The study demonstrated a relationship between higher physical activity and less reported MSD. The significant associations were found between physical activity and reported MSD in all parts of the body, except elbows, knees and feet when adjusted for employment status, age, gender, BMI, smoking, education and physical stressors. Light physical activity strengthened the inverse association between physical activity and MSD.
A comparable study among aluminium workers in Norway also found a relationship between higher physical activity and less reported MSD without differentiating between light and heavy activity [40
]. However, this study was limited to physical activity during leisure. The finding that light physical activity was associated with less MSD, is supported by studies on other health benefits [11
]. A volume of exercise half the amount of physical activity that is currently recommended [13
] may be of importance for several health outcomes. Anyway, since the majority of our study population performed heavy physical activity, the impact of light physical activity should be studied in a population performing predominantly light activity.
The fact that the significant association between higher physical activity and less MSD was present when adjusted for the physical stressors index supports the independent importance of physical activity versus physical stressors. A follow-up study by Leino-Arjas [41
] found that high physical strenuousness at work increased the risk of later poor functioning, whereas physical activity at leisure was protective. Physical strenuousness could be compared with heavy lifting and working with the hands above the shoulders. The opposite effects of physical activity and physical strenuousness on functioning might be similar for the development of MSD, as our study indicates.
The lack of positive effect of physical activity on MSD in the knees and feet is supported by studies that demonstrate physical training as the cause of injuries in the lower limbs [27
]. Athletic sports may increase the possibility of some lower-extremity disorders, and this might counteract the expected effect of physical activity on musculoskeletal health. A study on Australian Defence Force members found a high frequency of musculoskeletal problems, especially ankle, knee and spinal strains, among military personnel compared with a population in general practice [26
]. Other studies have found high rates of military training-related injuries at or below the knee [28
]. A literature study found some evidence that at least 4 hours per day of heavy physical activity increased the risk of osteoarthritis of the knees and running at least 20 miles (32 km) per week increased the risk of hip or knee osteoarthritis [17
In the Navy, civilians had a higher prevalence of MSD than military personnel. An association between being a civilian and having MSD persisted for the neck and lower back when adjusted for age, gender, physical activity, BMI, smoking, education and physical stressors. The difference between civilians and military personnel might be explained by the selection of military personnel due to requirements for fitness for duty. The requirement of fitness among military personnel probably leads to a selection of healthier persons to the military group. Work factors not included in the analysis, such as psychosocial factors [5
], might also differ among the two groups. This relationship should be studied further.
The one-year prevalence of 32% of frequent MSD in one or more parts of the body among workers in the Navy is rather low compared to other working populations [40
]. In a study among aluminium workers in Norway using the same questionnaire on MSD, the comparable number was 49% [40
]. The working populations both in the Navy and in the aluminium industry were dominated by men and had the same age. One explanation for the low prevalence of MSD in the Navy might be less physical risk factors in the work environment and more physical activity or exercise at work. Due to the demand of fitness test for military workers, this occupational group might differ from other working groups and these populations cannot be compared directly.
The cross-sectional study design does not allow us to draw conclusions on a causal relationship between physical activity and MSD. It is not clear whether physical activity influences the MSD or, conversely, the MSD influence the amount of physical activity. On the one hand physical activity might result in less MSD, and on the other hand MSD might result in less physical activity. However, the study population consists of people at work, and people with severe and chronic pain unable to perform physical activity are probably very few. In any case, a hypothesis that physical activity prevents MSD should be followed up by prospective studies.
We are faced with selection bias at two levels. Firstly, the employees in the army might differ from other workers due to health certificate requirements (the healthy worker effect). The healthy worker effect has been shown to significantly influence results of cross-sectional studies dealing with low back pain [44
]. Secondly, a selection bias due to the low response of 58% may be present if the non-responders had a different prevalence of MSD and a different degree of physical activity compared with those who responded. Similarly a different association between physical activity and MSD for the responders and non-responders could create such bias. However, the similar mean age and percentage of women and men among the respondents compared to the total Navy population may indicate that the respondents were representative for the Navy population. In addition, since the survey was part of a general health survey among employees and not specifically about MSD and physical activity we do not believe that the selection bias was a major issue.
Physical activity was measured by a questionnaire, and the validity and reliability of this method may be questionable [45
]. However, a questionnaire may be the only feasible method of assessing physical activity in large populations. We asked about physical activity during work, and this might include physical activities that both benefit and harm musculoskeletal health. Anyway, physical stressors, like twisted positions and heavy lifting, was associated with more MSD, but higher physical activity was associated with less MSD, also when adjusted for the physical stressors index. More precise information on the type of activity could improve the validity for both the physical activity questions and the questions about physical stressors. Recall bias and social desirability bias can lead to misclassification of the amount of activity.