Decline of substance P levels after stress management with cognitive behaviour therapy in women with the fibromyalgia syndrome

2.1 Study population

A detailed description of the study population and the stress management CBT intervention program has been given elsewhere [15]. Briefly, the study population was recruited in a municipality in central Sweden having 22,000 residents. Recruitment was done via, advertising in the local newspaper and through information meetings with the local branch of the Fibromyalgia Patient Association. Fifty-four potential female subjects were assessed for FMS with the 1990 American College of Rheumatology (ACR) criteria [17]. A physician (BK) manually assessed the tender points required for diagnosis using a fingertip pressure of 40 N. Information was sought on generalised pain duration as well as time since the FMS diagnosis. A physiotherapist experienced in FMS tender point assessment validated the diagnostic procedure.

After these procedures, six patients were excluded: two did not fulfil the diagnostic criteria, two had a serious mental disorder, and two declined participation after receiving further information about the study. The remaining 48 women agreed to participate and were allocated into two groups, group 1 (n=24) and group 2 (n=24) using a random block design. They constituted the study population. The randomization was performed with the SAS function “ranuni” that produces random numbers with equal distribution, i.e. all numbers appear with the same probability. According to this design for every four consecutive patients two were randomly allocated to group 1 and the remaining two were allocated to group 2. The allocations were indicated on paper sheets and put in sealed envelopes with a patient serial number on the outside. The envelopes were stored with the study monitor. When patients were included in the study they were given a serial number, the corresponding serial number envelope was opened and the patient allocation was noted in the study chart. One of the authors (KS) generated the random allocation sequence, a nurse at the investigating centre enrolled patients and one of the authors (BK) assigned participants to intervention.

2.2 Study design

Two analytical designs were used. The first design was a classical two-arm parallel group randomised controlled trial (RCT) with group 1 serving as the intervention group and group 2 as the control group (Fig. 1). The intervention group received CBT treatment for 6 months while the control group was waitlisted during the same time. Due to the study design it was obvious for the subjects as well as the staff what group the subjects were allocated to. Therefore no blinding was applied. When the RCT was concluded, group 2 received the same CBT treatment as group 1 had been given. Both group 1 and group 2 were followed up with 1 year after the start of CBT treatment. A second design, before-and-after treatment, was then possible for the two groups together (n=48), before CBT, after CBT and after a follow-up period of 6 months (1 year after the start of CBT treatment) (Fig. 1). The purpose of the two designs was to use data more efficiently and to crosscheck the results, a standard procedure in RCTs with small study populations.

Fig. 1:

Study population flow chart.

2.3 Data collection

The subjects were included in the study from November 2001 to May 2003. The last follow up was performed in November 2003. Data were collected on three occasions in group 1: at randomization, but before CBT treatment (at 0 months), after the CBT treatment period (at 6 months) and after the follow-up period (at 12 months). Data were collected on four occasions in group 2: at randomization, but before the waitlist period (at 0 months), after the waitlist period but before CBT treatment (at 6 months), after the CBT treatment period (at 12 months) and after the follow-up period (18 months) (Fig. 1).

At the baseline assessment before randomization, an extensive history was taken, including medical history and questions regarding mandatory education (yes or no), marriage status (yes or no), employed in gainful work (yes or no) and smoking habits (yes or no). At baseline and at each follow-up examination, information on anthropometric data, including height, weight, waist and hip circumference, co-morbidity, current medication, and other treatments was obtained. At the baseline assessment, information was also sought on menstrual status (i.e. whether the participant was pre-menopausal, menopausal or post-menopausal), and whether the patients were using hormone replacement therapy. For pre-menopausal women, all measurements were performed 9–14 days after the end of the last menstrual period to avoid interference regarding endocrine variables. Data collections were always performed between 8.00 and 10.00 a.m. Tender points were re-evaluated at the last examination.

Data from four psychometric questionnaires were obtained at each measurement occasion. The West Haven-Yale Multidimensional Pain Inventory (MPI) [18] is a self-report instrument for the comprehensive assessment of individuals with chronic pain. The instrument has been translated and tested in many languages, including Swedish (MPI-S) [19]. In this study, the psycho-social dimensions (including “life control”, “support from spouses or significant others”, “affective distress”, “interference” and “pain severity”) were used. Responses were given on seven-point scales (range 0–6).

The Maastricht Questionnaire measured “vital exhaustion” (fatigue) [20], [21]. This instrument is extensively used to evaluate treatment effects, for instance, in Swedish studies on cardiac disease [22], [23], [24]. Responses to the 19 items were given on three-point scales (0–2), with a total range of 0–38 and, high scores indicating a high degree of vital exhaustion.

The Everyday Life Stress instrument was used to assess the level of self-rated “stress behavior” [25]. Responses to the 20 items were given on four-point scales (range 0–3). On this scale, higher scores indicated more stressful reactions.

The Montgomery-Åsberg Depression Rating Scale – Self Reported (MADRS-S) is a validated instrument for evaluating treatment effects on depression symptoms. Responses were given on seven-point scales (range 0–6) [26], [27], [28].

Venous blood for the SP analysis was sampled at baseline and at each follow-up examination. The blood was collected in chilled EDTA tubes and was immediately centrifuged at +4 °C, and kept frozen at −70 °C before processing. The analyses were performed using a radioimmunoassay (RIA), based on the charcoal adsorption technique, as previously described [29]. All samples were saved and frozen, and all samples were analysed at the same time to avoid differences in the analytical procedure.

The patients’ local physicians were informed about the study and were responsible for the everyday care of the patients. There were no restrictions on changing medication or other treatment modalities. However, all such changes were documented in the study protocol.

2.4 CBT treatment program

The intervention has been described in detail elsewhere [15], [30]. Briefly, it consisted of 20 CBT treatment sessions during a 6-month period. The women were allocated into treatment subgroups with five to seven women in each. Each subgroup received one 3-h group session per week. When the 20 sessions were completed, three equally long CBT group booster sessions were given, within the next 6 months.

CBT treatment components included knowledge, self-monitoring, behavioural exercises, cognitive restructuring and life value issues. Homework assignments were given between sessions and included self-monitoring using simple diaries, as well as a booklet with behavioural and cognitive exercises. The patients learned applied relaxation techniques for active use as a coping mechanism for daily life situations [31]. The overall goal of the CBT treatment was to develop emotional, behavioural, and cognitive coping strategies for dealing with stress and pain.

Two psychologists trained in CBT provided the treatment. Continuous supervision by one of the co-authors (GB) was given to assure the quality of the therapy. Visual analogue scale estimates of pain intensity, stress, and wellbeing were made in or as close as possible to the actual therapy session to monitor the therapeutic process.

2.5 Statistical analysis

The statistical analyses were conducted using the Statistical Analysis System (SAS) software, version 9.3 [32]. Summary statistics, such as means, were computed using standard methods. The primary outcome variable was SP levels. Psychometric variables were used as covariates to adjust for individual and group differences.

In group 1, one person, and in group 2, three persons did not start CBT. Data were obtained from these persons as well, using the-intention-to treat approach. Analyses of plasma SP were missing at 12 occasions (of 168 occasions, 7%). Eight times due to laboratory failure and at four times blood samples were not obtained. Consecutive inclusion of subjects and follow up during the 2 years in different seasons was used to minimise the effects of possible seasonal variation in exposure variables and outcome.

The analyses were performed in order to clarify whether plasma SP was affected by CBT treatment and were conducted using the linear regression technique. We used linear regression technique throughout the analyses for two reasons. First, the study population was fairly small, and because of that initial SP differences had to be taken into account, which means that some form of multivariate analysis model had to be used. We chose a regression model.

In both the RCT design analysis and the before-and-after treatment design, the analysis outcome (plasma SP level) was entered as a dependent variable. In the RCT design analysis, group 1 (CBT treatment intervention group) and group 2 (waitlist control group) were included as independent variables. In the before-and-after treatment design analysis, time was added as an independent variable.

The problem with the initial plasma SP differences between the groups was handled in the RCT design by entering the initial plasma SP level as an independent variable in the analysis. In the before-and-after-treatment analyses, no such adjustment was necessary since each patient was her own control.

Analyses of possible co-variations between plasma SP and clinical behavioural correlates were performed, with plasma SP as a dependent variable and the possible psychometric variables as independent variables, one at a time. All tests were two-tailed. The level of significance was set at p<0.05.

3.1 Characteristics of the study population

Characteristics of the study population are shown in Table 1. The patients had, on average, a pain history of more than 10 years. In addition, they were, on average, more than 5 years post-fibromyalgia diagnosis, and had an average of 16 out of 18 diagnostic tender-points. There were no significant differences between the two groups in these variables at baseline. The crude mean (±SD) plasma SP levels in group 1 and group 2 were 9.09 (4.96) fmol/mL and 8.29 (4.96) fmol/mL, respectively.

3.2 Outcome in the RCT design

After adjustment for initial differences, in group 1, the intervention group’s mean plasma SP level was reduced by 39%, (Fig. 2). In group 2, the waitlist group, the mean plasma SP level was reduced by 5%. The difference in the trend was statistically significant (p=0.02).

Fig. 2:

Randomised controlled trial design with 6 months of cognitive behaviour therapy intervention. Substance P levels before intervention and after intervention in group 1 (n=24) and group 2 (n=24).

3.3 Outcome in the before-and-after treatment design

In the whole study population, group 1 and group 2 together (n=48), the plasma SP level (±SD) was reduced by 33%, from 8.32 (5.49) fmol/mL before the CBT treatment to 5.54 fmol/mL (2.15) after the 6-month CBT treatment period (p<0.01). At follow-up 1 year after the start of the CBT treatment the plasma SP was 6.98 (4.96) fmol/mL (p=0.23) as compared to the start of CBT treatment. When group 1 and group 2 were analysed separately, the same pattern of having a reduction in plasma SP levels was shown after CBT treatment. At follow-up 1 year after the start of CBT treatment, the reduction of plasma SP levels was no longer significant in either of the groups (Fig. 3).

Fig. 3:

Before-and-after design with 6 months cognitive behaviour therapy (CBT) intervention. Substance P levels in group 1 (n=24) at 0 months (before CBT), at 6 months after CBT, and at 12 months after start of CBT, and for group 2 (n=24) at corresponding times with start of CBT at 6 months from baseline.

3.4 Associations between plasma SP and psychometric variables

The course of the plasma SP levels was significantly and consistently associated with the course of “support from spouses or significant others” – no matter which of the two design analyses were used (Table 2). Plasma SP levels were also significantly associated with “life control” in the RCT design and after 6 months of CBT treatment in the before-and-after design. No other significant associations were observed before CBT treatment compared with after CBT treatment at 6 months and before CBT treatment compared with follow-up 1 year after the start of the CBT treatment.