- Email: [email protected]
Pain, somatic focus, and emotional adjustment in children of chronic headache sufferers and controls
Sot.Sri.,Med.Vol.31.No. I.pp. 51-59.1990
0277s9536:90 53.00+0.00
Printed in Great Britain. All tights reserved
Copyright C 1990 Pergamon Press plc
PAIN, SOMATIC FOCUS, AND EMOTIONAL ADJUSTMENT IN CHILDREN OF CHRONIC HEADACHE SUFFERERS AND CONTROLS SAMUELF. MIKAIL* and CARL L. VONBAEYER Department
of Psychology, University of Saskatchewan,
Saskatoon, Saskatchewan, Canada S7N OWO
Abstract-Recent epidemiological studies have indicated that as many as 78% of individuals with chronic pain come from families in which at least one other family member has chronic pain. This suggests that children of individuals with chronic pain may be at particular risk for developing chronic pain conditions in the future. This study examined the relationship between parental chronic pain and children’s general adjustment. A group of parents reporting chronic tension or migraine headache and their children were compared to a group of illness-free parents and their children. Results revealed that children of chronic headache sufferers were more somatically focused than their control counterparts. It is suggested that this heightened concern with their health status may be either a reaction to, or a means of coping with ongoing stress. A number of significant correlations between parents’ and their children’s level of emotional adjustment are also reported. The implications of these findings for general family health status, and characteristic ways of dealing with illness are discussed. Key )vor&--family
coping, familial pain, children’s adjustment
Several studies have demonstrated that the severity and duration of various forms of pain are related tb general marital and family adjustment of pain patients. In one such study [3] the relationship between psychological distress levels of pain patients and their spouses was examined. A sample of chronic pain patients and their spouses completed the Symptom Checklist-90 (SCL-90), a widely used measure of psychologic symptom severity. Results revealed that both pain patients and their spouses exhibited a significantly higher global severity index (GSI) score than nonpatients. Furthermore, there was a significant correlation between pain patient and spouse GSI scores when the pain patients’ GSI score was above the mean for normative psychiatric patients. Based on this result it was concluded that the correlation between patient and spouse scores is not linear, but rather, is dependent on the patient having high symptom severity scores. This result suggests that exposure to a person having high levels of symptom severity could influence the emotional adjustment of the spouse, and potentially other family members. Investigations have also revealed high levels of maladjustment in marriages of chronic pain patients [4,5]. It has been shown that as many as two-thirds of the couples of which one member has chronic pain reported deterioration in sexual activity, including decreased frequency and quality of sexual relations, while more than one-third reported deterioration in the marriage itself [4]. Furthermore, ratings of marital adjustment reflected a consistent trend toward deterioration following the onset of pain, although there was a tendency for patients to minimize the changes taking place in comparison to their partners [5]. These results suggest that the presence of a pain problem in a family member has an effect on family
INTRODUCl’ION Pain researchers are beginning to recognize the importance of familial factors associated with the onset, maintenance and treatment of chronic pain syndromes (1,2]. This research has generally taken one of two forms. It has either focused on retrospective analysis of familial incidence of pain or its has examined the effects of chronic pain on the psychological well being of spouses of chronic pain patients. In an effort to examine the nature of familial patterns of chronic pain Mohamed, Weisz, and Waring [l] compared the incidence of pain problems among family members of depressed chronic pain patients and depressed patients without pain. The study demonstrated that there was a greater incidence of pain problems among spouses, family members of spouses, and family members of the chronic pain patients, than among the family members of depressed patients. Furthermore, there was greater consistency of pain location among family members of the depressed chronic pain patients than among the depressed patients. In their conclusions the authors stress the potential importance of family dynamics in the propagation of pain symptoms. Using a similar approach, Violon and Giurgea [2] compared the family histories of a group of 40 chronic pain patients and a group of 50 patients with chronic but pain-free disease. Results revealed of chronic pain patients 78% reported that at least one member of their family had chronic pain, while 44% of the patients in the control group reported such a history. *Address correspondence to: Dr Samuel F. Mikail, Department of Psychology, The Rehabilitation Centre, Ottawa, Ontario, Canada KIH 8M2. 51
52
SAMUELF. MIKAIL and CARL L. VON BAEYER
adjustment, without full recognition or admission of such difficulties on the part of the patient. In a subsequent study [6] the relationship between level of pain and general family functioning was examined. It was found that certain family characteristics, as measured by the Family Environment Scale [7] were predictive of pain rating indices on the McGill Pain Questionnaire [S]. Several components of family functioning were related to the various dimensions of the pain experience of chronic pain patients. Higher levels of pain were associated with greater levels of family independence, more conflict and control, and a greater tendency toward achievement and organization. These findings provided empirical support for the contention that the psychosocial environment of the pain patient is related to his/her perceptions, rating and overall experience of pain. In summary, research examining the reiationship between chronic pain and family functioning has led to two conclusions: families of individuals with chronic pain have a higher incidence of pain conditions than families of patients with depression or pain-free chronic illness; and chronic pain appears to have an adverse effect on the marital relationship of pain patients and their partners. Although an obvious explanation for the first conclusion is genetic vulnerability, a recent review [9] emphasizes that at present there is little support for this contention. A number of authors have suggested that a coping and stress model may have heuristic value in accounting for patterns of illness observed in family groups. In one such review [lo] it is suggested that family units develop characteristic levels of health and illness which tend to be stable over time. The authors propose that such patterns arise out of patterned regularities in interactions among family members, and between the family and the community. In support of this view, it has been demonstrated that when the unit of analyses is the family, characteristic family coping strategies can be identified [l I]. Such coping strategies were found to vary across a variety of situations, and in turn predict level of perceived stress and adaptation. One difficulty with the research described is that few investigators have explored general family functioning in families of chronic pain patients. Only one study to date [12] has examined the level of children’s emotional adjustment when one of the parents had chronic pain. The study was limited to an examination of children’s emotional adjustment when the mother had chronic pain. Results revealed that families with a mother who had chronic pain had poorer perceived family environments, and higher levels of depression and anxiety among the parents than control families. Furthermore, children in these families were also found to be more depressed than their control counterparts. While the results of this study are important, they are limited by the relatively small sample size employed. The present study adds to the findings of this investigation by employing a larger sample size and varying the methodology somewhat. The aim of the present study was two fold: (1) to compare the incidence of pain related illness and general behavioral disturbance among children of
two parent groups; a chronic headache group and an illness-free group, and (2) to compare parent ratings of family functioning across groups. In addition, the relationship between ratings of family functioning and severity of parental pain complaint was examined in the chronic headache group. A number of specific predictions were made. First, it was hypothesized that there would be a greater incidence of pain related illness, such as headaches and abdominal pain, among children having a parent with chronic headaches, than among children of illness-free parents. Second, it was expected that there would be a greater incidence of general behavioral disturbance among children of chronic headache parents than among children of illness-free parents. It was also predicted that there would be a positive correlation between parent’s symptom severity and incidence of general behavioral disturbance of the child. Third, it was predicted that among chronic headache parents, ratings of family functioning would correlate negatively with pain rating indices. Finally, it was expected that ratings of family environment would reflect more dysfunction among families of chronic headache parents when compared with families having illness-free parents. METHOD
Subject
Patients treated for chronic headache in the Pain Management Service, University Hospital, Saskatoon, were contacted by mail and asked to volunteer their participation. Subjects had to be living with a spouse who did not suffer from a chronic illness. Subjects for the control group were obtained through a general optometry practice in the city of Saskatoon. Names were reviewed by one of the optometrists such that names, addresses and phone numbers of all families with illness-free parents, and children falling in the appropriate age range were extracted. Agreement to participate was obtained by the office staff before the names of the families were released to the research team. The study was described as investigating the effects of various health conditions on general family functioning. Children of both genders between the ages of 9 and 17 were included. This was to ensure that they were old enough to understand all instructions, and young enough to be living at home. Of a total of 38 chronic pain patients contacted, 14 declined participation, resulting in a 63.2% acceptance rate. A total of 49 illness-free families were contacted. Of these 19 declined participation, resulting in a 61.2% acceptance rate. Generally, refusal to participate was due to either other commitments, distances involved in traveling to the hospital, or scheduling difficulties. No one refused to participate due to the nature of the study. Thus, it can be assumed that the samples were fairly representative of the populations from which they were drawn. Procedure
Parent-child pairs underwent individual testing sessions at the hospital, with the entire procedure taking approx. 90-120min. Children were asked to respond to a number of questions concerning their
53
Pain in children of chronic headache sufferers health history and the frequency with which they experience headaches. This information was corroborated with parents. Parents were also asked to indicate the frequency with which they experienced headaches. Additionally, parents were asked to complete the short form of the Personality Inventory for Children (PIC), the Family Environment Scale (FES), and the Symptom Checklist-90 @CL-90). Chronic pain sufferers were also asked to complete the West Haven-Yale Multidimensional Pain Inventory (WHYMPI). As parents completed these questionnaires their children completed the Comprehension and Similarities subtests of the Wechsler Intelligence Scale for Children-Revised, and the Quick Test. These tests were administered in order to ensure that children in both groups did not differ with regard to level of cognitive complexity, or intellectual development. Instruments The PIG-short f&m consists of responses to 280 items, answered ‘True’ or ‘False’ by the parent of each child [ 131.The inventory provides scores on four factor scales, a Lie scale, a Development scale, and 14 clinical scales. The FES is a go-item self-report instrument measuring the social-environmental characteristics of families. The questionnaire is composed of 10 scales assessing three underlying dimensions including: Relationship, Personal growth, and System Maintenance [I. The SCL-90, is composed of a list of 90 different symptoms [14]. Using a 5-point Likert scale, subjects are required to indicate the degree to which they have been bothered by each of the symptoms in the last month. The scale provides scores on 9 factors including somatization, obsessive-compulsiveness, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. A General Symptom Index (GSI) based on the total score is also calculated. The WHYMPI is a 52-item pain assessment inventory composed of 12 scales falling into one of three categories; the impact of pain on the patient’s life, the responses of others to the patient’s communications of pain, and the extent to which the patient participates in common daily activities [15]. The WISC-R
Subtests
The similarities and comprehension subscales of the WISC-R were administered to each of the children [ 161. The similarities is composed of a series of 17 word pairs. The child is asked to indicate the manner in which each of the words in a given pair are alike. Word pairs are presented in order of increasing difficulty. The comprehension subscale is made up of 17 questions, each reflecting increasingly complex levels of knowledge. The Quick Test
This general measure of intelligence requires matching pictures with words from a list arranged in order of increasing difficulty [17].
RESULTS Data analyses are presented in the following order: data assessing the initial equivalence of groups, preliminary analyses, and tests of the hypotheses. Tests of the hypotheses were based on l-tailed tests of significance, while initial equivalence data, preliminary analyses and post-hoc analyses were based on 2-tailed tests of significance. Initial equivalence of the two parent groups was assessed using the following variables: age; gender; employment status of the primary wage earner in the family; highest level of education attained by the primary wage earner; and number of children in the family. No significant differences were found between the two groups on any of these variables with the exception of number of previous marriages. Chronic headache patients had significantly more previous marriage (M = 0.43, SD = 0.51) than control subjects (M = 0.07, SD = 0.25), t(49) = 3.36, P c 0.002. This finding is consistent with reports in the literature pointing to the high degree of marital discord reported by chronic pain patients and their spouses (e.g. I41). Initial equivalence of the two groups of offspring was assessed using the following variables: age; gender; birth order; Quick Test IQ; WISC-R comprehension and similarities scores; and health history. A significant age difference was found, with children of pain patients having a mean age of 13.67 years (SD = 2.71), while children of controls had a mean age of 12.23 years (SD = 1.75), t(49) = 2.30, P < 0.026. No other significant differences were found on the remaining initial equivalence variables. Preliminary
analyses
In order to determine any potential effects of offspring group differences in age, this variable was correlated with the major dependent variables including children’s scores on the PIC, children’s monthly frequency of headaches, and a number of parental variables. None of these correlations was found to be significantly different from zero. Thus, age differences between the two groups of children were ignored with regard to all subsequent analyses. Parental group composition
In order to verify the chronicity of headaches in the chronic pain group the average monthly frequency of headaches of the two parent groups was compared. The mean number of headaches per month for members of the chronic pain group was 16.63 (SD = 9.17), as compared to 0.90 (SD = 1.27) for members of the control group [F(l, 53) = 86,487, P < O.OOOl]. Thus, the experimental group represented a chronic pain sample in comparison to the illness-free controls. Tests of the hypotheses Hypothesis I. It was hypothesized that children of headache sufferers would exhibit a greater incidence of pain related illness than children of illness-free controls. Group differences on two components of the PIC; the somatic concerns scale and factor 3 (internalization/somatic symptoms) were examined. As can be seen from the results presented in Table 1,
SAMWL F. MIKAILand CARL L. VON BAEYER
54
Table I. Comparison
of child illness categories by group Pain (N = 24) Control (N = 30)
Scale Frequency of headaches (per month) Factor 3 (intemalizing~somatization) Somatic concerns scale
Table 2. Comparison
Mean (SD)
Mean (SD)
I
One-tailed P
2.16 (3.96) 5.83 (5.03) 52.88 (10.96)
l.ZO(l.35) 3.90 (2.33) 47.70 (8.66)
2.01 1.87 1.95
0.03 0.03 0.03
of somatic concerns subscale scores by group Pain (N = 24) Control (N = 30)
Scale
Mean (SD)
Mean (SD)
Frequency of illness Infrequent symptoms Somatizing Visual problems Excessive constipation Poor stamina Body temperature Muscular spasms Attention seeking Other
0.42 (0.58) 1.29(1.04) 0.67(1.17) 0.25 (0.53) 0.25 (0.44) 0.88 (0.80) 0.38 (0.71) 0.46 (0.78) 0.71 (0.75) 0.42 (0.50)
0.33 (0.55) 0.90(1.35) 0.30 (0.47) 0.23 (0.50) 0.27 (0.45) 0.93 (0.87) 0.13 (0.43) 0.23 (0.43) 0.53 (0.63) 0.17 (0.38)
a significant difference was found for both of these two components. Children of chronic pain sufferers had a significantly greater degree of somatic concerns as compared to children of healthy controls. Group differences on factor 3 scores were also significant, with the patterns of means being in the same direction. For purposes of clarification the somatic concerns scale was broken down into its component subscales (see Table 2). The two groups were compared on these in order to determine specific symptom groups which accounted for the observed difference on the scale. A significant difference between group means was found on the other subscale. This subscale was made up of two items: “my child has as much pep and energy as most children” and “my child talks a lot about his (her) size or weight”. In comparison to controls, parents in the chronic headache group were more likely to endorse the first item as false and the second item as true of their child. The second step in testing this hypothesis was to compare the monthly frequency of headaches reported by the two offspring groups. Analysis revealed that children of chronic headache sufferers reported significantly more headaches per month than did children of illness-free parents (see Table 1).
, 0.54 1.17 1.57 0.12 -0.14 -0.25 1.54 1.35 0.93 2.08
Two-tailed P 0.59 0.25 0.12 0.91 0.89 0.80 0.13 0.18 0.36 0.04
Thus, children of chronic headache sufferers reported experiencing more headaches per month, appeared to be more concerned with their body image, and were reported to have less energy than other children. However, these group differences did not generalize to other pain related illnesses. Hypothesis 2. This hypothesis was divided into two parts. First, it was suggested that there would be a higher incidence of general behavioral disturbance among children of chronic pain sufferers than among children of illness-free parents. This part of the hypothesis was tested by comparing group means using the PIC clinical scales. As indicated above, children of chronic pain sufferers demonstrated higher mean scores on the somatic concerns scale than did children of healthy controls. Additionally, a significant difference was found on delinquency scores. The items on the scale present a broad range of antisocial and acting-out behaviors (see Table 4 for a listing of subscales). It was found that children of chronic headache sufferers had a significantly higher delinquency score than children of healthy controls. As can be seen in Table 3, significant differences were also found on the general adjustment and social skills scales. Specifically, children of chronic headache sufferers were reported as having higher maladjustment and
Table 3. Comparison of level of behavioral disturbance among children of chronic headache sufferers and healthy controls
Pain (N = 24) Control (N = 30) Scale General adjustment Intellectual screening Developmental screenir ‘B Somatic concerns Family relations scale Delinquency Withdrawal Anxiety Psychoticism Hyperactivity Social skills Reported means are based on T-scores.
Mean (SD)
Mean (SD)
52.65 (12.39) 50.84 (10.70) 52.46 (9.30) 52.88 (10.96) 52.37 (12.82) 53.55 (13.36) 51.13(9.85) 51.83(11.19) 49.11 (9.15) 50.17(11.43) 47.40 (8.99)
47.91(7.29) 49.33 (9.54) 48.04 (10.26) 47.70 (8.66) 48. I2 (6.69) 47.40 (5.48) 49.14(10.19) 48.54 (8.86) 50.69(10.71) 49.87 (8.89) 51.99 (10.42)
I 1.72 0.55 1.64 1.94 1.54 2.28 0.72 I .20 -0.57 0.11 -1.68
One-tailed P 0.05 0.59 0.11 0.03 0. I3 0.01 0.48 0.23 0.57 0.91 0.05
Pain in children Table 4. Comparison
of chronic
headache
sufferers
55
of mean subscale scores of the PIC delinquency scale for children of chronic headache sulTercrs and healthy controls
Scale Disregard for limits,mterpersonal Ann-social tendeno:‘s Irritability/limited tolerance Sadness Lack of interest/impulsivity Interpersonal hostility Other
insensitivity
Pain (N = 24)
Control (IV = 30)
Mean (SD)
Mean (SD)
I
Two-tailed P
1.75 (2.05) 1.30 (2.32) 0.58 (0.88) 0.29 (0.46) 1.08 (0.58) 0.08 (0.28) 1.17(0.49)
0.93 (I .39) 0.60(1.10) 0.30 (0.60) 0.03 (0.18) 0.70 (0.47) 0.00 (0.00) 0.90(0.31)
1.74 I .47 1.41 2.79 2.68 I .62 2.49
0.09 0.15 0.17 0.01 0.01 0.1 I 0.02
Reported means are based on raw scores. Table 5. Comparison
of level of psychological
symptom severity of chronic headache sufferers and healthy controls Pain (N = 24) Control (N = 30)
SCL-90 scale
Mean (SD)
Somatization Obsessive-compulsive Interpersonal communication Depression Anxiety Hostility Phobic anxiety Paranoia Psychoticism GSI
55.53 (I 1.70) 53.80(11.45) 53.80(13.07) 54.01 (12.34) 53.50(13.09) 53.57 (12.82) 54. I I (14.05) 50.66(12.19) 52.42 (13.41) 53.60 (34.10)
I
Two-tailed P
3.70 2.38 2.38 2.52 2.17 2.21 2.59 0.39 1.46 2.63
0.02 0.02 0.02 0.04 0.03 0.01 0.70 0.15 0.01
Mean (SD) 46.13 47.34 47.34 47.19 47.55 47.50 47.13 49.54 48.31 34.78
(6.35) (8.01) (6.07) (6.90) (6.26) (6.60) (3.94) (8.33) (6.41) (16.32)
0.00
Reported means are based on T-scores.
lower social skills scores than children of illness-free controls. Once again, in order to clarify these findings, post-hoc analysis were conducted in which group differences were examined by breaking down the three scales into their component parts. As can be seen from Table 4, a number of group differences were found on the subscales of the delinquency scale. Children of chronic headache sufferers exhibited significantly higher scores on the sadness subscales, lack of interest and impulsivity subscale, and the ‘other’ subscale. The ‘other’ subscale includes three items; “my child can’t sit still in school because of nervousness”, “several times my child had threatened to run away”, and “the child’s father has very little patience with the child”. No individual subscale differences were found on the general adjustment or the social skills scales. Second, it was suggested that there would be a positive correlation between measures of parent and Table 6. Significant correlations
child nsvchological symptom severity. Since the parental *groups differed -on a number of SCL-90 subscales (see Table 5), correlational analyses were conducted separately for each of the groups. Results of these analyses are presented in Tables 6 and 7. For both groups a number of moderate to strong correlations were observed between measures of parent and child psychological symptom severity. Generally, these correlations were higher for parentchild pairs in the chronic pain group. However, it must be noted that variances in SCL-90 scale scores of pain patients, and PIC scale scores of their offspring, were greater than in the case of their control counter parts. This greater variability in scores may have contributed to the higher correlations in this group. Generally, however, the data corroborate the hypothesized relationships. Hypothesis 3. A number of predictions were made regarding the relationship between ratings of family environment and parental health status. A negative
between parents’s symptom severity and incidence of general behavioral disturbance of the child in .oain _. group SCL-90 symptom scales (parent)*
PIC symptom scales (child) Intellectual screening Adjustment Development screening Somatization Family relations Withdrawal Anxiety Delinquency Hyper&ti&y Psychoticism
I
2
3
4
0.40
5
6
7
8
9
GSI
0.58
0.41
0.45
0.59 0.61
0.59 0.73
0.62 0.71
0.49
0.46 0.52
0.44 0.65
0.51 0.80
0.67 0.57
0.59
0.62 0.58 0.49 0.58
0.48 0.38 0.46 0.67
0.39
0.42
0.47
0.44
0.43
0.57
0.66
0.56
0.67
0.70
0.69
0.63
0.69
0.63
0.71
0.73
0.70
0.74
0.72 0.62
*I = Somatization, 2 = obsessive-compulsive, 3 = interpersonal sensitivity, 6 = hostility, 7 = phobic anxiety, 8 = paranoid ideation, 9 = psychoticism. Test of significance of correlations is based on a i-tailed r-test, and P < 0.05.
4 = depression,
5 = anxiety,
56
SAMUEL Table
7. Significant
correlations
F.
MIKAIL and
between disturbance
CARL L. vo~i BAE\EYER
parent’s symptom severity and incidence of the child in the control group SCL-90 symptom
PIC symptom (child)
scales (parent)*
I
2
3
4
5
6
7
West Haven-Yale Multidimensional Pain Inventory Scales Interference Pain severity Negative mood Se&on 3 Household chores Outdoor activity Activity away from home Social activity
GSI
0.34
0.40
0.31 0.4-l
0.43 0.60 0.56 0.33 0.36
-0.33
0.46 - 0.36
-0.34 -0.42
-0.34
correlation between positive ratings of family functioning and pain rating indices was predicted for chronic pain sufferers. Pain rating indices were obtained using the West Haven-Yale Multidimensional Pain Inventory. A number of scales of this questionnaire measure level of pain severity. For purposes of the present study these included scale 1 (degree to which pain interferes with daily activities), scale 3 (pain severity), scale 5 (negative mood), and all four scales of section 3 (these measure the degree to which the individual engages in a variety of activities including household chores, outdoor work, activities away from home, and social activities). Ratings of family functioning were obtained by use of the Family Environment Scale. The pattern of correlations obtained (see Table 8) suggests that the predicted relationship between pain severity and positive family functioning is not as robust and direct as expected. Specifically, not all dimensions of family environment seemed to be related to pain severity. Scale 1 of the WHYMPI achieved a significant negative correlation with level of family cohesion and active-recreational orientation of a family (r= -0.39,P < 0.05; r = -0.44,P < 0.05respectively). Thus, the greater the degree of cohesion in a family, the lower the degree to which pain is reported to interfere with daily activities. Similarly, the greater the degree to which a family is oriented toward recreational interests, the less pain interferes with daily activity. Pain severity was negatively correlated with level of expressiveness and active-recreational correlations
9
0.53 0.36 0.41
I = Somatization. 2 = obsessivesompulsive, 3 = interpersonal sensitivity, 6 = hostility, 7 = phobic anxiety, 8 = paranoid ideation, 9 = psychoticism. Test of significance of correlations is based on a l-tailed r-test, with P < 0.05.
8. Significant
8
0.38
l
I = Cohesion,
behavioral
scales
Adjustment Developmental screening Intellectual screening Somatization Family relations Delinquency Withdrawal Anxiety Psychoticism Hyperactivity Social skills
Table
of general
between
DISCUSSION
The present investigation examined the incidence of pain related illness and general behavioral disturbance among children of two parent groups.
rating of family environment headache sufferers
2
3
4
-0.39 -0.40
-0.53 0.39
5 = anxiety.
orientation, and positively correlated with level of family organization. The presence of negative mood was negatively correlated with level of family cohesion and active-recreational orientation, and positively correlated with level of expressiveness. There were almost no significant correlations between the 4 scales of section 3 of the WHYMPI and the various dimensions of family functioning. The only significant correlation achieved was a positive correlation between the extent to which an individual engages in outdoor activity and the tendency to encourage independence of family members. Collectively, these results partially support the proposed hypothesis. However, it is clear that the relationship between pain intensity, or level of disability, and family functioning is less direct than expected. A second component of the hypothesis was the prediction that ratings of family environment would reflect more dysfunction among families of chronic headache sufferers than families of healthy controls. A series of oneway analyses of variance, comparing mean scores on the various subscales of the Family Environment Scale for the two groups reveal no significant differences. Thus, this hypothesis was not substantiated by the data.
Family
I
4 = depression,
0.37
and pain
environment 5
6
rating
indices
of chronic
scales* 7 -0.44 -0.34 -0.40
8
9
IO
-0.36
0.41
2 = expressiveness, 3 = conflict. 4 = independence, 5 P achievement orientation, 6 = intellectualcultural orientation, 7 = active-recreational orientation, 8 = moral religious orientation, 9 = organization. 10 = control. Test of significance for all correlations is based on a l-tailed f-test, with P < 0.05 level of significance. l
Pain in children of chronic headache sufferers
Additionally, a number of family environment variables were examined as they related to family illness patterns. With regard to the first hypothesis, it was expected that children of chronic headache sufferers would exhibit a greater incidence of pain related illness than children of illness-free controls. There was only partial support for this hypothesis. Children of chronic headache sufferers did achieve higher scores on the somatic concerns scale and factor 3 (intemalizing/somatization) of the PIC. However, items differentiating between the two groups dealt with concerns regarding body image and lack of energy. Specifically, children of chronic headache sufferers were reported to have less pep or energy than most children, as well as being more concerned about their size or weight. In addition, children of headache sufferers reported having a higher frequency of headaches per month than did children of headachefree parents. The somatic concerns scale was developed with the intention of identifying children who tend to have an exclusively ‘neurotic adjustment’ to stressful situations [ 131. Specifically, children who achieve a high score on this scale tend to respond to stress through a heightened concern over body functioning, energy level, diffuse physical symptoms, and so on. One can argue that such a focus may develop in response to repeated illness, prolonged episodes of pain, or in this case, frequent headaches. However, it is important to note that children suffering from chronic health conditions or illness do not achieve elevated scores on this scale [ 131. Collectively, these results suggest that children of chronic headache sufferers are more concerned with their health status and bodily functioning than children of healthy controls. That is, they tend to be more somatically focused. As noted by the developers of the PIC, becoming somatically focused tends to occur in response to stress [13]. Although the present study included no measures of perceived stress, several of the reported findings are consistent with this relationship. It was found that children of chronic headache sufferers have significantly higher scores on the general adjustment and delinquency scales than did controls. When this result was examined more closely it became apparent that the components of the delinquency scale which accounted for this group difference included the sadness subscale, the lack of interest/ impulsivity subscale, and the ‘other’ subscale. Items on these subscales focused on such factors as negative or depressed mood, distractability, inability to concentrate or remain focused, and poor family relations. As can be seen, these two scales, and their component subscales, can be considered to reflect either sources of stress, or end products of stress. In summary, it would appear that children of chronic headache sufferers tend to be more somatically focused than their control counterparts, and that this heightened concern with their health status may be either a reaction to, or a means of coping with ongoing stress. A number of hypotheses were made regarding family functioning and illness behavior. First, it was suggested that a positive correlation would be
57
observed between measures of parent and child psychological symptom severity. This relationship was examined separately for each of the parentxhild groups for reasons outlined earlier. A number of significant relationships were identified. The most interesting of these was a positive correlation between parents’ somatization scores and children’s somatic concerns scores. This correlation was found to be significant in both groups (r = 0.72 in the chronic headache group and r = 0.35 in the control group). If somatization is viewed as a characteristic style of coping [13], then this finding provides tentative support for the contention that there exists characteristic family coping strategies, and that these may affect the vulnerability of family members to certain illnesses [l 11. For example, individuals having chronic headaches tend to respond to stressful situations through increased muscle contraction [18], and may be more predisposed to headaches following such periods. It is conceivable that family members may exhibit similar responses to stressful situations. Another finding of interest is the pattern of correlations between parent and child symptom severity in the two groups. In the chronic headache group positive correlations were found between all scales of the SCL-90 and several of the PIC scales (i.e. somatic concerns, family relations, anxiety, and psychoticism). No such pattern was observed in the control group. It will be recalled that with the exception of paranoid ideation and psychoticism, chronic headache sufferers have significantly higher elevations on all scales of the SCL-90, including the general severity index (GSI). In a previous investigation it was found that there was a significant correlation between pain patient and spouse GSI scores when the patient’s scores were above the mean for normative psychiatric patients [3]. Results of the present study can be considered analogous to this finding. In fact, given that normal, as opposed to psychiatric controls were used in the present study, the results suggests that emotional adjustment of children may be more influenced by levels of parental symptom severity, than the emotional adjustment of spouses. This would make sense given that a child’s capacity to deal with and filter out another family member’s behavior is likely to be less developed than that of a mature adult. One potential limitation of these results is that the parent was required to fill out both the PIC and the SCL-90. Thus, it may be argued that at least part of the observed correlation in parent and child emotional functioning is due to shared method variance. However, this is not a serious limitation as all scales of the PIC have been shown to correlate highly with more objective ratings of children’s behavior [13]. Furthermore, inter-rater agreement for the instrument as a whole has been found to be quite acceptable [13]. The prediction was made that ratings of family environment would reflect more dysfunction among families of chronic headache sufferers in comparison to families of healthy controls. This hypothesis was not substantiated by the results. The absence of an effect can be explained in at least two ways. First, it may reflect the tendency of chronic pain patients to idealize relationships both previous to and following
58
SAMUEL
F.
MIKAIL
and
CARL
L. VON BAEYER
the onset of pain. This tendency was
[ 111. In studying the nature of family environments, it is essential that multiple ratings of family functioning be obtained. If pain patients are found consistently to present a more optimistic account of their relationships than do other family members, we may need to begin paying closer attention to this potential blind spot in our treatment of these patients. It may well be that the higher incidence of marital discord in these families is due to the patient’s inability to recognize, and in turn deal with, such interpersonal difficulties. Finally, a word about the focus of research and treatment in the area of chronic pain. For the most part the efforts of researchers and clinicians alike, have focused on gaining an understanding of chronic pain from the perspective of patients and their partners. Undoubtedly, this work is important and needs to continue. However, to date, the effects of chronic pain on patients’ children has not been adequately investigated. The above results suggest that children of pain patients may differ on several dimensions from children of healthy controls. Furthermore, the data suggest that these differences may in part be related to the manner in which these children respond to their parents’ condition. Children may experience their parents’ suffering as an ongoing stressor.
Pain in children of chronic headache sufferers However, they may be brought up in an environment in which (a) they are not taught how to adequately deal with such a stressor and/or (b) they are not exposed to effective coping models. Furthermore, they are not included in any component of their parents’ treatment program. Although results of the present study do not allow for any specific suggestions as to how children could be included in their parents’ treatment program, they do suggest that future research should address this issue. authors wish to extend their gratitude to Dr Kenneth Craig, Dr James Pond, Dr Michael Sullivan and Dr Joyce D’Eon for their helpful comments. This research was carried out as part of the first author’s doctoral dissertation in psychology with support of a training fellowship from the Saskatchewan Health Research Board.
7. 8. 9. 10.
Acknowledgements-The
REFERENCES
1. Mohamad S. N., Weisz G. M. and Waring E. M. The relationship of chronic pain to depression, marital adjustment, and family dynamics. Pain 5, 285-292, 1978. 2. Violin A. and Giurgea D. Familial models for chronic pain. Puin 18, 199-203, 1984. 3. Shanfield S. B., Heilman E. M., Cope D. N. and Jones J. R. Pain and the marital relationship: psychiatric distress. Pain I, 343-351, 1979. 4. Maruta T. and Osborne D. Sexual activity in chronic pain patients. Psychosomatics 19, 531-537, 1979. 5. Maruta T., Osborne D., Swanson D. W. and Halling J. A. Chronic pain patients and spouses: marital and sexual adjustment. Mayo Clinic Proc. 56,307-310, 1981. 6. Feuerstein M., Suit S. and Houle M. Environmental
11. 12.
59
stressors and chronic low back pain: life events, family and work environment. Puin 22, 295-307, 1985. Moos R. H. and Moos B. S. Family Environment Scale: Manual. Consulting Psychologist’s Press, Palo Alto, CA, 1981. Melzack R. The McGill pain questionnaire: major properties and scoring methods. Pain 1, 275-299, 1975. Turk D. C., Flor H. and Rudy T. E. Pain and families. I. Etiology, maintenance, and psychosocial impact. Puin 30, 2-28, 1987. Coyne J. and Holroyd K. Stress, coping and illness: a transactional perspective. In Handbook of Clinical Healfh Psychology (Edited by Millon T., Green C. and Meagher k.). Plenum Press, New York, 1982. McCubbin H. I. Integrating coping behavior in family stress theory. J. Marriage & Family 41, 237-244, 1979. Dura J. R. and Beck S. R. A comparison of family functioning when mothers have chronic pain. Pain 35,
79-90, 1988. 13. Wirt R. D.. Lachar D.. Klinedinst J. K. and Seat P. D. Multidimensional Description of Child Personality: A Manual for the Personality Inventory for Children.
Western Psychological Service, Los Angeles, CA, 1984. 14. Derogatis L. R. XL-90 Adminisfration, Score and Procedures Manual. Johns Hopkins University Press, Baltimore, MD, 1977. 15. Kerns R. D., Turk D. C. and Rudy T. E. The West Haven-Yale Multidimensional Pain Inventory. Pain 23, 345-356, 1985. 16. Wechsler D. Manualfor the Wechsler Intelligence ScaleRevised. The Psychological Corporation, New York, 1974. 17. Ammons R. B. and Ammons C. H. The Quick Test (QT): provisional manual. Psychol. Rep. 11, 11I-161, 1962.
18. Traue H. C., Gottwald A., Henderson P. R. and Bakal D. A. Nonverbal expressiveness and EMG activity in tension headache sufferers and controls. J. psychosom. Res. 29, 375-381, 1985.
0277s9536:90 53.00+0.00
Printed in Great Britain. All tights reserved
Copyright C 1990 Pergamon Press plc
PAIN, SOMATIC FOCUS, AND EMOTIONAL ADJUSTMENT IN CHILDREN OF CHRONIC HEADACHE SUFFERERS AND CONTROLS SAMUELF. MIKAIL* and CARL L. VONBAEYER Department
of Psychology, University of Saskatchewan,
Saskatoon, Saskatchewan, Canada S7N OWO
Abstract-Recent epidemiological studies have indicated that as many as 78% of individuals with chronic pain come from families in which at least one other family member has chronic pain. This suggests that children of individuals with chronic pain may be at particular risk for developing chronic pain conditions in the future. This study examined the relationship between parental chronic pain and children’s general adjustment. A group of parents reporting chronic tension or migraine headache and their children were compared to a group of illness-free parents and their children. Results revealed that children of chronic headache sufferers were more somatically focused than their control counterparts. It is suggested that this heightened concern with their health status may be either a reaction to, or a means of coping with ongoing stress. A number of significant correlations between parents’ and their children’s level of emotional adjustment are also reported. The implications of these findings for general family health status, and characteristic ways of dealing with illness are discussed. Key )vor&--family
coping, familial pain, children’s adjustment
Several studies have demonstrated that the severity and duration of various forms of pain are related tb general marital and family adjustment of pain patients. In one such study [3] the relationship between psychological distress levels of pain patients and their spouses was examined. A sample of chronic pain patients and their spouses completed the Symptom Checklist-90 (SCL-90), a widely used measure of psychologic symptom severity. Results revealed that both pain patients and their spouses exhibited a significantly higher global severity index (GSI) score than nonpatients. Furthermore, there was a significant correlation between pain patient and spouse GSI scores when the pain patients’ GSI score was above the mean for normative psychiatric patients. Based on this result it was concluded that the correlation between patient and spouse scores is not linear, but rather, is dependent on the patient having high symptom severity scores. This result suggests that exposure to a person having high levels of symptom severity could influence the emotional adjustment of the spouse, and potentially other family members. Investigations have also revealed high levels of maladjustment in marriages of chronic pain patients [4,5]. It has been shown that as many as two-thirds of the couples of which one member has chronic pain reported deterioration in sexual activity, including decreased frequency and quality of sexual relations, while more than one-third reported deterioration in the marriage itself [4]. Furthermore, ratings of marital adjustment reflected a consistent trend toward deterioration following the onset of pain, although there was a tendency for patients to minimize the changes taking place in comparison to their partners [5]. These results suggest that the presence of a pain problem in a family member has an effect on family
INTRODUCl’ION Pain researchers are beginning to recognize the importance of familial factors associated with the onset, maintenance and treatment of chronic pain syndromes (1,2]. This research has generally taken one of two forms. It has either focused on retrospective analysis of familial incidence of pain or its has examined the effects of chronic pain on the psychological well being of spouses of chronic pain patients. In an effort to examine the nature of familial patterns of chronic pain Mohamed, Weisz, and Waring [l] compared the incidence of pain problems among family members of depressed chronic pain patients and depressed patients without pain. The study demonstrated that there was a greater incidence of pain problems among spouses, family members of spouses, and family members of the chronic pain patients, than among the family members of depressed patients. Furthermore, there was greater consistency of pain location among family members of the depressed chronic pain patients than among the depressed patients. In their conclusions the authors stress the potential importance of family dynamics in the propagation of pain symptoms. Using a similar approach, Violon and Giurgea [2] compared the family histories of a group of 40 chronic pain patients and a group of 50 patients with chronic but pain-free disease. Results revealed of chronic pain patients 78% reported that at least one member of their family had chronic pain, while 44% of the patients in the control group reported such a history. *Address correspondence to: Dr Samuel F. Mikail, Department of Psychology, The Rehabilitation Centre, Ottawa, Ontario, Canada KIH 8M2. 51
52
SAMUELF. MIKAIL and CARL L. VON BAEYER
adjustment, without full recognition or admission of such difficulties on the part of the patient. In a subsequent study [6] the relationship between level of pain and general family functioning was examined. It was found that certain family characteristics, as measured by the Family Environment Scale [7] were predictive of pain rating indices on the McGill Pain Questionnaire [S]. Several components of family functioning were related to the various dimensions of the pain experience of chronic pain patients. Higher levels of pain were associated with greater levels of family independence, more conflict and control, and a greater tendency toward achievement and organization. These findings provided empirical support for the contention that the psychosocial environment of the pain patient is related to his/her perceptions, rating and overall experience of pain. In summary, research examining the reiationship between chronic pain and family functioning has led to two conclusions: families of individuals with chronic pain have a higher incidence of pain conditions than families of patients with depression or pain-free chronic illness; and chronic pain appears to have an adverse effect on the marital relationship of pain patients and their partners. Although an obvious explanation for the first conclusion is genetic vulnerability, a recent review [9] emphasizes that at present there is little support for this contention. A number of authors have suggested that a coping and stress model may have heuristic value in accounting for patterns of illness observed in family groups. In one such review [lo] it is suggested that family units develop characteristic levels of health and illness which tend to be stable over time. The authors propose that such patterns arise out of patterned regularities in interactions among family members, and between the family and the community. In support of this view, it has been demonstrated that when the unit of analyses is the family, characteristic family coping strategies can be identified [l I]. Such coping strategies were found to vary across a variety of situations, and in turn predict level of perceived stress and adaptation. One difficulty with the research described is that few investigators have explored general family functioning in families of chronic pain patients. Only one study to date [12] has examined the level of children’s emotional adjustment when one of the parents had chronic pain. The study was limited to an examination of children’s emotional adjustment when the mother had chronic pain. Results revealed that families with a mother who had chronic pain had poorer perceived family environments, and higher levels of depression and anxiety among the parents than control families. Furthermore, children in these families were also found to be more depressed than their control counterparts. While the results of this study are important, they are limited by the relatively small sample size employed. The present study adds to the findings of this investigation by employing a larger sample size and varying the methodology somewhat. The aim of the present study was two fold: (1) to compare the incidence of pain related illness and general behavioral disturbance among children of
two parent groups; a chronic headache group and an illness-free group, and (2) to compare parent ratings of family functioning across groups. In addition, the relationship between ratings of family functioning and severity of parental pain complaint was examined in the chronic headache group. A number of specific predictions were made. First, it was hypothesized that there would be a greater incidence of pain related illness, such as headaches and abdominal pain, among children having a parent with chronic headaches, than among children of illness-free parents. Second, it was expected that there would be a greater incidence of general behavioral disturbance among children of chronic headache parents than among children of illness-free parents. It was also predicted that there would be a positive correlation between parent’s symptom severity and incidence of general behavioral disturbance of the child. Third, it was predicted that among chronic headache parents, ratings of family functioning would correlate negatively with pain rating indices. Finally, it was expected that ratings of family environment would reflect more dysfunction among families of chronic headache parents when compared with families having illness-free parents. METHOD
Subject
Patients treated for chronic headache in the Pain Management Service, University Hospital, Saskatoon, were contacted by mail and asked to volunteer their participation. Subjects had to be living with a spouse who did not suffer from a chronic illness. Subjects for the control group were obtained through a general optometry practice in the city of Saskatoon. Names were reviewed by one of the optometrists such that names, addresses and phone numbers of all families with illness-free parents, and children falling in the appropriate age range were extracted. Agreement to participate was obtained by the office staff before the names of the families were released to the research team. The study was described as investigating the effects of various health conditions on general family functioning. Children of both genders between the ages of 9 and 17 were included. This was to ensure that they were old enough to understand all instructions, and young enough to be living at home. Of a total of 38 chronic pain patients contacted, 14 declined participation, resulting in a 63.2% acceptance rate. A total of 49 illness-free families were contacted. Of these 19 declined participation, resulting in a 61.2% acceptance rate. Generally, refusal to participate was due to either other commitments, distances involved in traveling to the hospital, or scheduling difficulties. No one refused to participate due to the nature of the study. Thus, it can be assumed that the samples were fairly representative of the populations from which they were drawn. Procedure
Parent-child pairs underwent individual testing sessions at the hospital, with the entire procedure taking approx. 90-120min. Children were asked to respond to a number of questions concerning their
53
Pain in children of chronic headache sufferers health history and the frequency with which they experience headaches. This information was corroborated with parents. Parents were also asked to indicate the frequency with which they experienced headaches. Additionally, parents were asked to complete the short form of the Personality Inventory for Children (PIC), the Family Environment Scale (FES), and the Symptom Checklist-90 @CL-90). Chronic pain sufferers were also asked to complete the West Haven-Yale Multidimensional Pain Inventory (WHYMPI). As parents completed these questionnaires their children completed the Comprehension and Similarities subtests of the Wechsler Intelligence Scale for Children-Revised, and the Quick Test. These tests were administered in order to ensure that children in both groups did not differ with regard to level of cognitive complexity, or intellectual development. Instruments The PIG-short f&m consists of responses to 280 items, answered ‘True’ or ‘False’ by the parent of each child [ 131.The inventory provides scores on four factor scales, a Lie scale, a Development scale, and 14 clinical scales. The FES is a go-item self-report instrument measuring the social-environmental characteristics of families. The questionnaire is composed of 10 scales assessing three underlying dimensions including: Relationship, Personal growth, and System Maintenance [I. The SCL-90, is composed of a list of 90 different symptoms [14]. Using a 5-point Likert scale, subjects are required to indicate the degree to which they have been bothered by each of the symptoms in the last month. The scale provides scores on 9 factors including somatization, obsessive-compulsiveness, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. A General Symptom Index (GSI) based on the total score is also calculated. The WHYMPI is a 52-item pain assessment inventory composed of 12 scales falling into one of three categories; the impact of pain on the patient’s life, the responses of others to the patient’s communications of pain, and the extent to which the patient participates in common daily activities [15]. The WISC-R
Subtests
The similarities and comprehension subscales of the WISC-R were administered to each of the children [ 161. The similarities is composed of a series of 17 word pairs. The child is asked to indicate the manner in which each of the words in a given pair are alike. Word pairs are presented in order of increasing difficulty. The comprehension subscale is made up of 17 questions, each reflecting increasingly complex levels of knowledge. The Quick Test
This general measure of intelligence requires matching pictures with words from a list arranged in order of increasing difficulty [17].
RESULTS Data analyses are presented in the following order: data assessing the initial equivalence of groups, preliminary analyses, and tests of the hypotheses. Tests of the hypotheses were based on l-tailed tests of significance, while initial equivalence data, preliminary analyses and post-hoc analyses were based on 2-tailed tests of significance. Initial equivalence of the two parent groups was assessed using the following variables: age; gender; employment status of the primary wage earner in the family; highest level of education attained by the primary wage earner; and number of children in the family. No significant differences were found between the two groups on any of these variables with the exception of number of previous marriages. Chronic headache patients had significantly more previous marriage (M = 0.43, SD = 0.51) than control subjects (M = 0.07, SD = 0.25), t(49) = 3.36, P c 0.002. This finding is consistent with reports in the literature pointing to the high degree of marital discord reported by chronic pain patients and their spouses (e.g. I41). Initial equivalence of the two groups of offspring was assessed using the following variables: age; gender; birth order; Quick Test IQ; WISC-R comprehension and similarities scores; and health history. A significant age difference was found, with children of pain patients having a mean age of 13.67 years (SD = 2.71), while children of controls had a mean age of 12.23 years (SD = 1.75), t(49) = 2.30, P < 0.026. No other significant differences were found on the remaining initial equivalence variables. Preliminary
analyses
In order to determine any potential effects of offspring group differences in age, this variable was correlated with the major dependent variables including children’s scores on the PIC, children’s monthly frequency of headaches, and a number of parental variables. None of these correlations was found to be significantly different from zero. Thus, age differences between the two groups of children were ignored with regard to all subsequent analyses. Parental group composition
In order to verify the chronicity of headaches in the chronic pain group the average monthly frequency of headaches of the two parent groups was compared. The mean number of headaches per month for members of the chronic pain group was 16.63 (SD = 9.17), as compared to 0.90 (SD = 1.27) for members of the control group [F(l, 53) = 86,487, P < O.OOOl]. Thus, the experimental group represented a chronic pain sample in comparison to the illness-free controls. Tests of the hypotheses Hypothesis I. It was hypothesized that children of headache sufferers would exhibit a greater incidence of pain related illness than children of illness-free controls. Group differences on two components of the PIC; the somatic concerns scale and factor 3 (internalization/somatic symptoms) were examined. As can be seen from the results presented in Table 1,
SAMWL F. MIKAILand CARL L. VON BAEYER
54
Table I. Comparison
of child illness categories by group Pain (N = 24) Control (N = 30)
Scale Frequency of headaches (per month) Factor 3 (intemalizing~somatization) Somatic concerns scale
Table 2. Comparison
Mean (SD)
Mean (SD)
I
One-tailed P
2.16 (3.96) 5.83 (5.03) 52.88 (10.96)
l.ZO(l.35) 3.90 (2.33) 47.70 (8.66)
2.01 1.87 1.95
0.03 0.03 0.03
of somatic concerns subscale scores by group Pain (N = 24) Control (N = 30)
Scale
Mean (SD)
Mean (SD)
Frequency of illness Infrequent symptoms Somatizing Visual problems Excessive constipation Poor stamina Body temperature Muscular spasms Attention seeking Other
0.42 (0.58) 1.29(1.04) 0.67(1.17) 0.25 (0.53) 0.25 (0.44) 0.88 (0.80) 0.38 (0.71) 0.46 (0.78) 0.71 (0.75) 0.42 (0.50)
0.33 (0.55) 0.90(1.35) 0.30 (0.47) 0.23 (0.50) 0.27 (0.45) 0.93 (0.87) 0.13 (0.43) 0.23 (0.43) 0.53 (0.63) 0.17 (0.38)
a significant difference was found for both of these two components. Children of chronic pain sufferers had a significantly greater degree of somatic concerns as compared to children of healthy controls. Group differences on factor 3 scores were also significant, with the patterns of means being in the same direction. For purposes of clarification the somatic concerns scale was broken down into its component subscales (see Table 2). The two groups were compared on these in order to determine specific symptom groups which accounted for the observed difference on the scale. A significant difference between group means was found on the other subscale. This subscale was made up of two items: “my child has as much pep and energy as most children” and “my child talks a lot about his (her) size or weight”. In comparison to controls, parents in the chronic headache group were more likely to endorse the first item as false and the second item as true of their child. The second step in testing this hypothesis was to compare the monthly frequency of headaches reported by the two offspring groups. Analysis revealed that children of chronic headache sufferers reported significantly more headaches per month than did children of illness-free parents (see Table 1).
, 0.54 1.17 1.57 0.12 -0.14 -0.25 1.54 1.35 0.93 2.08
Two-tailed P 0.59 0.25 0.12 0.91 0.89 0.80 0.13 0.18 0.36 0.04
Thus, children of chronic headache sufferers reported experiencing more headaches per month, appeared to be more concerned with their body image, and were reported to have less energy than other children. However, these group differences did not generalize to other pain related illnesses. Hypothesis 2. This hypothesis was divided into two parts. First, it was suggested that there would be a higher incidence of general behavioral disturbance among children of chronic pain sufferers than among children of illness-free parents. This part of the hypothesis was tested by comparing group means using the PIC clinical scales. As indicated above, children of chronic pain sufferers demonstrated higher mean scores on the somatic concerns scale than did children of healthy controls. Additionally, a significant difference was found on delinquency scores. The items on the scale present a broad range of antisocial and acting-out behaviors (see Table 4 for a listing of subscales). It was found that children of chronic headache sufferers had a significantly higher delinquency score than children of healthy controls. As can be seen in Table 3, significant differences were also found on the general adjustment and social skills scales. Specifically, children of chronic headache sufferers were reported as having higher maladjustment and
Table 3. Comparison of level of behavioral disturbance among children of chronic headache sufferers and healthy controls
Pain (N = 24) Control (N = 30) Scale General adjustment Intellectual screening Developmental screenir ‘B Somatic concerns Family relations scale Delinquency Withdrawal Anxiety Psychoticism Hyperactivity Social skills Reported means are based on T-scores.
Mean (SD)
Mean (SD)
52.65 (12.39) 50.84 (10.70) 52.46 (9.30) 52.88 (10.96) 52.37 (12.82) 53.55 (13.36) 51.13(9.85) 51.83(11.19) 49.11 (9.15) 50.17(11.43) 47.40 (8.99)
47.91(7.29) 49.33 (9.54) 48.04 (10.26) 47.70 (8.66) 48. I2 (6.69) 47.40 (5.48) 49.14(10.19) 48.54 (8.86) 50.69(10.71) 49.87 (8.89) 51.99 (10.42)
I 1.72 0.55 1.64 1.94 1.54 2.28 0.72 I .20 -0.57 0.11 -1.68
One-tailed P 0.05 0.59 0.11 0.03 0. I3 0.01 0.48 0.23 0.57 0.91 0.05
Pain in children Table 4. Comparison
of chronic
headache
sufferers
55
of mean subscale scores of the PIC delinquency scale for children of chronic headache sulTercrs and healthy controls
Scale Disregard for limits,mterpersonal Ann-social tendeno:‘s Irritability/limited tolerance Sadness Lack of interest/impulsivity Interpersonal hostility Other
insensitivity
Pain (N = 24)
Control (IV = 30)
Mean (SD)
Mean (SD)
I
Two-tailed P
1.75 (2.05) 1.30 (2.32) 0.58 (0.88) 0.29 (0.46) 1.08 (0.58) 0.08 (0.28) 1.17(0.49)
0.93 (I .39) 0.60(1.10) 0.30 (0.60) 0.03 (0.18) 0.70 (0.47) 0.00 (0.00) 0.90(0.31)
1.74 I .47 1.41 2.79 2.68 I .62 2.49
0.09 0.15 0.17 0.01 0.01 0.1 I 0.02
Reported means are based on raw scores. Table 5. Comparison
of level of psychological
symptom severity of chronic headache sufferers and healthy controls Pain (N = 24) Control (N = 30)
SCL-90 scale
Mean (SD)
Somatization Obsessive-compulsive Interpersonal communication Depression Anxiety Hostility Phobic anxiety Paranoia Psychoticism GSI
55.53 (I 1.70) 53.80(11.45) 53.80(13.07) 54.01 (12.34) 53.50(13.09) 53.57 (12.82) 54. I I (14.05) 50.66(12.19) 52.42 (13.41) 53.60 (34.10)
I
Two-tailed P
3.70 2.38 2.38 2.52 2.17 2.21 2.59 0.39 1.46 2.63
0.02 0.02 0.02 0.04 0.03 0.01 0.70 0.15 0.01
Mean (SD) 46.13 47.34 47.34 47.19 47.55 47.50 47.13 49.54 48.31 34.78
(6.35) (8.01) (6.07) (6.90) (6.26) (6.60) (3.94) (8.33) (6.41) (16.32)
0.00
Reported means are based on T-scores.
lower social skills scores than children of illness-free controls. Once again, in order to clarify these findings, post-hoc analysis were conducted in which group differences were examined by breaking down the three scales into their component parts. As can be seen from Table 4, a number of group differences were found on the subscales of the delinquency scale. Children of chronic headache sufferers exhibited significantly higher scores on the sadness subscales, lack of interest and impulsivity subscale, and the ‘other’ subscale. The ‘other’ subscale includes three items; “my child can’t sit still in school because of nervousness”, “several times my child had threatened to run away”, and “the child’s father has very little patience with the child”. No individual subscale differences were found on the general adjustment or the social skills scales. Second, it was suggested that there would be a positive correlation between measures of parent and Table 6. Significant correlations
child nsvchological symptom severity. Since the parental *groups differed -on a number of SCL-90 subscales (see Table 5), correlational analyses were conducted separately for each of the groups. Results of these analyses are presented in Tables 6 and 7. For both groups a number of moderate to strong correlations were observed between measures of parent and child psychological symptom severity. Generally, these correlations were higher for parentchild pairs in the chronic pain group. However, it must be noted that variances in SCL-90 scale scores of pain patients, and PIC scale scores of their offspring, were greater than in the case of their control counter parts. This greater variability in scores may have contributed to the higher correlations in this group. Generally, however, the data corroborate the hypothesized relationships. Hypothesis 3. A number of predictions were made regarding the relationship between ratings of family environment and parental health status. A negative
between parents’s symptom severity and incidence of general behavioral disturbance of the child in .oain _. group SCL-90 symptom scales (parent)*
PIC symptom scales (child) Intellectual screening Adjustment Development screening Somatization Family relations Withdrawal Anxiety Delinquency Hyper&ti&y Psychoticism
I
2
3
4
0.40
5
6
7
8
9
GSI
0.58
0.41
0.45
0.59 0.61
0.59 0.73
0.62 0.71
0.49
0.46 0.52
0.44 0.65
0.51 0.80
0.67 0.57
0.59
0.62 0.58 0.49 0.58
0.48 0.38 0.46 0.67
0.39
0.42
0.47
0.44
0.43
0.57
0.66
0.56
0.67
0.70
0.69
0.63
0.69
0.63
0.71
0.73
0.70
0.74
0.72 0.62
*I = Somatization, 2 = obsessive-compulsive, 3 = interpersonal sensitivity, 6 = hostility, 7 = phobic anxiety, 8 = paranoid ideation, 9 = psychoticism. Test of significance of correlations is based on a i-tailed r-test, and P < 0.05.
4 = depression,
5 = anxiety,
56
SAMUEL Table
7. Significant
correlations
F.
MIKAIL and
between disturbance
CARL L. vo~i BAE\EYER
parent’s symptom severity and incidence of the child in the control group SCL-90 symptom
PIC symptom (child)
scales (parent)*
I
2
3
4
5
6
7
West Haven-Yale Multidimensional Pain Inventory Scales Interference Pain severity Negative mood Se&on 3 Household chores Outdoor activity Activity away from home Social activity
GSI
0.34
0.40
0.31 0.4-l
0.43 0.60 0.56 0.33 0.36
-0.33
0.46 - 0.36
-0.34 -0.42
-0.34
correlation between positive ratings of family functioning and pain rating indices was predicted for chronic pain sufferers. Pain rating indices were obtained using the West Haven-Yale Multidimensional Pain Inventory. A number of scales of this questionnaire measure level of pain severity. For purposes of the present study these included scale 1 (degree to which pain interferes with daily activities), scale 3 (pain severity), scale 5 (negative mood), and all four scales of section 3 (these measure the degree to which the individual engages in a variety of activities including household chores, outdoor work, activities away from home, and social activities). Ratings of family functioning were obtained by use of the Family Environment Scale. The pattern of correlations obtained (see Table 8) suggests that the predicted relationship between pain severity and positive family functioning is not as robust and direct as expected. Specifically, not all dimensions of family environment seemed to be related to pain severity. Scale 1 of the WHYMPI achieved a significant negative correlation with level of family cohesion and active-recreational orientation of a family (r= -0.39,P < 0.05; r = -0.44,P < 0.05respectively). Thus, the greater the degree of cohesion in a family, the lower the degree to which pain is reported to interfere with daily activities. Similarly, the greater the degree to which a family is oriented toward recreational interests, the less pain interferes with daily activity. Pain severity was negatively correlated with level of expressiveness and active-recreational correlations
9
0.53 0.36 0.41
I = Somatization. 2 = obsessivesompulsive, 3 = interpersonal sensitivity, 6 = hostility, 7 = phobic anxiety, 8 = paranoid ideation, 9 = psychoticism. Test of significance of correlations is based on a l-tailed r-test, with P < 0.05.
8. Significant
8
0.38
l
I = Cohesion,
behavioral
scales
Adjustment Developmental screening Intellectual screening Somatization Family relations Delinquency Withdrawal Anxiety Psychoticism Hyperactivity Social skills
Table
of general
between
DISCUSSION
The present investigation examined the incidence of pain related illness and general behavioral disturbance among children of two parent groups.
rating of family environment headache sufferers
2
3
4
-0.39 -0.40
-0.53 0.39
5 = anxiety.
orientation, and positively correlated with level of family organization. The presence of negative mood was negatively correlated with level of family cohesion and active-recreational orientation, and positively correlated with level of expressiveness. There were almost no significant correlations between the 4 scales of section 3 of the WHYMPI and the various dimensions of family functioning. The only significant correlation achieved was a positive correlation between the extent to which an individual engages in outdoor activity and the tendency to encourage independence of family members. Collectively, these results partially support the proposed hypothesis. However, it is clear that the relationship between pain intensity, or level of disability, and family functioning is less direct than expected. A second component of the hypothesis was the prediction that ratings of family environment would reflect more dysfunction among families of chronic headache sufferers than families of healthy controls. A series of oneway analyses of variance, comparing mean scores on the various subscales of the Family Environment Scale for the two groups reveal no significant differences. Thus, this hypothesis was not substantiated by the data.
Family
I
4 = depression,
0.37
and pain
environment 5
6
rating
indices
of chronic
scales* 7 -0.44 -0.34 -0.40
8
9
IO
-0.36
0.41
2 = expressiveness, 3 = conflict. 4 = independence, 5 P achievement orientation, 6 = intellectualcultural orientation, 7 = active-recreational orientation, 8 = moral religious orientation, 9 = organization. 10 = control. Test of significance for all correlations is based on a l-tailed f-test, with P < 0.05 level of significance. l
Pain in children of chronic headache sufferers
Additionally, a number of family environment variables were examined as they related to family illness patterns. With regard to the first hypothesis, it was expected that children of chronic headache sufferers would exhibit a greater incidence of pain related illness than children of illness-free controls. There was only partial support for this hypothesis. Children of chronic headache sufferers did achieve higher scores on the somatic concerns scale and factor 3 (intemalizing/somatization) of the PIC. However, items differentiating between the two groups dealt with concerns regarding body image and lack of energy. Specifically, children of chronic headache sufferers were reported to have less pep or energy than most children, as well as being more concerned about their size or weight. In addition, children of headache sufferers reported having a higher frequency of headaches per month than did children of headachefree parents. The somatic concerns scale was developed with the intention of identifying children who tend to have an exclusively ‘neurotic adjustment’ to stressful situations [ 131. Specifically, children who achieve a high score on this scale tend to respond to stress through a heightened concern over body functioning, energy level, diffuse physical symptoms, and so on. One can argue that such a focus may develop in response to repeated illness, prolonged episodes of pain, or in this case, frequent headaches. However, it is important to note that children suffering from chronic health conditions or illness do not achieve elevated scores on this scale [ 131. Collectively, these results suggest that children of chronic headache sufferers are more concerned with their health status and bodily functioning than children of healthy controls. That is, they tend to be more somatically focused. As noted by the developers of the PIC, becoming somatically focused tends to occur in response to stress [13]. Although the present study included no measures of perceived stress, several of the reported findings are consistent with this relationship. It was found that children of chronic headache sufferers have significantly higher scores on the general adjustment and delinquency scales than did controls. When this result was examined more closely it became apparent that the components of the delinquency scale which accounted for this group difference included the sadness subscale, the lack of interest/ impulsivity subscale, and the ‘other’ subscale. Items on these subscales focused on such factors as negative or depressed mood, distractability, inability to concentrate or remain focused, and poor family relations. As can be seen, these two scales, and their component subscales, can be considered to reflect either sources of stress, or end products of stress. In summary, it would appear that children of chronic headache sufferers tend to be more somatically focused than their control counterparts, and that this heightened concern with their health status may be either a reaction to, or a means of coping with ongoing stress. A number of hypotheses were made regarding family functioning and illness behavior. First, it was suggested that a positive correlation would be
57
observed between measures of parent and child psychological symptom severity. This relationship was examined separately for each of the parentxhild groups for reasons outlined earlier. A number of significant relationships were identified. The most interesting of these was a positive correlation between parents’ somatization scores and children’s somatic concerns scores. This correlation was found to be significant in both groups (r = 0.72 in the chronic headache group and r = 0.35 in the control group). If somatization is viewed as a characteristic style of coping [13], then this finding provides tentative support for the contention that there exists characteristic family coping strategies, and that these may affect the vulnerability of family members to certain illnesses [l 11. For example, individuals having chronic headaches tend to respond to stressful situations through increased muscle contraction [18], and may be more predisposed to headaches following such periods. It is conceivable that family members may exhibit similar responses to stressful situations. Another finding of interest is the pattern of correlations between parent and child symptom severity in the two groups. In the chronic headache group positive correlations were found between all scales of the SCL-90 and several of the PIC scales (i.e. somatic concerns, family relations, anxiety, and psychoticism). No such pattern was observed in the control group. It will be recalled that with the exception of paranoid ideation and psychoticism, chronic headache sufferers have significantly higher elevations on all scales of the SCL-90, including the general severity index (GSI). In a previous investigation it was found that there was a significant correlation between pain patient and spouse GSI scores when the patient’s scores were above the mean for normative psychiatric patients [3]. Results of the present study can be considered analogous to this finding. In fact, given that normal, as opposed to psychiatric controls were used in the present study, the results suggests that emotional adjustment of children may be more influenced by levels of parental symptom severity, than the emotional adjustment of spouses. This would make sense given that a child’s capacity to deal with and filter out another family member’s behavior is likely to be less developed than that of a mature adult. One potential limitation of these results is that the parent was required to fill out both the PIC and the SCL-90. Thus, it may be argued that at least part of the observed correlation in parent and child emotional functioning is due to shared method variance. However, this is not a serious limitation as all scales of the PIC have been shown to correlate highly with more objective ratings of children’s behavior [13]. Furthermore, inter-rater agreement for the instrument as a whole has been found to be quite acceptable [13]. The prediction was made that ratings of family environment would reflect more dysfunction among families of chronic headache sufferers in comparison to families of healthy controls. This hypothesis was not substantiated by the results. The absence of an effect can be explained in at least two ways. First, it may reflect the tendency of chronic pain patients to idealize relationships both previous to and following
58
SAMUEL
F.
MIKAIL
and
CARL
L. VON BAEYER
the onset of pain. This tendency was
[ 111. In studying the nature of family environments, it is essential that multiple ratings of family functioning be obtained. If pain patients are found consistently to present a more optimistic account of their relationships than do other family members, we may need to begin paying closer attention to this potential blind spot in our treatment of these patients. It may well be that the higher incidence of marital discord in these families is due to the patient’s inability to recognize, and in turn deal with, such interpersonal difficulties. Finally, a word about the focus of research and treatment in the area of chronic pain. For the most part the efforts of researchers and clinicians alike, have focused on gaining an understanding of chronic pain from the perspective of patients and their partners. Undoubtedly, this work is important and needs to continue. However, to date, the effects of chronic pain on patients’ children has not been adequately investigated. The above results suggest that children of pain patients may differ on several dimensions from children of healthy controls. Furthermore, the data suggest that these differences may in part be related to the manner in which these children respond to their parents’ condition. Children may experience their parents’ suffering as an ongoing stressor.
Pain in children of chronic headache sufferers However, they may be brought up in an environment in which (a) they are not taught how to adequately deal with such a stressor and/or (b) they are not exposed to effective coping models. Furthermore, they are not included in any component of their parents’ treatment program. Although results of the present study do not allow for any specific suggestions as to how children could be included in their parents’ treatment program, they do suggest that future research should address this issue. authors wish to extend their gratitude to Dr Kenneth Craig, Dr James Pond, Dr Michael Sullivan and Dr Joyce D’Eon for their helpful comments. This research was carried out as part of the first author’s doctoral dissertation in psychology with support of a training fellowship from the Saskatchewan Health Research Board.
7. 8. 9. 10.
Acknowledgements-The
REFERENCES
1. Mohamad S. N., Weisz G. M. and Waring E. M. The relationship of chronic pain to depression, marital adjustment, and family dynamics. Pain 5, 285-292, 1978. 2. Violin A. and Giurgea D. Familial models for chronic pain. Puin 18, 199-203, 1984. 3. Shanfield S. B., Heilman E. M., Cope D. N. and Jones J. R. Pain and the marital relationship: psychiatric distress. Pain I, 343-351, 1979. 4. Maruta T. and Osborne D. Sexual activity in chronic pain patients. Psychosomatics 19, 531-537, 1979. 5. Maruta T., Osborne D., Swanson D. W. and Halling J. A. Chronic pain patients and spouses: marital and sexual adjustment. Mayo Clinic Proc. 56,307-310, 1981. 6. Feuerstein M., Suit S. and Houle M. Environmental
11. 12.
59
stressors and chronic low back pain: life events, family and work environment. Puin 22, 295-307, 1985. Moos R. H. and Moos B. S. Family Environment Scale: Manual. Consulting Psychologist’s Press, Palo Alto, CA, 1981. Melzack R. The McGill pain questionnaire: major properties and scoring methods. Pain 1, 275-299, 1975. Turk D. C., Flor H. and Rudy T. E. Pain and families. I. Etiology, maintenance, and psychosocial impact. Puin 30, 2-28, 1987. Coyne J. and Holroyd K. Stress, coping and illness: a transactional perspective. In Handbook of Clinical Healfh Psychology (Edited by Millon T., Green C. and Meagher k.). Plenum Press, New York, 1982. McCubbin H. I. Integrating coping behavior in family stress theory. J. Marriage & Family 41, 237-244, 1979. Dura J. R. and Beck S. R. A comparison of family functioning when mothers have chronic pain. Pain 35,
79-90, 1988. 13. Wirt R. D.. Lachar D.. Klinedinst J. K. and Seat P. D. Multidimensional Description of Child Personality: A Manual for the Personality Inventory for Children.
Western Psychological Service, Los Angeles, CA, 1984. 14. Derogatis L. R. XL-90 Adminisfration, Score and Procedures Manual. Johns Hopkins University Press, Baltimore, MD, 1977. 15. Kerns R. D., Turk D. C. and Rudy T. E. The West Haven-Yale Multidimensional Pain Inventory. Pain 23, 345-356, 1985. 16. Wechsler D. Manualfor the Wechsler Intelligence ScaleRevised. The Psychological Corporation, New York, 1974. 17. Ammons R. B. and Ammons C. H. The Quick Test (QT): provisional manual. Psychol. Rep. 11, 11I-161, 1962.
18. Traue H. C., Gottwald A., Henderson P. R. and Bakal D. A. Nonverbal expressiveness and EMG activity in tension headache sufferers and controls. J. psychosom. Res. 29, 375-381, 1985.