norm. Further, while an association between indicators of musculoskeletal dysfunction and physical quality of life is logically to be expected, an association with the other factors identified here is less so. Depression and anxiety have been associated with musculoskeletal pain in patients attending an outpatient rehabilitation clinic  using the HADS questionnaire. Unlike this study, however, scores were outside the normative range for both anxiety and depression. Pain has also been associated with disruption of cognitive function, attention and memory [24–26]. The level of pain considered in these studies, however, was greater than that in the present study, being sufficient to cause victims to seek medical treatment and associated with significant departures from normality. While there may be an association between pain and the factors studied, an alternative hypothesis might be considered more likely at the level of effect seen here. The divers that expressed symptoms of moderate to severe musculoskeletal symptoms were more likely to express symptoms of any kind. It might be expected, therefore, that any questionnaire which required them to report symptoms would return a higher score than in a control population. Rather than being a specific indicator of any dysfunction associated with pain, therefore, higher scoring might well be a non-specific response indicating more health-related concern and somatisation in this group. This hypothesis is supported by the absence of any association with more objective measures of memory and executive function by formal neuropsychological testing.
There were occupation-related risk factors for moderate to severe musculoskeletal symptoms identified in the study. Musculoskeletal symptoms were more likely in divers with experience in police and oilfield diving and in divers who had stopped diving. Police officers are exposed to a number of work-related stress factors  that might be expected to increase any tendency to somatise , and UK oilfield divers are known to have a greater tendency to express symptoms than other divers without any overt indication of any underlying illness . Retiral from diving in this group of UK divers indicates a change in job status rather than a complete retiral from work, and retraining, a drop in income or a move to a less secure job may be factors underlying increased somatisation for this group.
Not only was general symptom reporting associated with musculoskeletal symptoms, it was also associated with a positive physical examination for divers. This effect was only seen, however, where the physical abnormality included pain or tenderness. This might be expected since pain identified on physical examination is yet another symptom which would be commoner in a group which had a tendency more readily to report symptoms of any kind. This concept is supported to some degree by the observation that the reduction in SF36 PCS associated with a positive physical examination was less significant in divers than in offshore workers, indicating that the abnormalities identified were not physically as important. The results of physical examination might be considered to offer an objective measure in the assessment of musculoskeletal symptoms. It is clear, however, that this may not be the case and that any subjective response from the persons examined needs to be considered alongside an assessment of their threshold for expressing symptoms of any kind. This may well be true for other modalities of physical examination. Neurological examination, in particular, requires a high degree of subjective response from the patient, and the effect identified here might underlie the high prevalence of unexplained minor physical findings on neurological examination in a sample of Norwegian divers . This Norwegian study used observers who were not blinded to subject identity and medical history, and any tendency for the participant to influence the outcome of a physical examination would have been greater than in the present study.
Clearly, divers are more likely to experience symptoms than the control group, and positive findings on physical examination were related to this tendency. There was no indication, however, that symptom reporting necessarily indicated any significant medical problem. This study did not examine whether participants perceived that symptoms were related to any occupational factors, but in the UK, there is no strong perception that problem-free diving is injurious to health. The same is not true in Norway. Out of a population of 375 Norwegian divers working in the offshore oil and gas industry prior to 1990, at least 104 have been referred to a specialist hospital-based unit for the investigation of perceived diving-related problems . Unsurprisingly, health-related quality of life scores were below population norms, but there was no indication of any association with observable current abnormality or disease. This population also reported a high prevalence of neurological and musculoskeletal symptoms and had a higher than expected number of people on disability benefit in a Government-sponsored commission report . The release of this report was followed by the launching of a generous compensation scheme for oilfield divers with diving-related injury and medico-legal action against the Norwegian Government by a number of divers. It may be that some divers have a tendency to somatise and that when certain psychosocial pressures are applied, this converts to overt illness. Accordingly, UK divers with a tendency to somatise may be at the same risk, and any exercise, such as unnecessary screening, may further increase this risk by generating increased health-related anxiety.
This study may be considered to have weakness and strengths. The cross-sectional design does not allow any attribution of cause. A main purpose of the study, however, was to use physical examination to clarify the degree to which symptom reporting reflected physically identifiable disorder in the groups studied. Although the groups studied were randomly sampled from the background population, there may have been a degree of responder bias since we identified anxiety as a possible basis of symptomatic complaint and this is known to be associated with earlier presentation of diseases such as cancer . From this, it might be expected that anxious people would be more likely to attend for examination and that they would, accordingly, be over-represented. Random sampling with a control group from an equivalent industry, however, allowed for such responder bias. Somatisation was not assessed using one of the standard questionnaires for this purpose, and this might be seen as a shortcoming. The standard instruments, however, would have been inappropriate for this study’s purpose since joint or limb pain symptoms are included in all the most commonly used questionnaires . In fact, the scoring system used in the study covers all the symptoms generally elicited in somatisation questionnaires and allowed their assessment in the context of a standard medical examination, thus avoiding unnecessary repetition. The general symptom score used did not permit any putative diagnosis of somatoform disorder since it has not been validated. This was not one of the study aims, however, since our initial questionnaire study did not indicate a degree of symptom reporting that could be taken to indicate a somatoform disorder .
Medical examination does not provide a ‘gold standard’ for the identification of physical issues since abnormal physical signs are not necessarily associated with symptoms of pain or loss of function at the time it is conducted and, as detected here, it can be influenced by the subjective responses of the person examined. We wished, however, to study the basis of a range of musculoskeletal complaints compatible with a wide aetiology but predominantly manifesting as pain or discomfort, and the choice of a standardised, objective examination to provide this is justifiable . A blinded physical examination technique minimised participant-doctor interaction as a source of bias in the demonstration of physical signs, and the use of doctors from outside the study team avoided the impact of any researcher preconceptions. In spite of these precautions, it was clear that participants with a tendency to express symptoms could influence the outcome of the examination.
In summary, questionnaire data reflected the prevalence of physically identifiable musculoskeletal abnormalities in divers and non-divers, but this relationship was not strong enough to offer a basis for future health screening. Musculoskeletal symptom reporting was also linked to a general tendency to express symptoms of any kind and to have less favourable scores on questionnaires for mood, memory and executive function. Scores, however, did not depart from the normal range, and there was no underlying abnormality of tests of neuropsychological function. It is suggested that high scoring on symptom questionnaires is an expression of a tendency to report symptoms of any kind or somatisation rather than any indication of abnormality in the modality addressed by the questionnaire. Multiple symptom reporting was also related to a positive physical examination when the abnormality detected included pain or tenderness. It is suggested that multiple symptom reporting may indicate a tendency to respond positively to subjective elements of a physical examination and that this effect may underlie the high prevalence of minor physical abnormalities detected in divers in other studies.