This study reports on outcomes included as exploratory endpoints in the FINAL trial, which investigated whether 12 weeks of treatment with naltrexone 6 mg was superior to placebo in reducing average pain intensity in 99 women with fibromyalgia [12]. The FINAL study was a randomized, double-blind, placebo-controlled superiority trial conducted at Pain Center South, Odense University Hospital, a tertiary pain rehabilitation center in Denmark. In Denmark, about 95% of the patients diagnosed with fibromyalgia are women, which was the rationale for recruiting only women in this study [1]. The study received approval from The Ethical Committee of Southern Denmark (S-20190133) and was registered in The European Union Drug Regulation Authorities Clinical Trials Database (EudraCT-no. 2019-000702-30). The study protocol was uploaded to clinicaltrials.gov (NCT04270877) before the commencement of the study. A comprehensive protocol paper was published before the completion of participant recruitment [7].

Participants were recruited from the study site and through advertisements in National Patient Association magazines, both in print and online. To be eligible for participation, individuals had to meet the following criteria: they had to be women aged 18–64 years diagnosed with fibromyalgia according to the American College of Rheumatology 1990 criteria and the 2016 diagnostic criteria for fibromyalgia [30, 31]. In addition, participants were required to have at least moderate pain intensity, defined as an average pain score of at least 4 on a 0–10 numeric rating scale (NRS) over the past week [16]. Participants could continue their regular care and pain medication if these treatments remained stable. However, individuals who had used opioids within 4 weeks before the trial were excluded owing to the potential interaction between opioids and naltrexone. The use of opioids during the trial was considered a violation of the protocol. Furthermore, participants were not permitted to use antiinflammatory medication or nonsteroidal antiinflammatory drugs (NSAIDs) 4 weeks before and during the trial, as one of the exploratory outcomes involved assessing inflammation biomarkers. Information about concomitant pain medication was reported with the main results [12].

Using a 1:1 group allocation, participants were randomly assigned to receive either 6 mg of naltrexone or a placebo medication. The randomization process was conducted using a computerized algorithm, with no stratifications applied. A data manager without clinical involvement in the study created a sequential randomization list using permuted blocks of 2–6 individuals to ensure allocation concealment. The allocation information was stored in a password-protected computer file accessible solely by the data manager. The primary investigator enrolled the participants and assigned them sequential randomization numbers, allocating them to one of the two treatment groups. Participants, investigators, and outcome assessors were all blinded regarding the allocation and the permuted blocking strategy [12].

Glostrup Pharmacy, an independent compounding pharmacy in Denmark, manufactured tablets containing 1.5 mg of naltrexone and visually indistinguishable placebo tablets. The trial medication was then sent to Hospital Pharmacy Funen, along with a copy of the randomization list. The pharmacy blinded the medicine by using identical containers labeled with randomization numbers. The study duration was 16 weeks, which included a 12-week treatment period (including a 4-week titration phase) followed by a 4-week washout period (weeks 13–16). All participants initially started with a daily oral dose of one tablet of either 1.5 mg LDN or placebo. Over the 4-week titration phase, the dose was increased by one tablet each week, reaching four tablets by week 4. The dose escalation was determined on the basis of safety and tolerability, allowing for delayed increments in case of unacceptable side effects. After week 4, a maintenance dose was established, corresponding to the highest tolerated dose. The trial medication was taken once daily in the evening.

Cuff pressure pain tolerance (cPTT), TSP, and CPM were assessed using a computer-controlled cuff algometry system (CPAR, NociTech Denmark). These validated tests were used to assess spinal and supraspinal pain mechanisms [14]. Two 13 cm wide cuffs were positioned around the left and right gastrocnemius muscles to measure cPTT on both legs, approximately 8 cm distal to the apex patella. During the tests, cuffs applied pressure at a rate of 1 kPa/s, with a maximum pressure limit of 100 kPa. The test was performed on the left leg first and subsequently on the right leg. Participants were instructed to use an electronic visual analog scale (VAS) ranging from “no pain” (0 cm) to “worst pain imaginable” (10 cm) to continuously rate the intensity of induced pressure pain from its onset. When the participant stopped the test by pressing a button, the pressure level was recorded as the cPTT. When participants pressed the button, the cuffs were immediately deflated.

To assess the TSP, the cuff system was programmed to administer a series of ten consecutive stimulations to the participant’s left leg. The pressure applied for each stimulation corresponded to the individual’s pain tolerance score of the left leg, delivered at a frequency of 1 Hz (i.e., 1 s of inflation to the target pressure followed by 1 s of deflation to 5 kPa to ensure the cuff did not move). Before initiating the test, participants received instructions on utilizing the VAS to continuously rate the perceived pain intensity throughout the repeated stimulations, noting any changes over time. The average pain intensity scores for the first four and last three stimulations were calculated, and TSP was determined as the difference between these two values. A score above 0 reflected increased pain ratings during repeated stimulations, while a score below 0 reflected decreased pain ratings.

To assess CPM, the same procedure as described for pain tolerance was repeated on the left leg, followed by an identical pain tolerance test with a conditioning stimulus (CS) applied in parallel to the right leg. The pressure of the CS was set to correspond to the pressure for VAS = 5 during the assessment of cPTT of the right leg. The CS was maintained consistently on the right leg. Meanwhile, the test cuff on the left leg gradually increased pressure incremented at a rate of 1 kPa per s, starting from 0 kPa. Throughout the process, participants continuously rated the intensity of pain experienced in the left leg using the VAS. The CPM score was determined as the absolute difference in cPTT during stimulations with and without CS. A score above 0 reflected increased pain tolerance during conditioned stimulation (hypoalgesia), while a score below 0 reflected decreased pain tolerance.

The 30-s chair stand test, which has shown good feasibility and reliability in women with fibromyalgia, was applied to assess muscular endurance in the lower extremities [8, 19]. This test has been suggested as a tool for monitoring fibromyalgia [3, 8]. The test began seated, with the participant securing contact with the chair’s backrest. Initially, the participant was asked to rise to a full stand, with arms crossed over the chest, and to sit down again to familiarize themselves with the test. Subsequently, this sit-stand-sit cycle had to be repeated as many times as possible in 30 s. The count of completed cycles defined the test result.

Muscular endurance in the upper extremities was assessed using an isometric exhaustion test of the deltoid muscle [4]. The participant was seated without a backrest and with full contact of both feet to the ground. The test position was a 90° shoulder abduction with an extended arm and the palm facing downwards. To assess the time (s) to maximal exhaustion, the participant was asked to hold the position for as long as possible. This test result was defined as “arm hold time”, i.e., when the participant could no longer keep the arm position.

The same observer (a trained study nurse blinded to treatment allocation) assessed all exploratory outcomes at baseline and at the 12-week follow-up.

The main analyses were conducted on the basis of the “complete case” population to explore the potential impact of sophisticated pain and functional markers rather than to assess the intervention’s effectiveness. All continuous endpoints were analyzed using an analysis of covariance (ANCOVA) model, with randomized treatment as a factor and the baseline value of the endpoint as a covariate. Least squares means were predicted to assess the effects of LDN compared with placebo on five selected exploratory secondary outcomes (https://clinicaltrials.gov/study/NCT04270877). The change from baseline to 12 weeks was used as the dependent variable in each analysis, with baseline values included as covariates to adjust for initial differences [29]. This method provided a more precise estimation of the treatment effect by accounting for baseline variability in outcome measures. In addition to the main comparisons between the two treatment groups, two sensitivity analyses were conducted to explore potential differences in outcomes according to clinical pain reduction (30% pain responders) versus no reduction (not 30% pain responders) within LDN and placebo groups, separately. All estimations and hypothesis tests were conducted using a two-sided significance level of 0.05. Results are presented as least squares means for each group, along with between-group differences and corresponding 95% confidence intervals. Statistical analyses were performed using IBM SPSS Statistics version 28.