Abstract
The differences in quality of life and school absence were studied in one hundred adolescents from the open population who had reported chronic headache or chronic, physically unexplained, pain at other locations.
The adolescents kept a 3-week diary about their pain and completed a quality of life questionnaire. Of all chronic pain sufferers in this study, it was the adolescents with headache who showed the least frequent pain, but they reported the poorest quality of life and the largest school absence due to their pain.
Adolescents with headache or adolescents with back pain showed the highest negative correlations between pain parameters and quality of life. Headache sufferers showed highly negative relationships between pain parameters and most quality of life domains (median r = -0.46), and only in headache sufferers did this involve both intensity and frequency of pain.
Qualitative studies are needed to reveal the background to these differences.
Introduction
Chronic pain, defined as pain that is either continuous or recurrent (i.e. pain with pain-free intervals) for three months or longer and physically unexplained, is common in adolescents. Our previous prevalence study on chronic pain in a representative sample of the open Dutch population showed that over one third of all Dutch adolescents aged between 12 and 18 years of age indicated chronic pain (see below). Most prevalent chronic pain locations were the head (18.9%), a limb (16.2%), abdomen (13.2%), and back (10.9%) (1).
Although prevalence rates for physically unexplained chronic pain in adolescents are rather high, research on the impact of chronic pain is scarce, let alone studies on the differences in impact for different chronic pain locations. These studies focus mainly on subtypes of chronic musculoskeletal pain or headache. Mikkelsson et al. (2) showed that functional disability reported on a subjective functional disability index, was more severe in school children with musculoskeletal pain in more than one area than in children with neck or back pain. The former group reported more difficulty in getting to sleep, sitting during lessons, walking, physical exercise and hobbies because of pain. The majority of studies on differences in impact of subtypes of headache reported no significant differences between migraine and tension-type headache, except for school absence. School absence was more often reported in children and adolescents with migraine than in patients having tension-type headache only (3–7). On the other hand, Andrasik et al. (8) showed that while migraine attacks may lead to short but severe interruptions of the activities of daily life, tension headache had its negative effects in particular on the emotional functioning of the subject (i.e. increased depression). In one of the few studies on differences in life impact between different chronic pain types, Brattberg (9) compared back pain and headache and demonstrated that headache troubled school children more than back pain, in terms of affecting their ability to concentrate, requiring analgesics or leading to absence from school (33% vs. 11%).
In fact, many studies have focused on factors (stress, anxiety and depression) that precede or accompany the most common pain types, such as headache (10), abdominal pain (11–12) and musculoskeletal pain (13), than on the consequences of pain.
Ten years ago Goodman & McGrath (14) concluded in their review on the prevalence of various types of pain in children and adolescents (1991) that the epidemiology of pain in children and adolescents is still relatively undocumented. This conclusion is still valid and due in part to the fact that many of the studies utilized retrospective questionnaire measures for the assessment of pain, although the diary method is still conceived as the gold standard, given its dominant position in clinical trials on pain. Previously we found only modest correlations between the diary and the questionnaire method (r = 0.39) for the assessment of chronic pain (15). A further limitation of earlier work is that regarding the impact of chronic pain many studies evaluated only one life domain, such as functioning at school or psychological functioning using psychopathological measures, for instance the Child Behaviour Check List or the Beck Depression Scale (16). However, the evaluation of quality of life encompasses a number of domains, such as psychological, physical and social functioning, functional status, and satisfaction with life and health. We are of the opinion that quality of life should be assessed accordingly.
Studies using an overall view on quality of life as a consequence of pain, within or among juvenile groups with physically unexplained chronic pain, are scarce. This is due in part to the lack of suitable instruments for the overall assessment of quality of life.
Langeveld et al. (5, 17) developed the first comprehensive generic measure for quality of life assessment in 12–18 years-old headache sufferers (QLH-Y). They showed that nearly all quality of life domains (psychological functioning, physical and functional status) discriminated between the headache and nonheadache group. Headache sufferers experienced more stress, fatigue, depression, and somatic symptoms; they felt less strong, and their mood was less cheerful. In addition, they reported diminished satisfaction with health and life in general than the subjects who never had headaches. The availability of this broad and generic quality of life measure makes it possible to evaluate quality of life as a consequence of juvenile headache and compare it with quality of life resulting from other chronic pains. In adolescents with physically unexplained chronic headache or limb pain we used the same quality of life measure and found a negative relationship between pain and quality of life (15). In the present study, this data set was used to focus on the adolescents with headache in comparison with adolescents who had chronic pain at their back, limbs or abdomen. We therefore addressed the following questions
Is there a difference in quality of life between adolescents with chronic headache and those with pain at other locations (i.e. abdomen, limb, back)?;
What is the relationship between pain frequency and intensity with quality of life in juvenile headache patients?;
Is this relationship different from that in other pain patients?.
We hypothesized that the more intense and the more frequent the headache, the lower the quality of life, but we had no expectations concerning the differences between the pain locations.
Method
Subjects
We recruited adolescents (aged 12–18 years) from among those who, having already given their informed consent in advance while participating in our previous prevalence study on chronic pain in a representative community sample of 5423 Dutch children and adolescents (aged 0–18 years) (1), had indicated chronic pain and who were willing to participate in the current study. The current study sample consists of a quarter of the adolescents who had indicated chronic pain in the prevalence study. Main reasons for drop out were non response; cessation of pain and lack of motivation to complete the pain diary. For the prevalence study the adolescents had marked ‘chronic pain’ to the precoded question ‘How long has the pain been present?’ (1. Previous four weeks; 2. Longer than four weeks; 3. Three months or longer. We considered the latter as chronic pain (18).)
Procedure
The adolescents were sent a pain booklet by mail and a self-addressed envelope. They were invited first to answer the questionnaires and then to keep a diary of the intensity of their pain for three successive weeks. Pain was recorded three times a day, at breakfast, dinnertime and bedtime. Although a measurement point at lunchtime would have yielded more equally distributed data through the day, the measurement point at dinnertime was chosen instead of lunchtime because a pilot study (19) showed that the activities of the adolescents at lunchtime (i.e. during school hours) interfered with adequate diary keeping. It was emphasized that the diary should be filled in at the above mentioned fixed times and relate to the pain intensity at that time. When answering the questionnaires and keeping the diary, adolescents with more than one chronic pain location were asked to refer only to the pain that generally troubled them most. It was decided to restrict the record to only one pain type in the light of a pilot study (19), which showed that completion of the Visual Analogue Scale three times a day for a period of three weeks for more than one pain location resulted in a large reduction in compliance. At the end of the second week of the diary recording-period, the adolescents were sent a postcard reminding them that they had only one more week to go. On return of the pain booklets, they received a CD voucher worth 10 Dutch guilders (i.e. 4 euros).
Instruments
Questionnaires
The questionnaires comprised an information form to register the adolescent's age, gender, history of pain (number of years with pain) and days of school absence resulting from pain in the month prior to the assessment. To assess quality of life, the adolescents completed the Quality of Life Pain – Youth (QLP-Y) questionnaire. The scale was based on the Quality of Life Headache – Youth (QLH-Y) questionnaire, developed by Langeveld et al. (5) and adapted to suit all pain locations. The scale measured the impact of pain on day-to-day functioning in six domains:
psychological functioning (34 items), including the subscales stress, harmony, fatigue, strength, cheerfulness, optimism and depression;
social functioning (12 items) with the subscales functioning at home and interactions with siblings and peers;
functional status (14 items) with the subscales daily living and leisure activities;
physical status (9 items).
Four response categories, ranging from zero to three (i.e. 0=rarely or never; 1=sometimes; 2=often; 3=very often; or: 0 = not at all; 1 = a little; 2=quite a bit; 3=very much), were assigned to the above items. In addition, the QLP-Y included two visual analogue scales to measure (5) satisfaction with life and (6) health (the anchor points for both items ranged from 0 mm ‘completely dissatisfied’ to 100 mm ‘completely satisfied’). The subscale scores were calculated by summing the items of a subscale and dividing the sum score by the number of items of that subscale. The domain scores were obtained by averaging the subscale scores pertaining to a domain. This resulted in domain scores which can range from 0 to −3. The reference period for answering the QLP-Y was the previous week.The higher the scores, the better the self-reported quality of life. QLH-Y is reported to be reliable and valid. The scale has been found to differentiate between adolescents with headache and healthy adolescents, with the former showing significantly poorer quality of life (17).
Diary
The diary consisted of visual analogue scales (VAS) for assessing the intensity and frequency of the pain. The VAS was a horizontal line of 100 mm with ‘no pain’ and ‘the worst imaginable pain’ as anchor points at the left and right ends of the line, respectively. The adolescents were asked to mark a position on the VAS that best matched the intensity of their pain at the moment they had to fill in the VAS. Jensen et al. (20) have demonstrated that the VAS is a valid measure for the assessment of self-reported pain intensity in chronic pain patients.
Statistics
In order to obtain pain intensity scores, the VAS markings were first converted to a score from 0 to 100 by reading off each mark against a millimeter ruler. In addition, VAS scores below 5 mm were coded as zero (no pain). This was done because our previous data showed that only a score of 5 mm or above on the VAS indicated the presence of pain (19). Mean intensity and mean frequency consisted of the means of all pain intensity recordings (≥5 mm or more on VAS) and the percentage of all recordings with pain during the three week diary recording period.
We adapted the threshold for statistical significance to a P-value of less than 0.05 per individual test.
Since the appropriateness of multiple correction for correlated hypotheses is still under debate (21), we chose to refrain from post hoc tests for multiple comparisons (Bonferroni and other procedures).
The description of the statistical analyses is incorporated at the beginning of each of the Results sections.
Results
The group consisted of 95 girls and 33 boys. The average age was 14.7 years (s.d. = 1.4 years), for both boys and girls. Ten of the girls had not yet started to menstruate; the average postmenarche period of the other 85 girls was 3.2 years (s.d. = 2.2 years, range 0.3–9 years). The estimated average history of the pain was 3.3 years (s.d. = 3.1, range 0.3–15 years), for both boys and girls and was not significantly different for different pain locations.
Pain location: impact on pain parameters, quality of life and school absence
To compare adolescents with pain at different locations on quality of life and school absence univariate analyses of variance were used with pain locations as between-subjects factors and quality of life measures (with and without correction for pain parameters) as dependent variables. The univariate analyses of variance were based on the diary records of 111 of the participants (87%), the diaries (VAS) of 17 adolescents having contained more than 25% missing values. To obtain an adequate sample size for statistical analysis, the analyses were additionally restricted to headache (n = 31), abdominal pain (n = 16), back pain (n = 15) and limb pain (n = 38), leaving a group of 100 adolescents, 26 boys and 74 girls. Eleven adolescents were excluded because the sample size regarding their pain type was too small (i.e. neck pain, n = 6 and chest pain, n = 3) or because they had too many missing values in the quality of life questionnaire to obtain valid data (n = 2).
Table 1 shows the means of pain parameters, quality of life domains and school absence for the most frequently reported pain types. Overall, the mean intensity of the pain was low and differences in pain intensity between pain locations were not significant. Overall, the average pain frequency (the percentage of recordings with pain) was high, i.e. between more than half to more than 80 per cent of the pain recording time (total range 2–100%; headache 8–100%; limb pain 2–100%; back pain 43–100%; abdominal pain 12–100%).
Pain parameters, quality of life domains and school absence per pain location
Values are given as Mean (SD). Significant differences
P < 0.05;
P < 0.01.
Headache sufferers were among the adolescents who showed the least frequent pain; adolescents with back pain or limb pain reported pain significantly more often than adolescents with headache or abdominal pain, F (3,96) = 3.494, P = 0.02.
Pain location showed a significant effect on quality of life. Adolescents with headache or back pain showed the relatively poorest quality of life, as revealed by psychological functioning, F (3,89) = 3.838, P = 0.01 and functional status, i.e. the highest interference of their pain with their daily activities, F (3,93) = 2.705, P = 0.05. According to Cohen's δ (22) the effect sizes are medium to large on psychological functioning (explained variance 12%) and functional status (explained variance 8%).
In addition, pain location showed a significant effect on days absent from school as a result of pain (range 0–17 days). Again, headache sufferers reported more than twice as much school absence in the month prior to the assessment than adolescents from the other pain groups, F (3, 90) = 2.69, P = 0.05; a medium to large effect according to Cohen's δ (22) (explained variance 8%). When we looked at the percentage of adolescents with zero days absent from school this was 45% for the adolescents with headache which is almost twice as small as for adolescents with pain at the limbs (82%), abdomen (81%) or back (73%).
Pain location: the relationship between pain parameters and quality of life
Pearson correlations were performed to establish the relationship between pain parameters and quality of life measures within pain locations. Adolescents with headache or back pain showed the highest negative correlations between pain and quality of life (median r =−0.46 and median r =−0.86, respectively), that is, the more the pain the lower the quality of life. Only in adolescents with headache was the pain related to each of the domains of quality of life, except social functioning, and the relationship between pain and quality of life involved both intensity and frequency, except school absence which involved intensity of pain only (Table 2).
Relationship (Pearson correlations) between pain parameters, quality of life and school absence per pain location
VAS.
P<0.05;
P<0.01;
P<0.001.
Discussion
This is one of the few studies on different pain types using a pain diary instead of retrospective measurement for the assessment of pain.
It was salient that adolescents with headache had less pain than adolescents with chronic pain at other locations, but at the same time the poorest quality of life. Their quality of life scores were comparable to those of adolescents with back pain, but the latter group showed the highest pain intensity and frequency. While the differences in quality of life between adolescents with headache and other chronic pains may look small, they are consistent. When we included the intensity and frequency of the pain as covariants in the analyses, the effect of pain location on quality of life remained, indicating that it cannot completely be attributed to the pain parameters. Moreover school absence, which can be considered as a behavioural measure of interference with daily life, was also significantly greater in adolescents with headache. An explanation might be that headaches possibly in particular interfere with cognitive functioning. Adolescents with back pain may be particularly restricted in physical activities, leaving cognitive functioning relatively undisturbed, which might explain their relatively low school absence. Future studies should investigate the pain-location specific problems in more detail. Our finding on school absence is not in line with data from Collin et al. (23) who found very low school absence rate due to headache (0.05% over two 12-week periods vs. 10% during the month prior to assessment). The discrepancy may be attributed to differences in age: our research group was older (12–18 years vs. 5–14 years), which often means more pain report (17).
Our findings should be read with some reserve. First, one may wonder whether gender contaminated the effect of pain location on quality of life, since the ratio of boys to girls is different for the pain locations and girls reported lower quality of life. The girls to boys ratio was smaller in the headache and limb pain group than in adolescents with pain at other locations (2:1 vs. 12:1). Based on this difference in ratio, we would have expected that adolescents with headache reported higher quality of life compared with adolescents with pain at other locations. The fact that this group showed lower quality of life scores, notwithstanding the gender differences in this respect, strengthens the relevance of this finding.
We refrained from differentiating between headache types, such as migraine and tension type headaches, because particularly in adolescents this requires the assessment of an experienced (child) neurologist instead of selfreport.
We should also mention the correlational nature of the study, implying that causality cannot be inferred. Findings from adult studies suggest that emotional problems in pain patients are more likely to be a consequence than a cause of chronic pain (24, 25); other adult studies indicate that they may share common underlying aetiologic factors (26, 27). Prospective, environmental and genetic studies are needed to focus specifically on questions related to causality, divided into different pain types. In addition, interviews to assess the extent of the interference of pain with daily activities are needed. On the basis of such interviews the generic quality of life questionnaire used in this study can be expanded with specific modules related to different pain types.
Implications
Qualitative research is needed to determine the meaning of the relationship between pain and quality of life, particularly in headache sufferers. Our findings support the urgency of headache treatment including quality of life as outcome measure.