In this study, we found that close to one-third of community-living elderly men experienced neck pain during a 12-month period, but that more than 75% rated the morbidity to be none, minor or moderate. We also found that the proportion of men who rated the morbidity as severe was higher if the neck pain was accompanied by rhizopathy, motor deficit or thoracolumbar pain.
The one-year prevalence of neck pain in the adult population varied greatly between different studies, with some studies reporting prevalence as low as 16% and others as high as 75% [1, 8, 10, 13]. However, the lack of a general definition of neck pain makes the utility of results and the comparison between studies difficult. There is a need for epidemiological studies in many populations, including in different settings and age groups. Nonetheless, our 12-month neck pain prevalence of 29% was similar to the 12-month neck pain prevalence in a Greek urban population of men and women aged 15–65 years [22] and a British population of men and women aged 16–64 years [23] (one-year neck pain prevalence 29% and 34% respectively). It was also similar to a Russian rural and urban population of men and women aged 40–94 years (mean lifetime neck pain prevalence 29%) [24]. Previously reported 12-month prevalence from different Swedish populations vary; Bergman et al. found a one-year neck pain prevalence of 14.5% in men between 20 and 74 years of age [25], while Linton et al. reported a much higher one-year prevalence, 66.3% in men and women 35–45 years of age [26]. The latter study made no clear distinction between back or neck pain, which may explain the high prevalence. To facilitate comparisons between studies of neck pain, the anatomic region of the neck needs to be defined in the same way, for example as defined by Guzman et al. [27].
Reports from other populations found a lower prevalence. For example, Genebra et al. reported one-year neck pain prevalence to be 20% in a Brazilian population of men and women older than 20 years [13] and Palacios-Ceña et al. reported a one-year neck pain prevalence of 16% in a Spanish population of men and women older than 16 years [10]. Since the prevalence of musculoskeletal disorders, including neck pain, differs and mainly increases with age [5, 10, 13], all comparisons between studies must take the age distribution of participants into account. According to a systematic analysis of the Global Burden of Disease Study, the point prevalence of neck pain increases with age until a cut-off of 70–74 years, where the prevalence decreases [3]. This, together with the narrow age span in our study compared with wider age spans in the above-mentioned studies, might explain why we did not find any influence of age on prevalence or severity of neck pain. This is in contrast to other published studies [1, 2, 7, 11, 13, 16,17,18]. Differences in the age of participants may thus have contributed to the higher prevalence in our study compared to those from Brazil and Spain.
Our study only includes elderly men, while the above-mentioned studies also include women; this will also contribute to prevalence differences compared to our study. The female gender has been reported as a risk factor for neck pain in various studies, but more recent epidemiological studies question whether this is true [5, 28]. We could not find any previous study of neck pain that focuses on elderly men > 69 years. As neck pain is a multifactorial disease, various risk factors and triggers must be evaluated to better understand how to prevent, diagnose and manage patients with neck pain.
We found that our hypothesis that neck pain accompanied by symptoms of rhizopathy/motor deficit/thoracolumbar pain is associated with a higher morbidity than neck pain alone seems to be true. We were unable to find prevalence and morbidity data in these groups of patients with neck pain in other studies and consider this information clinically important. For example, our results indicate that three times as many men with both neck and thoracolumbar pain rate their morbidity as severe compared to men with neck pain without thoracolumbar pain, and detailed questions on pain localisation may be more essential than previously known in patients with neck pain. It is also of great interest to find that, among men with both neck and thoracolumbar pain who rated the morbidity as severe, 73% rated the pain severity differently in the two anatomic regions. This may also be of clinical importance when identifying subgroups in need of specific targeted preventive and curative interventions. This also highlights the necessity of categorising patients with neck pain in future studies, i.e. with only neck pain or also with accompanying symptoms such as rhizopathy/motor deficit/thoracolumbar pain, to make comparisons of the prevalence and morbidity data across studies meaningful.
The strengths of this study include the large population-based sample with a high participation rate of elderly community-living men within a narrow age span. Therefore, we are of the opinion that the data can be generalised to similar cohorts, However, we must also emphasise that our inferences cannot be directly transferred to cohorts of younger or older men, men of other ethnic backgrounds, men living in other societies or women.
Study weaknesses include the inclusion of only ambulatory elderly men without bilateral hip replacements. Further, despite the population-based study design, we cannot exclude selection bias. We speculate that unhealthy and sick men may have declined participation, but we have no indication that this proportion is different in men with and without neck pain. In addition, the majority of our cohort were of Caucasian ethnicity within a small age span, and all lived in Sweden. Therefore, we cannot state that our inferences account for men of other ethnicities, men living in other countries and in men of other ages. Furthermore, with our approach, temporal and causal relationships are not possible to establish. We can only draw conclusions regarding associations. The limitations also include the retrospective study design that yields a risk of recall bias and the subjective patient-reported morbidity. It is also possible that the sensory symptoms occur due to nerve entrapment in the arms and not the neck, and the motoric symptoms are due to other disorders than nerve root compression in the neck. As the study was done without clinical examinations or other diagnostic tools, we can only report the subjective estimates by the patients. It would thus have been an advantage to have prospectively collected information on neck pain and limitations of daily living (ADL), instead of relying on subjective reported data. It would also have been advantageous to evaluate the number and duration of each neck pain episode, frequency of associated symptoms and recurrence rate.