Studying the association between musculoskeletal disorders, quality of life and mental health. A primary care pilot study in rural Crete, Greece

Setting and sample

Data was collected from one rural Primary Care Centre (PCC) on the island of Crete. This PCC covers a population of approximately 10,000 inhabitants, from a geographically defined area and is staffed by GPs, nurses, midwives, physiotherapists and laboratory technicians who provide primary and emergency care services around the clock. Over a period of 10 working days, all consecutive patients in the waiting room, aged 20-75 years, were eligible to participate in this study. This sample size was sufficient to show statistical significance at 80% for subjects with MSD (a quality of life measure was associated with the presence of MSD if responses to this measure were 50% worse than for subjects without the presence of MSD). The Greek version of the standardised Nordic questionnaire for the analysis of musculoskeletal symptoms (general form) also known as the Nordic Musculoskeletal Questionnaire (NMQ) was used to identify people with musculoskeletal problems [10]. The NMQ is a self-administered questionnaire, designed for the purpose of screening for MSD in epidemiological studies and was translated and validated into the Greek language [11]. The use of NMQ to measure MSD prevalence in a primary care population in Crete has been reported elsewhere [9]. See additional file 1.

Instruments

Greek versions of the medical outcomes study Short Form 36 measures of health status (SF-36) [12] and the General Health Questionnaire (GHQ-28) [13] were applied to investigate HRQL and mental distress. The SF-36 consists of 8 domains measuring physical functioning, role limitation due to physical health problems, bodily pain, general health, vitality, social functioning, role limitation due to emotional health problems and mental health. Scores are transformed from a range of 0 to 100, with the higher score indicating better HRQL for all domains, except 'bodily pain' where a lower score accounts for less pain and increased quality of life. Physical functioning, role limitation due to physical health problems and bodily pain correlate mainly to physical dimensions, whereas role limitation due to emotional health problems and mental health mostly to mental dimensions of health status. General health, vitality and social functioning correlate to both dimensions. SF-36 has been applied in general population surveys in many countries and has been used for general or specific MSD [14, 15]. SF-36 has been validated in the Greek language [16].

The GHQ-28 is a 28-item measure of emotional distress that is divided into four sub-scales: somatic symptoms, anxiety/insomnia, severe depression and social dysfunction. For statistical analysis the binary method of scoring the questionnaire was used. A score of 5 or more in GHQ-28 indicated mental distress. GHQ-28 has also been used to assess psychological distress in surveys of musculoskeletal disorders [17] and has also been validated in the Greek language [18].

Data on height and weight was self-reported via the NMQ and the body mass index (BMI) was calculated. Additional information concerning socio-demographic characteristics (age, gender, cohabitation, education, occupation, clinical co-morbidity and consultations to PCC -to GPs, nurses or physiotherapists) was gathered by reviewing medical records available at the PCC.

Bioethical committee

At first this study was approved by the Postgraduate Studies Committee of the Medical School of the University of Crete, as a part of a PhD thesis. Then it gained the approval of the Scientific Board of the Regional University Hospital of Heraklion, which serves as a bioethical committee. All the participants were fully informed, through a personal letter of agreement, about the purpose of the study and gave their written consent before the completion of the questionnaires.

Analysis of data

Cohabitation status was placed into 2 categories (cohabitation/no cohabitation). Educational level and employment status were categorized according to the International Standard Classification [International Standard Classification of Occupations, 88 (ISCO-COM 6) and (International Standard Classification of Education 1997, ISCED 0+1]. Multiple linear regression analyses with backward selection were conducted using physical and mental dimensions of SF-36 as dependent variables and musculoskeletal symptoms, age, gender, education, occupational status and clinical co-morbidity as independent variables. Since both SF-36 and GHQ-28 refer to health indicators during the last few weeks, the estimated 7 days prevalence of musculoskeletal symptoms were used. Regression analysis was performed with GHQ-28 scores as dependent variables to test the possibility of predicting the mental health status from the occurrence of musculoskeletal symptoms. In order to estimate the adjusted odds ratios of the consultations to PCC in relation to scores of SF-36 and GHQ-28 multiple logistic regressions were assessed. Continuous variables are expressed as means (SD). In this analysis the prevalence of pain during the previous 12-months period was used. All analyses were performed using SPSS version 16.0. An a = 0.05 level was set as significant.

Participation rate

Participation rate was high (91%), 176 patients (55% females, mean age 54.5 years) agreed to complete the questionnaires. 126 (71.6%) reported at least one musculoskeletal problem during the previous 12-month period (responses to the NMQ questions) with low back pain being the most frequent (n = 76, 43.2%), followed by knee (n = 55, 31.3%), shoulder (n = 53, n = 31.1%) and neck problems (n = 46, 26.1%). Less than half of those who reported MSD (n = 55, 31.3%) attributed to them restrictions in daily activities and 42% reported pain during the previous 7-day period (point prevalence). Women and the elder tended to report more symptoms for every pain site (p < 0.05).

The impact on quality of life

Musculoskeletal symptoms were generally associated with worsened HRQL. Subjects reporting neck pain over the previous 7-day period, had significantly lower scores in SF-36, particularly for physical functioning (SF-36 score, 42.9 versus 81.0, p < 0.0001), role limitation due to emotional problems (45.2 v 77.9, p < 0.0001), bodily pain (52.9 v 24.4, p < 0.001), general health (56.4 v 48.7, p = 0.023), vitality (47.5 v 62.5, p = 0.002) and role limitation due to physical problems (32.1 v 86.7, p < 0.001). Scores relating to other musculoskeletal symptoms are presented in Figure 1 (t-test analysis). Wrist pain did not show a significant effect on any dimension of SF-36. The social functioning, vitality and general health domains demonstrated the least association with MSDs and the mental health domain was affected only by knee pain. Impaired HRQL was particularly evident for the physical functioning, role limitations due to physical health problems and bodily pain domains in patients reporting any musculoskeletal problem comparing with those (n = 50) reporting no musculoskeletal problems at all. In general, HRQL for subjects with coexisting MSD were worse than those with only one disorder. Reporting of more than four musculoskeletal symptoms significantly deteriorated all SF-36 dimensions, except for vitality, social functioning and mental health.

Among subjects reporting musculoskeletal symptoms, age was found to influence physical functioning, role limitations due to physical problems and bodily pain. Obesity and lower education were both related to lower scores in physical functioning. Women reported worst HRQL than men as they scored lower in most SF-36 dimensions, except for vitality, social functioning and mental health. Occupation did not show any significant correlations in the study population. The presence of clinical co-morbidities did not influence the SF-36 scores.

The impact on mental health

According to the analysis of GHQ-28 scores, MSD patients were more likely than non-MSD patients to present symptoms of mental distress for every pain reported for the time period of the last week. In multivariate analysis, mental distress, as measured through GHQ-28, added a negative effect in HRQL dimensions of SF-36, except for general health (Table 1). Moreover, patients with mental distress (those who scored positive in GHQ-28) were more likely to be men (Beta = -1.25, 95% C.I. 0.12-0.65, p = 0.003) who suffer from neck (Beta = 1.92, 95% C.I. 1.21-38.40, p = 0.03) or shoulder pain (Beta = 1.18, 95% C.I. 1.40-7.47, p = 0.006), according to the multiple regression analysis.

The impact on seeking care patterns

Main findings

Implications of the study

The present findings may have implications both for public health planning and for primary care settings. Physical disability and depression are both predisposing factors for primary care consultations. General practitioners should consider screening for psychological symptoms in all patients with MSD symptoms. Apart from pain and disability management, patients with MSD also need psychological support. GPs need to be adequately trained to deal in a sensitive manner with these patients and their families.

Strengths and limitations of the study

The small size of the sample raises concerns of type II error and the inclusion of only one PCC raises issues of generalisation of other primary care settings. The fact that 15% of the sample population did not complete the SF-36 may have also introduced a potential bias response. However, non-responders did not significantly differ from responders in terms of age, gender and presence of MSD. Time needed for completion of the questionnaires and place (waiting room in the PCC) was the main issues for non-compliance, according to their reports. Household surveys or pre-arranged telephone interview could overcome this problem.

Although GHQ is not a diagnostic instrument, it can be used in the first stage to identify potential cases of depression, which must in turn be confirmed using clinical assessment [30]. This cross-sectional study reports on the impact of MSD on mental health as measured by a screening scale. Thus, we are unable to comment on whether the identified burden on the affected quality of life reflects the mental health problems experienced or other co-morbidities and conditions which may play a role. Our study cannot explain if mental disorders are a cause or a consequence of MSD factors and further studies are needed.

The fact that the consultation rates were estimated through the medical records may also introduce an information bias, since the credibility of the records was not previously checked. However, the current study contributes to the literature with regard to MSD, mental health and HRQL, by focusing on the relationship between these issues in a primary care setting in a rural location.