A population based study of the prevalence of pain in Iceland

2.1 Sample

A random sample of 1286 Icelandic adults, 18 years old and older, was drawn from the Icelandic National Registry, which holds information about names, birth-dates, and addresses of all residents of Iceland. The residents of Iceland numbered 288,201 as of December 1, 2002 [29] a month before data collection started. Of those invited 599 (46.6%) completed the survey.

2.2 Instruments

2.3 Pain severity and pain interference with life activities

The Icelandic version of the Brief Pain Inventory (BPI) [30,31] was used to assess prevalence, severity, and interference of pain. It contains items to evaluate presence and location of pain and four items assess pain severity (worst pain, least pain, average pain, and pain now). Response options range from 0 “no pain” to 10 “pain as bad as I can imagine”, with a single composite severity score (Pain Severity) created by taking the mean of the four severity items for analysis. One open-ended question obtains information about treatments or medications used to treat pain. One item addresses how much pain relief the individual has received, with response options ranging from 0% (no relief) to 100% (complete relief). Finally, seven items evaluate pain interference with life activities. Participants report how any type of pain interfered with various aspects of their lives with options from 0 “Does not interfere” to 10 “Completely interferes”. A single interference score (Pain Interference) was created by taking the mean of the seven interference items. Pain and pain interference are usually assessed with a 24-h timeframe, but in the present study the timeframe was changed from 24-h to “the past week”. Open-ended questions were added to assess how long participants had experienced their pain, and what participants believed caused their pain.

The BPI has been used extensively and found to be both reliable and valid in several languages [e.g. 32,33,34,35,36,37]. Psychometric properties of the Icelandic BPI have been assessed in a convenience sample of 244 individuals and found to be both reliable and valid [38]. In the present study the internal consistency of the severity scale was α = 0.86 and α = 0.78 for the interference scale. Only persons who had experienced pain in the past week were asked to fill out the BPI.

2.4 Mood

Mood was assessed with the Positive and Negative Affect Scale (PANAS) [39] a 20-item measure consisting of two subscales, positive and negative affect. Subjects rate each of the affect words with respect to how they felt during the past week, on a scale of 1 to 5 with anchors of “very slightly”, “a little”, “moderately”, “quite a bit”, and “extremely”. Mean scores for each subscale are used in analyses. The measure has reported reliability (internal consistency) and construct validity [39]. The PANAS was translated into Icelandic by the primary investigator for this study and back-translated into English by another bilingual person. These steps were repeated until agreement was reached. The instrument was pilot tested with 10 Icelandic adults who provided feedback regarding understand-ability of the items. In response to their feedback a new translation was generated for 5 items and this version was then completed by 5 additional Icelandic speaking adults. A new translation was sought for two items in response to their feedback. In this study the internal consistency of the positive scale was α = 0.81 and the negative scale α = 0.88. The correlation between the two scales was r = −0.12. This low correlation supports that the two scales measure two separate constructs.

2.5 Quality of life

Global quality of life was measured with a single item stating “I am content with the quality of my life right now”. This item is based on a single item from the Functional Assessment of Cancer Therapy (FACT) scales [40]. The response options in this study, were (1) “Not at all”, (2)“A little bit, (3)“Quite a bit”, and (4)“Very much”. Single item measures for QOL have been used with good results [41] and were chosen for this study instead of lengthy measures to limit response burden.

2.6 Perceived personal health

Personal health was evaluated with a single item generated by the research team. “How is your current health?” Response options were “Very good”, “Good”, “Neither good or poor” “Poor” and “Very poor”.

2.7 Procedure

After obtaining permission from appropriate Institutional Review Boards, specialists at the Icelandic National Registry generated a simple random sample of registered habitants of Iceland, 18 years or older, as of December 31, 2002. Data collection was conducted with a mailed survey from the beginning of January through mid March 2003. The questionnaire, cover letters, and mailing schedule were based on the Tailored Design Method [42].

3.1 Sample

The mean (SD) age of the respondents and non-respondents was 44.94 (17.12) and 44.25 (19.44) respectively which is not a significant difference. A majority of respondents were female (56%). Of the 654 women that were invited to participate, 333 returned the questionnaire (51%) while 264 of 618 men (43%) did so (data on gender were not available for two participants). The mean (SD) length of education for respondents was 13.81 (4.27) years. The majority of participants were married, or cohabiting (67.4%) and 68.6% worked either fullor part time outside of the home. 63.8% of participants reported living in Reykjavik and surrounding neighborhoods, 28.1% lived in villages, and 7.1% lived in rural areas. These figures accurately represent the population of Iceland at the time of data collection, with 62% living in Reykjavik and the surrounding neighborhoods, 31.6% living in urban areas throughout the country, and 6% living in rural areas [29]. Unfortunately, data about income were too incomplete to use for any analysis. For further information on demographics see Table 1.

3.2 Prevalence of pain

Participants were asked whether they had experienced pain in the past week and 575 participants of 599 answered this question (96%). Of those, 232 (40.3%) had experienced pain in the past week. Of the 232 individuals who reported pain in the past week, 183 (78.9%) indicated that they had experienced this pain for more than three months or 30.6% of the 599 participants. Those who had experienced pain for more than three months did not differ in terms of demographic characteristics, perceived health or QOL from those who had experience pain for less than three months.

Those who had experienced pain in the past week (n = 232) had experienced pain for a length of time ranging from one day to 50 years. The median length of their pain experience was 1095 days (3 years). Their worst pain in the past week had a mean (SD) of 4.85 (2.17). Their least pain in the past week had a mean (SD) of 1.92 (1.86) and their average pain had a mean (SD) of 3.35 (1.93). Participants’ mean (SD) pain now was 2.70 (2.27). Participants’ Pain Severity composite in the past week had a mean (SD) of 3.21 (1.73). Participants’ Pain Interference with life activities had a mean (SD) of 2.59 (1.98).

3.3 Relationships between demographic variables, pain, mood and QOL

Those who had experienced pain in the past week were compared to those who had not in terms of demographic variables. Comparisons were made with a Logistic regression and based on age, gender, education, marital status, habitation, personal health, QOL, and whether participants had been diagnosed with a chronic disease or not. The only significant difference between those who had experienced pain and those who had not was that those who had experienced pain rated their personal health worse [B = −0.91, SE = 0.15, Wald = 38.75, p = 0.00].

The item assessing perceived health was completed by 540 participants. Most participants indicated that their health was “Very good” (42%), or “Good” (39%), while a few indicated that it was “Neither good or poor” (16%) or “Poor” (4%) and none indicated that is was “Very poor”. The mean (SD) score was 1.82 (0.84). For those who had been in pain (n = 232) the findings differed in that, 22.8% said their health was “Very good”, 44% said it was “Good”, for 59% it was “Neither good or poor” and for 18% it was “Poor”. None of the participants who had been in pain said their health was “Very poor”.

The Positive and Negative Affect Scale (PANAS) was completed by 549 participants. The mean (SD) on the positive affect scale was 3.27 (0.66), with an observed range of 1-5 (possible range of 1-5 with higher scores indicating more positive mood) for the entire sample, and 3.21 (0.69) for those who had experienced pain in the past week. The mean (SD) score on the negative affect scale was 1.66 (0.67) with an observed range of 1.0-4.7 (possible range 1-5 with higher scores indicating more negative mood) for the entire sample and 1.74 (0.70) for those in pain. Scores on neither the positive nor negative affect scales differed for those who had pain and those who did not.

The global quality of life (QOL) item was completed by 590 participants. The mean (SD) score was 3.2 (0.78) for the total sample, with an observed range of 1-4 (possible range 1-4 with higher scores indicating better quality of life), and 3.06 (0.82) for those in pain. This was not a significant difference.

3.4 Pain variables as predictors of pain interference with life

The following hypothesis were tested; when controlling for gender, age and education (a) pain severity predicts pain related interference with life activities; (b) pain severity predicts positive and negative affect and QOL; and (c) pain related interference with life activities mediates the effects of pain severity on positive and negative affect and quality of life (Fig. 1).

When examining the effects of pain severity on pain interference with life activities we used the pain severity composite score for analysis. Since pain has been reported to differ based on age, gender and education we controlled for these variables in the analysis [e.g. 9,11,19]. Pain Severity significantly predicted Pain Interference with life [B = 0.71; F = 126.14; df = 1,206; p = 0.00] (Fig. 1).

The test of mediation was based on first testing the relationship between the predictor (Pain Severity) and the potential mediator (Pain Interference with Life) as described previously. The second test is of the path between the potential mediator and the dependent variable (mood and QOL) with all variables at an equal level or to the left partialled out. If these two relationships are significant, it will be concluded that Pain Interference with Life is a mediating variable; it is not considered necessary to test the direct path between the independent (Pain Severity) and dependent variables (mood and QOL), and in fact, this path is not considered to be of importance if a significant mediation effect is found [43].

We tested the mediation effect of pain interference between pain severity, mood and QOL. All three mediation effects were significant. Between Pain Interference with Life and Positive Affect [B = −0.06; F = 4.53; df = 1,196; p = 0.04], between Pain Interference and Negative Affect [B = 0.15; F = 23.21; df = 1,196; p = 0.00], and between Pain Interference and Global Quality of Life [B = −0.18; F = 29.11; df = 1,196; p = 0.00] (Fig. 1).

Since the mediation effect of Pain Interference was established it is not necessary to test the direct paths from Pain Severity to the outcome variables as described above. However, these tests are presented here to provide further information. As before all analysis controlled for the effects of age, gender and education. The direct path between Pain Severity and Positive affect was not significant [B = −0.03; F = 1.17; df = 1,197; p = 0.28], but the direct paths from Pain Severity to Negative affect [B = 0.08; F = 7.00; df = 1,197; p = 0.01] and QOL [B = −0.10; F = 8.30; df = 1,197; p = 0.00] were.

3.5 Causes of pain

A content analysis was performed on an open ended question asking participants to specify what they believed to be the cause of their pain. Two researchers conducted the analysis. First, a scoring sheet was generated independently from the data by each researcher, and then a single scoring sheet was generated from the two. The data were then analyzed separately by the two researchers and the analysis compared for consistency. A high degree of agreement was among the two researchers. Of those who had pain, 204 answered the question. Some participants provided multiple answers to those questions. In response to what the participants believed to have caused their pain the most frequently cited causes for pain were strain injuries resulting from work or sports activities (n = 79), arthritis (n = 39), mechanical problems (e.g. due to birth defects, curvature, slipped discs, etc.) (n = 37), disease (n = 31) and accidents (n = 30). A fair number (n = 19) cited that they did not know what caused their pain (Table 2). Many of the participants cited numerous causes for their pain and a few did not specify a cause for their pain but instead wrote in their type of pain (“back ache”, “headache”, etc.).

3.6 Methods used to treat pain

When asked whether participants had sought medical attention for their pain, all 232 individuals who reported having pain in the past week answered this question. Of those, 179 reported having sought medical attention (77.5%). Those who had used medication or other methods to treat their pain in the past week reported an average (SD) pain relief of 43.0% (34.7%).

A content analysis was conducted of responses to the question asking participants what treatments they received for their pain for which 200 participants provided an answer. The same approach was used to the analysis as described above. A majority indicated that they had used medications (n = 137), and most had used analgesic medications (n = 122). A fair number, 47 had been in physical therapy and/or massage, 40 had used exercise, 31 had been in physical therapy, 19 had used heat or cold and 21 had not used anything (Table 3). Many had used a combination of treatments. Most had only used medications alone (n = 76), 61 had used mediations in combination with other treatments, and 42 had used other methods without medications.