The association between cold exposure and musculoskeletal disorders: a prospective population-based study

Main findings

High occupational ambient cold exposure was associated with NSP, LBP, and radiating LBP. When stratifying on gender, associations were observed between occupational ambient cold exposure and NSP among men, and LBP and radiating LBP among women.

Similar associations between cold exposure and long lasting (> 3 months) NSP (OR 1.46; 95% CI 1.13–1.89) was seen in a study on a general working population in northern Norway (Farbu et al. 2019), adjusting for age, gender, smoking, educational level, physical activity level, and insomnia, while they did not find any associations with LBP (OR 1.18, 95% CI 0.91–1.52). A longitudinal study based on the same sample found that working in a cold environment at baseline was associated with MSDs lasting three months or more, 7–8 years later (incidence rate ratio 1.15; 95% CI 1.03–1.29) (Farbu et al. 2021a, b). However, anatomical location was not specified. A Swedish study compared construction workers from regions with different climate in Sweden and found that working in the coldest region was associated with NSP (OR 1.57: 95% CI 1.47–1.67) as well as LBP (OR 1.19: 95% CI 1.14–1.24) (Burström et al. 2013). In a study on Russian mine workers, working in cold conditions (< 10 °C) was associated with reporting LBP (OR 1.82; 95% CI 1.55–2.15), after adjusting for gender, BMI, duration of work, physical fitness level, and stress (Skandfer et al. 2014).

Furthermore, working in cold indoor environments has also been associated with MSDs. A study on Israeli male cold store workers, exposed to indoor temperatures around − 20 °C, had an increased risk of back pain (OR 2.9; 95% CI 1.3–6.7) compared to their unexposed colleagues who performed similar tasks (Dovrat and Katz-Leurer 2007). A study on Pakistani workers compared those working in cold storage facilities with indoor temperatures ranging from − 20 to − 30 ºC with those who did not have any exposure to cold indoor environments and found a 15-times increased risk of repeated NSP or upper extremities (RR 15.00; 95% CI 6.33–35.51) (Ghani et al. 2020). There were, however, differences in physical load between the cold-exposed and unexposed groups that could explain part of the increased risk. A Norwegian study on people working in the seafood industry found strong associations for the subjective feeling of being cold and LBP (OR 11.0; 95% CI 4.5–26.8), and NSP (10.5; 95% CI 3.1–35.3) (Aasmoe et al. 2008). Subjectively feeling cold was associated with both NSP and LBP in a study of Finnish meat processing industries (Sormunen et al. 2009a, b). Those who experienced “extensive low back cooling” had the highest risk of LBP (OR 3.88; 95% CI 1.82–8.25) and those who experienced “extensive neck–shoulder cooling” had an increased risk of NSP (OR 6.47; 95% CI 2.79–14.99) (Sormunen et al. 2009a, b).

The prevalence of NSP at baseline was 15.8% in the whole population (women 20.1%; men 10.6%), which is lower than the official Swedish statistics, reporting that 45% of the women and 25% of the men have pain in the neck and the upper back (The Swedish Work Environment Authority 2020). This difference is probably due to the cutoff for defining pain used in the present study with only the highest category (four-grade scale, as “none”, “insignificant”, “somewhat”, or “a lot”). When changing the cutoff and including the “somewhat” category the baseline prevalence was 40.2%, closely resembling the national statistics. The reason for using the higher cutoff was to ascertain that the case group included individuals with more severe symptoms. Another explanation to the difference in prevalence rates could be that we used a more defined anatomic location in our survey, i.e., not including the upper part of the back. Fejer et al. (2006) reported a point prevalence of neck pain more in line with ours, ranging from 5.9% to 22.2% in the adult population (15–74 years), while Safiri et al. (2020) reported a point prevalence in western Europe as low as 4.6%. However, in the latter study, no age restrictions were applied, nor criteria on being currently working.

The prevalence of LBP at baseline was 13.4% (women 16.2%; men 10.1%), which is in line with Ihlebaek et al. (2006) reporting a point prevalence of 13.4% (women 16.8%; men 9.9%) in southern Norway and slightly lower than the 18.2% (women 20.4%; men 14.6%) in southern Sweden. In addition, a systematic review of global prevalence of LBP presented a mean point prevalence of 11.9% (Hoy et al. 2012), which is well in line with our results.

The prevalence of radiating LBP at baseline was 5.9% (women 5.9%; men 5.8%) which is in the same range as have been reported earlier (Berry et al. 2019), but substantially higher than what was reported by Younes et al. (2006) who studied a Tunisian urban population with a point prevalence of 0.75% and the review by Konstantinou and Dunn (2008) that reported a point prevalence of 1.6%. The difference in prevalence of radiating LBP is believed to be explained by differences in sampling and case definitions (i.e., demanding radicular pain below the hip or knee).

The annual incident proportion of NSP in our study was 1.5% (women 1.8%; men 1.1%), for LBP 1.5% (women 1.8%, men 1.3%), and radiating LBP 0.6% (women 0.7%, men 0.5%), highlighting the magnitude of the problem.

Gender

We found gender differences in NSP, LBP, and radiating LBP prevalence at baseline and follow-up as well as regarding incidence proportion. There were also gender differences in occupational ambient cold exposure with a smaller proportion of women being highly exposed (9.4% of women vs. 17.7% of men in NRS 8–10). Similar gender differences in prevalence of both NSP and LBP in the Swedish population have been shown by others (Bingefors and Isacson 2004; Wahlstedt et al. 2010) and several papers have reported a higher prevalence of musculoskeletal pain in general in women compared to men (Bingefors and Isacson 2004; Treaster and Burr 2004; Fillingim et al. 2009; Leboeuf-Yde et al. 2009; Leijon et al. 2009).

The gender stratified analysis showed that there were differences in the associations between ambient occupational cold exposure and the different pain outcomes for men and for women. For men, a significant association was found only between cold exposure and NSP, while there were significant associations between cold exposure and LBP and radiating LBP for women. Since the prevalence of NSP was substantially higher among women than men, the rather small added proportion related to occupational cold exposure might not have been discernable. Furthermore, we adjusted for physical exposures at work with a simple JEM, and it cannot be ruled out that there were gender differences that a more detailed physical exposure assessment could have revealed. Finally, due to low numbers of highly cold exposed females, associations with our outcomes for women could be hard to detect in our cohort.

The mechanisms for gender differences in the susceptibility for cold-induced discomfort are not clear. Sormunen et al. (2009a, b) reported that female workers experienced cold ambient temperature, and other environmental factors as significantly more harmful than their male counterparts. Pienimäki et al. (2014) pointed out that women may have a lower temperature threshold for reporting symptoms. In contrast, in an experimental study, working in cold, compared with thermoneutral conditions, increased muscular activity in the forearm and upper arm extensors only in men and not in women (Sormunen et al. 2009a, b).

Personal protective equipment (PPE), such as heavy caps, coats, and gloves can be used to protect workers from ambient cold exposure. However, it may alter working posture and hinder movements, thus increasing the physical workload (Piedrahita et al. 2004; Dovrat and Katz-Leurer 2007). In addition, gender differences in access to, as well as usage of PPE cannot be ruled out.

To conclude, further studies are needed to elucidate mechanisms behind gender differences regarding effects of cold work.

Mechanisms

The mechanism for cold-induced MSDs is not established, but different explanations have been suggested. An increase in muscle activation as well as a reduction in muscle activation gaps, due to exposure to moderately cold conditions in several upper extremity muscles during repetitive work in ambient temperatures of 4–10 °C has been shown (Oksa 2002; Oksa et al. 2002, 2006, 2012; Piedrahita et al. 2008; Sormunen et al. 2009a, b; Renberg et al. 2020a, b). Cold exposure has not been shown to have any effect on muscle activation in isometric muscle work. Some studies have even showed a beneficial effect, where the endurance time was increased and the rate of fatigue slower when muscle temperature was below normal but higher than 27 °C (Oksa 2002; Renberg et al. 2020a, b). On the other hand, working in cold environments may reduce the temperature of the working muscle tissue and slow nerve conduction velocity which in turn could be seen as a shift to lower frequencies in the frequency component of EMG (Petrofsky and Laymon 2005).

Cold-induced vasospasm has been found in 20% of patients with chronic LBP and in 38% of patients with fibromyalgia but in only 8% of healthy people (Lapossy et al. 1994). Thus, another plausible mechanism could be that reduced muscular blood flow induces ischemic nociceptive pain during situations with high physical demands. In support of this view, a significantly lower blood flow was seen in a laboratory study, during wrist flexion–extension repetitive work in two cold conditions (5 °C) compared to during a thermoneutral condition (25 °C) (Oksa et al. 2002). Finally, a study found that workers with chronic pain reported more indoor climate complaints than pain-free controls despite similar actual indoor climate and concluded that the difference was likely due to central sensitization (Sundstrup et al. 2015).

For lumbar disc disease, it has been suggested that cooling is unfavorable for the diffusion of the intervertebral disc fluid when combined with heavy work and static postures (Hildebrandt et al. 2002).

Methodological limitations

The rather low response rate may have introduced a sampling bias that was not controlled for. In addition, separating the exposure measure into four categories and stratifying for gender reduced the statistical power and increased the risk for type 2 error. The measure of cold exposure was subjectively reported, arbitrarily scaled, and could not be translated into exposure intensity or duration. In addition, the cut points for cold exposure were data-driven and not based on any a priori knowledge about physiological threshold effects. Furthermore, potential effects of leisure-time cold exposure were not investigated in this study. However, a previous cross-sectional study on the same cohort revealed no effect of exposure occurring outside of work (Stjernbrandt and Farbu 2022). In clinical practice, there is likely a rather large overlap between reporting LBP and radiating LBP. However, in our study we strived to separate these two entities, by defining them in two separate questions in the questionnaire. Separating LBP and radiating LBP based on subjective reporting of symptoms might not be motivated from a clinical standpoint and makes the interpretation of the results more challenging. However, the separation was motivated by the assumption that pathophysiological mechanisms may differ in the sense that radiating LBP is less related to cold-related effects on postural muscle activity and more associated with degenerative changes in intervertebral discs and adjacent joint and bony structures. Another limitation is the fact that the duration of pain was not investigated in our study. There might be differences in short-term and more chronic conditions regarding the effects of cold exposure. Furthermore, the JEM employed in the current study was based on the major and sub-major ISCO groups and resulted in a rather crude measure of physical workload. A more detailed JEM would have improved the assessment of the physical load. Finally, since the study focused on currently working subjects, there is a risk of a healthy worker effect which might have attenuated the effect sizes.

Methodological strengths

One major strength of our study is the fact that it was population-based and utilized a prospective approach. To the authors’ knowledge, only one previous study on occupational cold exposure and MSDs has been performed, and that study specified outcomes as single- or multi-site musculoskeletal complaints without any details on anatomical region (Farbu et al. 2021a, b). The present study was conducted in a subarctic and temperate climate, where manual work is common, and this was a suitable setting for a study aiming to explore effects of cold exposure on MSDs. Since both prevalence and mechanistic explanations may differ, reporting gender-stratified results was also relevant. The study collected ample data on known confounding factors to allow for adjusted regression models.

留言 (0)

沒有登入
gif