PATIENTS AND METHODS

Thirteen female patients with a mean (SD) age of 41.1 (6.3) years (range 30–49) who met the 1990 American College of Rheumatology (ACR)¹ criteria for the classification of FM and 10 healthy female subjects matched for age (37.9 (6.1) years, range 31–49) and education were enrolled. The patients had a mean education level of 7.5 (3.1) years (range 5–15) and the controls a mean of 7.8 (3.2) years (range 5–15). The inclusion criteria for the study group comprised a negative history for dementia, cerebrovascular disease, alcohol abuse, psychoactive drug treatment, and other neurological disorders. None of the patients had comorbid psychiatric disorders and none were receiving treatment with antidepressant drugs. The patients who had been using such drugs were only included if they had stopped using them at least three weeks before the study. All subjects gave their informed consent.

Assessing myalgic scores

The pain pressure threshold (PPT) measurements of patients with FM were performed in the same room in the early afternoon with a mechanical algometer. The same doctor carried out all of the measurements and the tests throughout the study. Before the evaluations, subjects were informed of the procedure. Pain threshold was explained as the amount of pressure adequate to induce a sensation of discomfort, and the subjects were warned that the aim was to determine the pain threshold but not pain tolerance.

Eighteen tender points (TPs) accepted by the ACR for FM¹ and three control points (CPs), which were generally agreed upon and used in several previous studies, were evaluated. The three control points were the mid-forehead, the two thirds distal portion of the dominant forearm, and the dominant thumbnail. The apparatus has a force-pressure handle connected to a rubber disk and calibrated in kg/cm². Pressure was increased at a rate of 1 kg/s, after vertically applied upon the TP, and in this course the subjects were asked to state when they felt pain. A positive TP was defined as a point at which the subject had mild or great pain with <4 kg/cm² pressure. The sum of the PPTs of 21 points (18 TPs, 3 CPs) was calculated as the total myalgic score (TMS in kg/cm²).¹⁷

The Hamilton Depression Rating Scale (HDRS)¹⁸ and Hamilton Anxiety Rating Scale (HARS)¹⁹ were used to evaluate the affective condition of patients with FM. The cut off score for mild depression on the HDRS was 27.

Auditory P300 event related potentials

P3 ERPs were recorded with Ag/AgCl electrodes (impedance was held below 5 kOhm) on a Dantec Keypoint (Medtronic, Denmark) device by the same doctor who was unaware of the subjects’ clinical data, in an electromagnetically isolated, sound attenuating room. Subjects lay down on an adjustable examination table with a semitilted headpiece and were allowed to relax. During recordings, subjects were instructed to minimise blinking and to fix their eyes on the wall to reduce eye movements as much as possible. The active electrode was placed at Cz (vertex) and referenced to the linked earlobe according to the international 10–20 system. P3 evoked potentials were generated after a binaurally presented tone discrimination paradigm through a headphone with frequent (85%) tones of 1000 Hz and rare oddball stimuli (15%) of 2000 Hz at 80 dB. Subjects were instructed to count rare stimuli—target tones at 2000 Hz—and report at the end of the session. When there was a discrepancy of >10% hits between the actual number of stimuli and the number reported by the subjects, recordings were repeated. Frequency limits were 0.1–50 Hz and analysis time was a total of one second, including the 100 ms baseline pre-stimulus. Latencies (ms) of the P3 peak and amplitudes (μV) of P3 were recorded.

Patients were administered the Turkish version of the Fibromyalgia Impact Questionnaire (FIQ)²⁰ after measurements of the ERPs, PPTs, HDRS, HARS at entry and in the fourth and eighth weeks of treatment (sertraline 100 mg daily).

Statistics

The data were analysed on a personal computer using the statistical package for social sciences (SPSS) software. Mann-Whitney U test was used for intergroup comparisons. Patients’ consecutive measurements were compared using the Wilcoxon test. Values were correlated with Spearman rank correlation coefficients. A two tailed p<0.05 was considered to be significant.

RESULTS

The entry measurements of P3 latencies of patients were not significantly different, whereas P3 amplitudes were significantly lower than those of the controls (table 1). Table 2 summarises three follow up measurements of TMS, FIQ-5, FIQ-6, HDRS, HARS and P300 potentials of the patients. A significant improvement was found in TMS, FIQ-5, FIQ-6, HDRS, HARS, and P3 amplitudes between their measurements at entry and at the end of eighth weeks of treatment (table 2). P3 latency did not change significantly with treatment, but amplitudes significantly improved after an eight week course of treatment with sertraline, almost reaching those of the controls (fig 1).

Relationship of measurements to P3 potentials at study entry

Significant correlation was found between P3 potentials and TMS (latency, r_s=−0.79, p<0.01; amplitude, r_s=0.53, p<0.05). Nevertheless, P3 potentials did not correlate with FIQ-5 (pain-visual analogue scale (VAS) scores), and FIQ-6 (fatigue-VAS scores), or HARS and HDRS.

DISCUSSION

Patients with FM usually complain of decline in their everyday cognitive functions. They often have trouble in remembering things and are unable to concentrate on demanding tasks; this sometimes is more disturbing and disabling than their pain.^8,9 Most patients are often unable to plan and execute daily work as they feel that they are in a “daze” or “mental fog”—so called “fibrofog”—in which connecting scenes to each other is difficult.^21,22 They complain of failing even in simple daily tasks, such as cooking, driving, or shopping without missing items on their shopping list, and complain of spending, consequently, much more time on tasks because of forgetfulness.

The P3 component of ERPs is a repeatable, relatively inexpensive, and useful method for the assessment of cognitive ability in normal subjects as well as in patients with neuropsychiatric disorders.^11,12,23 This method of evaluating cognitive functions is reliable from test to test, and its variability is highly comparable and sometimes better than routinely employed biomedical assays like cholesterol, glucose, or haemoglobin measurements.^12,23,24 P3 measurements have some advantages over neuropsychometric tests, in that they are less prone to practice variations and can be performed in a “blinded” manner.²⁵ However, the usefulness of P3 has been restricted because of contributing environmental factors or “biological determinants” or lack of standardised measurement protocols between laboratories.^11,12,26 For example, amplitudes may be influenced by age of the subject, time of day, season of the year, or recently ingested food and personality type of the subject, whereas latency may be influenced by age, heart rate, and body temperature.^27–29 Some reports suggest that these effects do not basically restrict the clinical utility of P3 and greater sensitivity could be gained by taking into account these contributing “biological determinants” in the research environment.^12,27,30,31

Park et al assessed cognitive functions by information processing, recognition memory, working memory function, free recall, verbal fluency, and vocabulary in patients with FM in comparison with age and education matched controls and education matched older controls.⁸ Patients with FM had a poorer performance in all these measures, with the exception of processing speed, than the controls matched for age and education. Patients with FM also had poorer vocabulary than older controls. Additionally, impaired cognitive performance in patients with FM has been found to correlate with pain—measured with Arthritis Impact Measurement Scales—but not with depression or anxiety scores. Similar results have been previously reported by Grace and coworkers, who found an intact speed of processing but decreased working memory and long term memory that correlated with pain scales.⁷

We found no difference in P3 latency, but reduced amplitudes in patients with FM with respect to controls matched for age and education. Significant correlation was also found between P3 potentials and TMS but not with HARS or HDRS. These results should be considered together with previous results^6–8 because it has been suggested that P3 latency reflects information processing speed and P3 amplitudes express memory functions more generally.^11,12

Gervais and coworkers suggested that incomplete effort and potential exaggeration of cognitive deficits had a role in the assessment of patients with FM, particularly those who claimed medicolegal benefits.³² It is crucial to be aware of response bias during the assessment of memory impairment in FM, especially when there is a disability claim, and study groups should consist of subjects without medicolegal incentives. Similar suspicions about cognitive testing in FM have been put forward by Leavitt and colleagues.²¹ Assessment of P3 potentials has the advantage of being free of possible exaggeration and response bias, but does not give as much detailed information about cognitive functions—that is, verbal fluency, verbal knowledge, or vocabulary, as neuropsychiatric tests.

Another important result of our study was the improvement in P3 amplitudes, but without any change in P3 latencies, as sertraline treatment was given. Previous limited data point out that sertraline treatment had no effect on cognitive function tests in a variety of clinical conditions other than FM.^33,34 However, Sanz and coworkers obtained increased P3 amplitudes and unchanged P3 latencies with SRI treatment in obsessive-compulsive disorder.³⁵ Serotonin (5-hydroxytryptamine) modulates brain electrophysiological activity³⁶ and, recently, the effect of acute tryptophan depletion on auditory ERPs has been investigated in bipolar disorders.³⁷ Young et al showed that acute tryptophan depletion caused reduced amplitudes of the N1P2 and P3 components of the auditory evoked potentials in bipolar patients, especially in the frontal and central scalp areas.³⁷ In our study design it is difficult to estimate the real mechanism of action of sertraline on P3 potentials because significant clinical improvement on pain, fatigue, or depression scales was also achieved with the treatment. It is unclear whether the effect of sertraline on P3 potentials was related to its serotoninergic activity in the central nervous system or to the improvement in clinical parameters, or whether it could solely be attributed to its placebo effect.

In conclusion, our results showed that P3 amplitudes were reduced in patients with FM in comparison with controls matched for age, sex, and education. We emphasise that these data should be followed up in further studies with a broader series of patients with FM—assessing the relationship of P3 ERPs and neuropsychiatric tests in an on line or off line design to ascertain the clinical relevance of P3 potentials in FM.