ABSTRACT
Background and Aims
Chronic low back pain (CLBP) with radiculopathy is a disabling, predominantly neuropathic condition that imposes a heavy burden in low‐ and middle‐income countries (LMICs). Combination pharmacotherapy targeting multiple pain pathways is increasingly used, yet head‐to‐head comparisons specifically in radicular CLBP are scarce. This prospective study compared pregabalin combined with duloxetine versus pregabalin combined with amitriptyline for pain reduction and functional improvement.
Methods
This study was conducted between July 2024 and December 2025, during which 76 adults with clinically diagnosed CLBP and radiculopathy were enrolled at Dhaka Medical College Hospital, Bangladesh. Patients were assigned non‐randomly, in the order of enrollment, to Group I (pregabalin with duloxetine) or Group II (pregabalin with amitriptyline), with 38 patients in each group. Pain and disability were assessed using the Numeric Pain Rating Scale (NPRS) and the Oswestry Disability Index (ODI) at baseline and at 4, 8, 12, and 24 weeks. Analyses used t tests, Mann–Whitney U tests, baseline‐adjusted ANCOVA, responder analyses, and linear mixed‐effects models, reporting effect sizes and 95% confidence intervals.
Results
Both groups improved substantially over time. However, the groups were markedly imbalanced at baseline (NPRS and ODI, both p < 0.001; standardized mean difference ≈0.9); the baseline‐adjusted ANCOVA was therefore the primary analysis, and after adjustment for baseline severity the between‐group differences were no longer statistically significant for either NPRS or ODI (all ANCOVA p > 0.40). In unadjusted, secondary analyses the crude reductions were larger with duloxetine (Week‐24 NPRS between‐group difference −1.54, 95% CI −2.72 to −0.37, Cohen's d = 0.60; ≥ 50% pain reduction 60.5% [23/38] vs. 34.2% [13/38], p = 0.04), but these crude comparisons are confounded by the baseline imbalance and are regarded as exploratory. Sedation (7.9% [3/38] vs. 31.6% [12/38], p = 0.02) and drowsiness (13.2% [5/38] vs. 39.5% [15/38], p = 0.02) were less frequent with duloxetine, although adverse‐event ascertainment was open‐label and incomplete.
Conclusion
In this non‐randomized cohort, the pregabalin–duloxetine combination was associated with greater crude improvement, and duloxetine‐treated participants reported fewer episodes of sedation and drowsiness; however, the apparent efficacy advantage was substantially attenuated after accounting for baseline imbalance, and because adverse‐event ascertainment was open‐label and incomplete, the tolerability signal also remains exploratory. These findings should be regarded as hypothesis‐generating and require confirmation in adequately powered randomized trials.
Keywords: amitriptyline, chronic low back pain, combination therapy, duloxetine, functional disability, non‐randomized comparative study, pregabalin, radiculopathy
1. Introduction
Chronic low back pain (CLBP) remains one of the leading global public health challenges, imposing a substantial burden on individuals and healthcare systems. According to the Global Burden of Disease (GBD) Study 2021, low back pain affected 619 million individuals worldwide in 2020 and is projected to reach 843 million by 2050, representing a 36.4% increase driven by population aging and growth [1]. It has consistently ranked as the leading cause of years lived with disability (YLDs) since 1990, with a global age‐standardized YLD rate of 832 per 100,000 in 2020 [1]. The total disability‐adjusted life years (DALYs) attributable to low back pain increased from 43.4 million in 1990 to 70.2 million in 2021 [2]. In Bangladesh, CLBP is a major contributor to musculoskeletal disability, with a weighted prevalence of 18.5% among adults, increasing to 27.8% in older populations [3, 4]. The resulting economic burden, including healthcare costs and productivity loss, is particularly pronounced in low‐ and middle‐income countries (LMICs).
Among its clinical subtypes, CLBP with radiculopathy represents a more severe and disabling condition characterized by nerve root compression or inflammation, most commonly at the L4–L5 or L5–S1 levels [5]. The underlying pathophysiology involves neuroinflammation, ischemia, and central sensitization, leading to persistent neuropathic pain [5]. Clinically, patients present with radiating leg pain, sensory disturbances, and occasionally motor deficits [6]. The lifetime prevalence of lumbar radicular pain ranges from 13% to 40%, with an annual incidence of disc‐related sciatica of approximately 2.2% [6]. Compared with nonspecific CLBP, radiculopathy is associated with greater pain intensity, functional impairment, and reduced quality of life [7]. This distinction is important because much of the pharmacological evidence for antidepressants and anticonvulsants derives from nonspecific CLBP or mixed neuropathic‐pain populations rather than from radicular CLBP specifically; findings from these broader populations should therefore be extrapolated to radiculopathy with caution.
The pharmacological management of CLBP with radiculopathy remains challenging. Current guidelines advocate a multimodal approach incorporating nonsteroidal anti‐inflammatory drugs (NSAIDs), anticonvulsants, and antidepressants [8]. While NSAIDs provide short‐term relief, their long‐term use is limited by gastrointestinal, cardiovascular, and renal risks [8]. Opioids are discouraged due to their limited efficacy and significant adverse effects, including dependence and overdose [9]. Moreover, fewer than half of patients achieve meaningful pain relief with monotherapy, highlighting the need for more effective treatment strategies [10].
Pregabalin, a gamma‐aminobutyric acid analog, acts by binding to the α2δ subunit of voltage‐gated calcium channels, thereby reducing excitatory neurotransmitter release and attenuating central sensitization [11]. It has demonstrated efficacy in treating neuropathic pain, including lumbar radiculopathy, with a recent meta‐analysis showing significant short‐term improvements in pain scores [12]. However, its effectiveness as monotherapy is limited, and a large randomized trial found pregabalin no more effective than placebo for sciatica and dose‐limiting adverse effects often restrict its use [11, 13].
Duloxetine, a serotonin–norepinephrine reuptake inhibitor (SNRI), enhances descending inhibitory pain pathways and significantly improves both pain and functional outcomes in patients with CLBP [14]. A crossover trial demonstrated a 32% reduction in pain scores in patients with neuropathic CLBP [15]. Beyond CLBP, duloxetine has established analgesic and functional benefits across a range of chronic pain pathologies, including diabetic peripheral neuropathy, fibromyalgia, osteoarthritis, and chemotherapy‐induced neuropathy, and recent randomized evidence indicates that adding a home‐based postural exercise program to duloxetine further improves sleep quality, physical quality of life, and trunk mobility in fibromyalgia [16]. Amitriptyline, a tricyclic antidepressant (TCA), exerts analgesic effects through similar mechanisms but is associated with greater anticholinergic and sedative side effects [17]. The evidence suggests that duloxetine provides more consistent short‐term pain relief than amitriptyline does, with a more favorable tolerability profile [17, 18].
Combination therapy targeting multiple pain pathways has emerged as a rational strategy for the management of neuropathic pain. The OPTION‐DM trial demonstrated that combining pregabalin with antidepressants resulted in superior pain reduction compared with monotherapy, supporting the concept of multimodal analgesia [19]. However, despite this evidence, direct comparative studies evaluating pregabalin combined with duloxetine versus pregabalin combined with amitriptyline in CLBP patients with radiculopathy remain scarce, particularly in LMIC settings [19, 20].
This gap is especially relevant in Bangladesh, where the burden of low back pain is high [3, 4], but existing local studies have examined prevalence and risk factors rather than comparative treatment effectiveness [20]. To our knowledge, no previous head‐to‐head study has directly compared these two pregabalin‐based combinations in patients with radicular CLBP; the present study was therefore designed to provide preliminary, real‐world comparative evidence rather than definitive efficacy estimates. Therefore, this study aimed to compare the effectiveness of pregabalin combined with duloxetine versus pregabalin combined with amitriptyline in reducing pain and improving functional outcomes among patients with chronic low back pain with radiculopathy.
2. Methods
2.1. Study Design and Setting
This prospective, non‐randomized comparative study (a quasi‐experimental design in which the two treatment regimens were allocated to consecutively enrolled patients by alternating order of enrollment) was conducted to evaluate the comparative effectiveness of two pharmacological regimens in patients with chronic low back pain with radiculopathy. The study was therefore not a randomized controlled trial. Although the reviewers suggested describing the study as a comparative cohort study, treatment was investigator‐allocated (by alternating order of enrollment) rather than observed under routine care, so the study is most accurately classified as a non‐randomized, open‐label, quasi‐experimental comparative study and is interpreted accordingly. Both regimens are established, guideline‐endorsed treatment options that were prescribed as part of routine clinical care; the study compared their effectiveness rather than testing an experimental agent. This investigator‐initiated academic study was not prospectively registered in a clinical‐trials registry. We acknowledge this openly as a limitation, note that the absence of an experimental intervention and the use of two standard‐of‐care regimens reduce (but do not eliminate) the applicability of trial‐registration requirements, and confirm that retrospective registration can be undertaken if required by the editor. All 76 enrolled patients received their allocated regimen, and none were excluded after enrollment or lost to follow‐up. The study was carried out at the Department of Pharmacology & Therapeutics, Dhaka Medical College, in collaboration with the Department of Physical Medicine and Rehabilitation, Dhaka Medical College Hospital (DMCH), Dhaka, Bangladesh. The study was conducted between July 2024 and December 2025. Patients were recruited from the outpatient department of physical medicine and rehabilitation.
2.2. Study Population and Sampling
The study population comprised adult patients who were diagnosed with chronic low back pain with radiculopathy and who presented at DMCH during the study period. Eligible participants were aged between 18 and 65 years, had a duration of low back pain of at least 3 months, and reported a pain intensity of ≥ 4 on the Numeric Pain Rating Scale (NPRS). Radiculopathy was diagnosed clinically using predefined criteria, namely radiating pain in a dermatomal distribution below the knee together with at least one corroborating neurological sign—a positive straight‐leg‐raise (Lasègue) test, or a segmental sensory, motor, or reflex deficit consistent with the involved nerve root. Magnetic resonance imaging was used to support the diagnosis where available; however, because the study was conducted in a resource‐limited public hospital, imaging was not obtained for every participant, and diagnosis therefore rested primarily on standardized clinical assessment. Patients with known hypersensitivity to study drugs; uncontrolled diabetes mellitus (HbA1c > 7%); significant cardiovascular or cerebrovascular disease; recent steroid use; pregnancy or lactation; or prior use of pregabalin, duloxetine, or amitriptyline were excluded. Individuals with other causes of neuropathy or psychiatric illness, such as bipolar disorder, were also excluded.
Allocation and concealment. Participants were assigned to the two treatment groups strictly according to their chronological order of enrollment (alternating allocation); no random‐sequence generation, allocation concealment, or masking of the allocation schedule was used. Because allocation was not concealed and was foreseeable, the possibility of selection bias cannot be excluded, and this is acknowledged as a principal limitation. A purposive (non‐probability) sampling technique was used in view of the clinical setting and patient availability. No formal a priori sample‐size calculation was performed; the sample size of 38 participants per group (76 in total) was determined pragmatically by the number of eligible, consenting patients presenting during the fixed enrollment period (feasibility‐based sampling). A post‐hoc analysis indicated that this sample provided approximately 74% power to detect the observed Week‐24 NPRS between‐group effect size (Cohen's d ≈ 0.60) at a two‐sided α of 0.05; approximately 45 participants per group would have been required for 80% power. The study should therefore be considered modestly powered, and non‐significant comparisons interpreted with caution. We note, however, that observed (post‐hoc) power computed from the study's own effect estimate is essentially a re‐expression of the p value and adds little independent statistical information; it is reported here only as a rough descriptive indicator of study size and should not be interpreted as evidence for or against any specific effect.
2.3. Data Collection and Treatment Protocol
After written informed consent was obtained, baseline data were collected via a structured case record form. Sociodemographic characteristics, clinical history, and duration of symptoms were recorded, and baseline pain intensity and functional status were assessed. The participants were assigned to two treatment groups on the basis of the order of enrollment. Group I received pregabalin with duloxetine, and Group II received pregabalin with amitriptyline. Both groups were managed via standardized treatment protocols. Pregabalin was initiated at 25 mg once daily for the first 4 weeks, increased to 50 mg once daily for up to 8 weeks, then increased to 75 mg once daily for up to 12 weeks, and subsequently to 75 mg twice daily for up to 24 weeks. In Group I, duloxetine was started at 20 mg once daily for 8 weeks and increased to 30 mg once daily thereafter until Week 24. In Group II, amitriptyline was initiated at 10 mg once daily and increased to 25 mg once daily from Week 12 onward. Patients were followed up at 4, 8, 12, and 24 weeks. At each visit, clinical assessments were performed, and treatment adherence was evaluated. All medications were obtained from the same manufacturer to maintain consistency.
2.3.1. Rationale for the Dosing Protocol
The doses used here are lower than the maxima commonly cited in Western guidelines (e.g., pregabalin 300–600 mg/day, duloxetine 60 mg/day, amitriptyline up to 75 mg/day). This conservative, gradual dose‐escalation schedule was deliberately adopted and guided primarily by tolerability and adherence rather than by efficacy maximization. The participants were treatment‐naïve outpatients in a resource‐limited LMIC setting, in whom higher starting doses of pregabalin, duloxetine, and amitriptyline are frequently associated with sedation, dizziness, and anticholinergic effects that lead to early discontinuation; lower, well‐tolerated doses reflect real‐world prescribing practice in this population and were intended to sustain adherence over the 24‐week follow‐up. Because the two regimens differed not only in the second agent (duloxetine vs. amitriptyline) but also in their specific doses and titration schedules, the comparison should be understood as a comparison of two complete low‐dose treatment strategies rather than of the two drugs at equipotent doses; the observed differences cannot therefore be attributed unambiguously to the pharmacological properties of duloxetine versus amitriptyline.
2.3.2. Blinding and Concomitant Care
This was an open‐label study: neither the participants nor the outcome assessors were blinded to treatment allocation. To limit performance and detection bias, all assessments used the same validated instruments (NPRS and ODI), administered through a standardized case record form by the same trained assessor across visits, and participants received identical counseling and follow‐up schedules. No structured physiotherapy, exercise program, epidural steroid injection, or additional analgesic regimen was prescribed as part of the study protocol; use of over‐the‐counter analgesics outside the protocol was discouraged but not systematically recorded, which is acknowledged as a limitation. Adverse events were assessed prospectively at every scheduled visit (Weeks 4, 8, 12, and 24) by open, non‐leading inquiry and clinical examination rather than by a structured, drug‐specific symptom checklist. For each participant, only the single predominant (most bothersome) adverse event over the whole follow‐up period was recorded; the categories in Table 7 are therefore mutually exclusive, a participant who experienced more than one type of event was counted only once, and “no adverse effect” denotes that no adverse event was reported at any of the four follow‐up visits across the 24 weeks. The severity, duration, and time of onset of adverse events were not systematically graded or recorded.
Table 7.
Comparison of adverse effects between treatment groups (N = 76).
| Adverse effect | Group I (n = 38) | Group II (n = 38) | RD (95% CI) | p‐value |
|---|---|---|---|---|
| Sedation | 3 (7.9) | 12 (31.6) | −23.7% (−40.8, −6.6) | 0.02 |
| Drowsiness | 5 (13.2) | 15 (39.5) | −26.3% (−45.2, −7.4) | 0.02 |
| Vertigo | 4 (10.5) | 10 (26.3) | −15.8% (−32.9, 1.3) | 0.14 |
| No adverse effects | 26 (68.4) | 1 (2.6) | 65.8% (50.2, 81.4) | < 0.001 |
Note: Data are presented as frequencies and percentages. Group I: pregabalin with duloxetine; Group II: pregabalin with amitriptyline. RD, risk difference (Group I − Group II), with 95% confidence interval; comparisons used Fisher exact tests. Categories are mutually exclusive (the predominant adverse effect recorded per participant).
2.4. Measures
Pain intensity was assessed via the Numeric Pain Rating Scale (NPRS), a brief self‐report tool that can be completed within 1 min. The scale consists of a horizontal continuum ranging from 0 to 10, where 0 represents no pain and 10 denotes the worst imaginable pain [21, 22, 23]. Functional disability was evaluated via the Oswestry Disability Questionnaire (ODQ), a widely used disease‐specific instrument for patients with low back pain. The Oswestry Disability Index (ODI) was derived from the ODQ, which includes 10 items assessing daily functional activities. Each item is scored from 0 to 5, and the total score is summed and converted to a percentage by multiplying by two, yielding a range from 0 to 100. Higher ODI scores indicate greater functional impairment [24]. Chronic low back pain was operationally defined as pain persisting for at least 3 months with an NPRS score of ≥ 4.
2.5. Statistical Analysis
Data were analyzed in SPSS 26.0 (IBM Corp., Armonk, NY, USA) and R 4.3.1, with reporting following the SAMPL guidelines and Assel et al. [25]. Distribution was assessed by the Shapiro–Wilk test. Continuous variables are reported as mean ± SD or median (IQR), and categorical variables as frequencies with numerators and denominators. Baseline groups were compared by independent‐samples t‐tests, Mann–Whitney U tests (BMI, symptom duration), and chi‐square or Fisher exact tests; baseline imbalance was additionally quantified by standardized mean differences (SMDs). Follow‐up NPRS and ODI were compared between groups by independent‐samples t‐tests with Cohen's d and 95% CIs. Because of the baseline imbalance, the emphasized (adjusted) analyses were analysis of covariance (ANCOVA), modeling each follow‐up score on treatment group and its baseline value, and a linear mixed‐effects model with group, categorical time, and group × time as fixed effects and a single random participant intercept (one variance component; independent, homoscedastic residuals; maximum‐likelihood estimation). AIC, BIC, and the intraclass correlation coefficient are reported descriptively; residuals showed only mild non‐normality (Shapiro–Wilk W = 0.99 for both models). Clinical significance was assessed by change from baseline (95% CIs), percentage improvement (per participant, [baseline − follow‐up]/baseline × 100, averaged by group), the proportion achieving ≥ 30% and ≥ 50% NPRS reduction, and the proportion meeting the MCID (NPRS ≥ 2 points; ODI ≥ 10 points) [26], each with numerators, denominators, and Fisher exact tests. Post‐hoc power for the NPRS comparison used the two‐sample t‐test framework (two‐sided α = 0.05). NPRS was the principal outcome and ODI the main secondary outcome, but as no statistical analysis plan was registered, no pre‐specified primary endpoint exists and all inferential comparisons are exploratory. Holm–Bonferroni correction was applied to the eight between‐group follow‐up comparisons (Table 2); other analyses are unadjusted. Baseline and mixed‐effects analyses were originally planned; ANCOVA, responder, MCID, effect‐size, and multiplicity analyses were added post hoc. All tests were two‐tailed (α = 0.05); P values follow standard conventions (p < 0.001; nearest thousandth for 0.001–0.01; nearest hundredth for ≥ 0.01). There were no missing data. Because of the baseline imbalance, the baseline‐adjusted ANCOVA was treated as the primary between‐group analysis and the unadjusted t‐test comparisons as secondary.
Table 2.
Between‐group comparison of NPRS and ODI scores at follow‐up.
| Variable | Group I (n = 38) | Group II (n = 38) | Diff. (I − II) | 95% CI (Cohen's d) | p‐value | Holm p |
|---|---|---|---|---|---|---|
| NPRS—1st FU | 7.42 ± 0.89 | 8.08 ± 0.88 | −0.66 | −1.06, −0.25 (0.74) | 0.002 | 0.013 |
| NPRS—2nd FU | 5.61 ± 1.57 | 6.71 ± 1.74 | −1.11 | −1.86, −0.35 (0.67) | 0.004 | 0.024 |
| NPRS—3rd FU | 4.30 ± 2.14 | 5.74 ± 2.43 | −1.44 | −2.49, −0.40 (0.63) | 0.008 | 0.030 |
| NPRS—4th FU | 3.54 ± 2.20 | 5.09 ± 2.88 | −1.54 | −2.72, −0.37 (0.60) | 0.01 | 0.030 |
| ODI—1st FU (%) | 57.78 ± 16.18 | 69.47 ± 16.23 | −11.70 | −19.10, −4.29 (0.72) | 0.002 | 0.014 |
| ODI—2nd FU (%) | 38.60 ± 17.10 | 53.45 ± 21.43 | −14.85 | −23.72, −5.99 (0.77) | 0.001 | 0.011 |
| ODI—3rd FU (%) | 42.05 ± 9.55 | 50.88 ± 17.36 | −8.83 | −15.23, −2.42 (0.63) | 0.008 | 0.030 |
| ODI—4th FU (%) | 41.49 ± 7.46 | 48.95 ± 16.80 | −7.46 | −13.40, −1.52 (0.57) | 0.01 | 0.030 |
Note: Values are mean ± SD. Group I: pregabalin with duloxetine; Group II: pregabalin with amitriptyline. The difference column is Group I − Group II; negative values indicate lower (better) scores in Group I. All comparisons remained significant after Holm–Bonferroni correction across this family of eight tests (Holm‐adjusted p 0.011–0.030).
Abbreviations: CI, confidence interval; FU, follow‐up.
2.6. Use of Artificial Intelligence
No artificial intelligence tool was used for study design, data collection, or statistical analysis. An AI‐based language tool (ChatGPT, OpenAI) was used only to improve the readability and language of the manuscript, and all outputs were reviewed and verified by the authors.
2.7. Ethical Considerations
Ethical approval for the study was obtained from the Ethical Review Committee of Dhaka Medical College (Memo No: ERC‐DMC/ECC/2024/211). Written informed consent was obtained from all participants prior to enrollment. The study was conducted in accordance with the principles of the Declaration of Helsinki. Confidentiality and anonymity were strictly maintained throughout the study. Participation was voluntary, and participants were allowed to withdraw at any stage without consequences.
3. Results
3.1. Sociodemographic and Clinical Characteristics of the Participants
Table 1 presents the baseline sociodemographic and clinical characteristics of the study participants. The mean age of the participants was comparable between the two groups, with 40.29 ± 11.92 years in the amitriptyline group and 41.21 ± 10.63 years in the duloxetine group, with no statistically significant difference (p = 0.72; SMD = 0.08). The median BMI was also similar between the groups [25.85 (24.00–27.31) vs. 25.83 (24.36–26.90), p = 0.77]. The duration of chronic low back pain was comparable, with a median of 12 months in both groups (p = 0.90). The cohort was predominantly female (81.6% [31/38] in the duloxetine group and 76.3% [29/38] in the amitriptyline group). At baseline, however, the duloxetine group demonstrated significantly lower pain and disability scores than did the amitriptyline group. The mean NPRS score was 9.50 ± 0.51 in the amitriptyline group and 9.03 ± 0.59 in the duloxetine group (p < 0.001; SMD = 0.86). Similarly, the mean ODI score was significantly greater in the amitriptyline group (82.63 ± 8.86) than in the duloxetine group (73.45 ± 11.46) (p < 0.001; SMD = 0.90). These two clinically meaningful baseline imbalances indicate that the duloxetine group was less severely affected at entry, and they are explicitly accounted for in the adjusted analyses below. Other categorical variables, including sex distribution, residence, marital status, BMI category, income category and duration of disease, were comparable between groups, with no statistically significant differences.
Table 1.
Baseline characteristics of the study participants (N = 76).
| Variables | Category | Group I (n = 38) | Group II (n = 38) | p‐value | SMD |
|---|---|---|---|---|---|
| Age (years) | Mean ± SD | 41.21 ± 10.63 | 40.29 ± 11.92 | 0.72 | 0.08 |
| BMI (kg/m2) | Median (IQR) | 25.83 (24.36–26.90) | 25.85 (24.00–27.31) | 0.77 | 0.18 |
| Duration of CLBP (months) | Median (IQR) | 12 (6–36) | 12 (7–36) | 0.90 | 0.05 |
| NPRS score | Mean ± SD | 9.03 ± 0.59 | 9.50 ± 0.51 | < 0.001 | 0.86 |
| ODI (%) | Mean ± SD | 73.45 ± 11.46 | 82.63 ± 8.86 | < 0.001 | 0.90 |
| Sex | Male | 7 (18.4) | 9 (23.7) | 0.57 | — |
| Female | 31 (81.6) | 29 (76.3) | |||
| Residence | Rural | 14 (36.8) | 13 (34.2) | 0.81 | — |
| Urban | 24 (63.2) | 25 (65.8) | |||
| Marital status | Married | 36 (94.7) | 36 (94.7) | > 0.99 | — |
| Unmarried | 2 (5.3) | 2 (5.3) | |||
| BMI category | Normal | 16 (42.1) | 14 (36.8) | 0.59 | — |
| Overweight | 17 (44.7) | 21 (55.3) | |||
| Obese | 5 (13.2) | 3 (7.9) | |||
| Income category | Low | 2 (5.3) | 2 (5.3) | 0.17 | — |
| Middle | 30 (78.9) | 23 (60.5) | |||
| High | 6 (15.8) | 13 (34.2) | |||
| Duration category | < 1 year | 14 (36.8) | 10 (26.3) | 0.46 | — |
| 1– < 3 years | 13 (34.2) | 18 (47.4) | |||
| ≥ 3 years | 11 (28.9) | 10 (26.3) |
Note: Data are presented as frequencies and percentages unless otherwise stated. Group I: pregabalin with duloxetine; Group II: pregabalin with amitriptyline. Independent samples t‐tests were used for normally distributed variables and Mann–Whitney U tests for nonnormally distributed variables; categorical variables were compared with chi‐square or Fisher exact tests. SMD, standardized mean difference (values ≥ 0.2, 0.5, and 0.8 denote small, moderate, and large imbalance, respectively).
3.2. Pain and Functional Outcomes
The comparison of pain intensity and functional status between the two treatment groups at different follow‐up points is summarized in Table 2. Both groups demonstrated progressive improvement over time; in unadjusted comparisons the duloxetine group showed numerically greater reductions in pain scores. At the first follow‐up, the mean NPRS score was lower in the duloxetine group (7.42 ± 0.89) than in the amitriptyline group (8.08 ± 0.88), with a between‐group difference (Group I − Group II) of −0.66 (95% CI −1.06 to −0.25; Cohen's d = 0.74; p = 0.002). This difference persisted at subsequent follow‐ups, reaching −1.54 (95% CI −2.72 to −0.37; Cohen's d = 0.60) at the fourth follow‐up (3.54 ± 2.20 vs. 5.09 ± 2.88, p = 0.01). Similarly, the duloxetine group demonstrated lower ODI scores at all follow‐up points. At the first follow‐up, the mean ODI score was 57.78 ± 16.18 in the duloxetine group and 69.47 ± 16.23 in the amitriptyline group, a difference of −11.70 (95% CI − 19.10 to −4.29; Cohen's d = 0.72; p = 0.002). By the fourth follow‐up, the ODI score was 41.49 ± 7.46 in the duloxetine group compared with 48.95 ± 16.80 in the amitriptyline group, a difference of −7.46 (95% CI −13.40 to −1.52; p = 0.01). All eight between‐group comparisons remained statistically significant after Holm–Bonferroni correction (adjusted p range 0.011–0.030); negative differences denote lower (better) scores in Group I. Changes in NPRS and ODI scores from baseline across follow‐up visits are presented in Supplementary Table 1.
3.3. Clinical Significance and Responder Analysis
At Week 24, the mean reduction in NPRS from baseline was 5.48 points (95% CI 4.84–6.13; 61.3% improvement) in the duloxetine group and 4.41 points (95% CI 3.59–5.23; 47.6%) in the amitriptyline group; both exceeded the NPRS MCID of ~2 points. All duloxetine patients (38/38, 100%) achieved a ≥ 30% reduction in pain versus 30/38 (78.9%) with amitriptyline (risk difference 21.1%, 95% CI 8.1–34.0; p = 0.005), and 23/38 (60.5%) versus 13/38 (34.2%) achieved a ≥ 50% reduction (risk difference 26.3%, 95% CI 4.7–48.0; p = 0.04). The proportion meeting the ODI MCID (≥ 10 percentage‐point improvement) at week 24 was 38/38 (100%) versus 33/38 (86.8%). These responder and MCID results are summarized in Table 3 and indicate that the between‐group differences were clinically, and not only statistically, meaningful, although they are subject to the same baseline‐imbalance caveat noted above.
Table 3.
Exploratory responder and minimal clinically important difference (MCID) analyses at Week 24: change from baseline, responder rates, and MCID.
| Outcome/responder criterion | Group I (n = 38) | Group II (n = 38) | RD (95% CI) | p‐value |
|---|---|---|---|---|
| Mean NPRS reduction, points (95% CI) | 5.48 (4.84–6.13) | 4.41 (3.59–5.23) | — | — |
| NPRS % improvement | 61.3% | 47.6% | — | — |
| ≥ 30% NPRS reduction, n/N (%) | 38/38 (100) | 30/38 (78.9) | 21.1% (8.1–34.0) | 0.005 |
| ≥ 50% NPRS reduction, n/N (%) | 23/38 (60.5) | 13/38 (34.2) | 26.3% (4.7–48.0) | 0.04 |
| NPRS MCID met (≥ 2 points), n/N (%) | 38/38 (100) | 32/38 (84.2) | 15.8% (4.2–27.4) | 0.03 |
| ODI MCID met (≥ 10 points), n/N (%) | 38/38 (100) | 33/38 (86.8) | 13.2% (2.4–23.9) | 0.05 |
Note: Group I: pregabalin with duloxetine; Group II: pregabalin with amitriptyline. Responder comparisons used Fisher exact tests.
Abbreviations: CI, confidence interval; MCID, minimal clinically important difference; RD, risk difference.
3.4. Longitudinal and Baseline‐Adjusted Analysis of the Treatment Effect
To account for repeated measurements and baseline differences, a mixed‐effects model was applied, as shown in Tables 4 and 5. For NPRS, time was a significant predictor of pain reduction at all follow‐up points (p < 0.001), indicating substantial improvement over time in both groups. The overall group effect was not statistically significant (β = −0.47, 95% CI − 1.26 to 0.31; p = 0.24), whereas the interaction effect between group and time became significant at later follow‐ups, particularly at the third (β = −0.97, 95% CI − 1.59 to −0.34; p = 0.002) and fourth follow‐ups (β = −1.07, 95% CI − 1.70 to −0.45; p = 0.001). Both time and treatment group were significant predictors of the ODI; the duloxetine group had significantly lower ODI scores overall (β = −9.18, 95% CI − 15.79 to −2.58; p = 0.007), and a significant group × time interaction was observed at the second follow‐up (β = −5.67, 95% CI − 10.52 to −0.82; p = 0.02). The random‐intercept models fitted the repeated‐measures structure well (NPRS: AIC 1276.6, BIC 1323.9, ICC 0.68; ODI: AIC 2850.1, BIC 2897.4, ICC 0.73), and residuals were approximately normally distributed on inspection. After adjustment for baseline scores (ANCOVA; Table 6), the between‐group differences were markedly attenuated and no longer statistically significant at any time point for either outcome (NPRS, all p > 0.55; ODI, all p > 0.46), indicating that much of the crude difference reflected the duloxetine group's more favorable baseline status rather than a definitive treatment effect.
Table 4.
Linear mixed‐effects model for NPRS.
| Variable | β | Std. error | 95% CI | p‐value |
|---|---|---|---|---|
| Group (Duloxetine) | −0.47 | 0.41 | −1.26, 0.31 | 0.24 |
| Time (1st FU) | −1.42 | 0.23 | −1.86, −0.98 | < 0.001 |
| Time (2nd FU) | −2.79 | 0.23 | −3.23, −2.35 | < 0.001 |
| Time (3rd FU) | −3.76 | 0.23 | −4.20, −3.32 | < 0.001 |
| Time (4th FU) | −4.41 | 0.23 | −4.85, −3.97 | < 0.001 |
| Group × Time (1st) | −0.18 | 0.32 | −0.81, 0.44 | 0.56 |
| Group × Time (2nd) | −0.63 | 0.32 | −1.26, −0.01 | 0.048 |
| Group × Time (3rd) | −0.97 | 0.32 | −1.59, −0.34 | 0.002 |
| Group × Time (4th) | −1.07 | 0.32 | −1.70, −0.45 | 0.001 |
Note: Random‐intercept model (maximum‐likelihood estimation): AIC 1276.6, BIC 1323.9, ICC 0.68.
Abbreviations: CI, confidence interval; FU, follow‐up.
Table 5.
Linear mixed‐effects model for the ODI.
| Variable | β | Std. error | 95% CI | p‐value |
|---|---|---|---|---|
| Group (Duloxetine) | −9.18 | 3.42 | −15.79, −2.58 | 0.007 |
| Time (1st FU) | −13.16 | 1.77 | −16.59, −9.73 | < 0.001 |
| Time (2nd FU) | −29.18 | 1.77 | −32.61, −25.75 | < 0.001 |
| Time (3rd FU) | −31.75 | 1.77 | −35.18, −28.32 | < 0.001 |
| Time (4th FU) | −33.68 | 1.77 | −37.11, −30.25 | < 0.001 |
| Group × Time (1st) | −2.51 | 2.51 | −7.37, 2.34 | 0.31 |
| Group × Time (2nd) | −5.67 | 2.51 | −10.52, −0.82 | 0.02 |
| Group × Time (3rd) | 0.35 | 2.51 | −4.50, 5.21 | 0.89 |
| Group × Time (4th) | 1.72 | 2.51 | −3.13, 6.57 | 0.49 |
Note: Random‐intercept model (maximum‐likelihood estimation): AIC 2850.1, BIC 2897.4, ICC 0.73.
Abbreviations: CI, confidence interval; FU, follow‐up.
Table 6.
Baseline‐adjusted between‐group differences (ANCOVA).
| Outcome/time point | Adjusted group difference (Duloxetine − Amitriptyline) | 95% CI | p‐value |
|---|---|---|---|
| NPRS—1st FU | −0.09 | −0.38 to 0.21 | 0.56 |
| NPRS—2nd FU | −0.14 | −0.75 to 0.47 | 0.65 |
| NPRS—3rd FU | −0.18 | −1.07 to 0.70 | 0.68 |
| NPRS—4th FU | −0.11 | −1.08 to 0.87 | 0.83 |
| ODI—1st FU | 0.58 | −3.80 to 4.96 | 0.79 |
| ODI—2nd FU | −2.57 | −9.51 to 4.36 | 0.46 |
| ODI—3rd FU | −2.18 | −8.19 to 3.83 | 0.47 |
| ODI—4th FU | −1.89 | −7.67 to 3.88 | 0.52 |
Note: Each follow‐up score was modeled with treatment group and the corresponding baseline score as covariates. After adjustment for baseline severity, no between‐group difference remained statistically significant, indicating that baseline imbalance accounts for much of the crude difference.
Abbreviations: CI, confidence interval; FU, follow‐up.
3.5. Adverse Effects
Table 7 summarizes the adverse effects observed in the study participants. Adverse effects were recorded at every follow‐up visit through open inquiry and clinical assessment, and for each participant the predominant adverse effect was documented; the categories in Table 7 are therefore mutually exclusive and sum to the group total. We acknowledge that systematic solicitation of the full spectrum of drug‐specific adverse effects (e.g., nausea, dry mouth, constipation, insomnia, orthostatic hypotension, and weight change) and objective measurement of blood pressure and weight were not undertaken; this limited adverse‐event ascertainment is noted as a limitation, and the safety findings below should be interpreted with corresponding caution. Sedation was significantly more common in the amitriptyline group (31.6% [12/38]) than in the duloxetine group (7.9% [3/38]) (risk difference −23.7%, 95% CI − 40.8 to −6.6; p = 0.02). Similarly, drowsiness was reported more frequently in the amitriptyline group (39.5% [15/38]) than in the duloxetine group (13.2% [5/38]) (risk difference −26.3%, 95% CI − 45.2 to −7.4; p = 0.02). Although vertigo was also more common in the amitriptyline group (26.3% [10/38]) than in the duloxetine group (10.5% [4/38]), the difference was not statistically significant (p = 0.14). Notably, a substantially greater proportion of participants in the duloxetine group reported no adverse effects (68.4% [26/38]) than did those in the amitriptyline group (2.6% [1/38]) (risk difference 65.8%, 95% CI 50.2–81.4; p < 0.001).
4. Discussion
In the present study, we evaluated the comparative effectiveness of pregabalin combined with duloxetine versus pregabalin combined with amitriptyline in patients with chronic low back pain with radiculopathy. Our findings showed a consistent pattern of improvement in both groups, with numerically greater unadjusted reductions in pain intensity and functional disability in the duloxetine group across follow‐up visits. Importantly, however, the two groups were not comparable at baseline, and after adjustment for baseline severity these between‐group differences were substantially attenuated and no longer statistically significant. The results should therefore be interpreted as preliminary and hypothesis‐generating rather than as definitive evidence that one combination is superior. The mixed‐effects analysis nonetheless indicated a significant group × time interaction for NPRS at the later follow‐ups and for ODI at the second follow‐up, suggesting a possible differential trajectory that warrants confirmation in randomized studies.
The pattern of greater crude analgesic improvement associated with the pregabalin–duloxetine combination is consistent with the mechanistic rationale of multimodal analgesia in neuropathic pain. Pregabalin exerts its effect through high‐affinity binding to the α2δ−1 subunit of voltage‐gated calcium channels, reducing excitatory neurotransmitter release and attenuating central sensitization, which is a key mechanism in radicular pain [27, 28]. Duloxetine, a serotonin–norepinephrine reuptake inhibitor, enhances descending inhibitory pain pathways by increasing serotonergic and noradrenergic transmission, thereby reducing nociceptive signaling at the spinal level [29, 30]. These central mechanisms underlie duloxetine's analgesic efficacy across diverse chronic pain pathologies—including diabetic peripheral neuropathy, fibromyalgia, osteoarthritis, and chronic low back pain—and recent randomized evidence in fibromyalgia shows that adding a home‐based postural exercise program to duloxetine further improves sleep quality, physical quality of life, and trunk mobility, underscoring the drug's broad role in chronic pain management [16]. The combination of these agents provides complementary action by targeting both presynaptic excitation and supraspinal inhibition [31]. The apparently greater benefit observed with duloxetine than with amitriptyline may be related to its more selective neurotransmitter modulation, whereas amitriptyline has broader receptor activity, including histaminergic and anticholinergic effects [18].
Another observation was the numerically more pronounced and sustained improvement in functional disability in the duloxetine group. Functional recovery is a critical outcome in patients with chronic low back pain, reflecting the restoration of daily activities and overall patient well‐being [32]. Previous randomized trials have demonstrated that duloxetine significantly improves functional outcomes and reduces pain in individuals with chronic low back pain [33, 34]. Ma et al. (2023) further confirmed that duloxetine provides consistent improvements in physical function and global outcomes [34]. In our cohort, the duloxetine group had lower ODI scores at all follow‐up points in unadjusted analyses; however, this apparent advantage was largely explained by baseline imbalance once ANCOVA adjustment was applied. This may partly be explained by duloxetine's additional effect on mood and psychological factors, which contribute significantly to perceived disability in chronic pain conditions [29].
From a clinical‐significance standpoint, the within‐group reductions in both groups exceeded the accepted MCIDs for NPRS (~2 points) and ODI (~10 percentage points) [35]; because the study lacked an untreated control group, however, these within‐group changes cannot by themselves establish a causal treatment benefit. The higher proportion of duloxetine patients achieving ≥ 30% and ≥ 50% pain reduction suggests a potentially greater likelihood of clinically meaningful response. Because responder analyses are based on percentage change, they are somewhat less sensitive to the absolute baseline gap; nevertheless, residual confounding by baseline severity cannot be excluded, and these exploratory findings require randomized confirmation.
Findings from the mixed‐effects longitudinal model provide additional context for these observations. Time was a strong predictor of improvement in both groups; however, in the absence of an untreated control group, this within‐group improvement over time cannot be attributed solely to the drugs and may also reflect natural history, regression to the mean, or co‐interventions. The significant group × time interaction for NPRS at later follow‐ups and for ODI at the second follow‐up suggests that the rate of improvement may have differed between regimens, although the non‐significant overall group effect for NPRS and the attenuation seen on ANCOVA temper this interpretation. Compared with simpler analytical approaches, mixed‐effects modeling allows for appropriate handling of repeated measurements, within‐subject correlations, and baseline differences, thereby providing more robust estimates of treatment effects [36]. Any divergence between groups over time is consistent with the delayed neuroadaptive changes associated with serotonin–norepinephrine modulation [29]. This caution is especially important because the baseline NPRS scores in both groups were exceptionally high and tightly clustered near the ceiling of the scale (9.03 ± 0.59 and 9.50 ± 0.51), even though the inclusion threshold was only NPRS ≥ 4. Such extreme baseline values make regression to the mean a particularly strong candidate explanation for the large within‐group improvements. In the absence of an untreated or placebo control group, the within‐group changes cannot distinguish a genuine pharmacological treatment effect from regression to the mean, the natural history of the condition, non‐specific therapeutic‐contact (placebo) effects, or the use of concomitant analgesics, all of which may have contributed to the improvements reported here.
In the present study, differences in adverse effect profiles between the two treatment groups were clinically important and, unlike the efficacy findings, are unlikely to be explained by baseline imbalance. The amitriptyline group experienced significantly higher rates of sedation and drowsiness, whereas a substantially larger proportion of participants in the duloxetine group reported no adverse effects. These findings are consistent with the pharmacological profile of tricyclic antidepressants, which exhibit strong antagonism at histaminergic and muscarinic receptors, leading to sedation and anticholinergic side effects [37]. Evidence from a meta‐analysis of randomized controlled trials involving over 4000 patients has shown that amitriptyline is associated with a significantly increased risk of such adverse effects [26]. Similarly, comparative studies have identified sedation and drowsiness as the most frequent adverse effects associated with tricyclic antidepressants in neuropathic pain management [38]. In contrast, duloxetine's selective mechanism of action results in a more favorable tolerability profile, as supported by network meta‐analyses identifying duloxetine as a consistently effective and well‐tolerated option [17]. It should be noted that the relatively low doses used, the open‐label design, and the limited adverse‐event ascertainment may all have influenced these safety findings. Nonetheless, duloxetine‐treated participants reported fewer episodes of sedation and drowsiness; while consistent, these observations should be interpreted cautiously. Improved tolerability is generally important in chronic pain management, as adverse effects are a major determinant of treatment discontinuation and poor adherence. However, we did not collect adherence or discontinuation data, so any link between the observed tolerability difference and long‐term treatment sustainability is speculative and was not tested here. In particular, the striking difference in the proportion reporting no adverse effect (68.4% vs. 2.6%) should not be interpreted as strong evidence of superior tolerability: because adverse events were captured only by open, non‐systematic inquiry and only the predominant event per participant was recorded, this figure may reflect differential elicitation and recording as much as a true difference in the underlying adverse‐event burden, and we therefore present it as a hypothesis‐generating observation rather than a robust safety conclusion.
4.1. Strengths and Limitations
This is among the few prospective longitudinal studies in Bangladesh comparing pregabalin–duloxetine with pregabalin–amitriptyline in radicular CLBP, with repeated multi‐time‐point follow‐up and analyses using mixed‐effects modeling and baseline‐adjusted ANCOVA. Several limitations apply. Treatment was allocated non‐randomly in the order of enrollment, without concealment or blinding (open‐label), so selection, expectation, performance, and detection bias cannot be excluded, and the study is not a controlled trial. The groups were markedly imbalanced at baseline (SMD ≈ 0.9); after adjustment the between‐group differences were no longer significant, showing that the crude comparisons overstated duloxetine's apparent benefit. The random participant intercept accounts only for within‐participant correlation and does not adjust for this baseline imbalance; adjustment relied solely on the ANCOVA, and the unadjusted, responder, and MCID analyses (Table 3) are therefore secondary and exploratory, with the 100% responder rate for some endpoints reflecting small‐sample instability rather than uniform efficacy. No a priori sample‐size calculation was performed, and the single‐center sample was modest (post‐hoc power ≈ 74%), so non‐significant findings do not indicate equivalence. Doses were lower than internationally recommended, potentially limiting efficacy. Radiculopathy was diagnosed clinically without universal MRI or electrophysiology, and concomitant interventions (physiotherapy, injections, over‐the‐counter analgesics) were not controlled. Because only the predominant adverse event per participant was recorded, Table 7 underestimates the frequency of individual symptoms. Finally, the predominantly female (≈80%) single‐center LMIC cohort and the specific low‐dose protocol limit generalizability, and these limitations should be weighed when interpreting the comparative effectiveness reported here.
5. Conclusion
The present study suggests that, in this non‐randomized cohort of patients with chronic low back pain with radiculopathy, within‐group reductions in pain and disability exceeded the minimal clinically important difference thresholds in both groups. After adjustment for the baseline imbalance, the apparent efficacy advantage of the pregabalin–duloxetine combination was largely attenuated and no longer statistically significant, so a definitive efficacy benefit over pregabalin–amitriptyline cannot be claimed. Duloxetine‐treated participants reported fewer sedation‐ and drowsiness‐related adverse effects; because ascertainment was open‐label and incomplete, this tolerability observation remains exploratory and requires confirmation. These findings suggest that pregabalin combined with duloxetine may be a reasonable option to consider when both sustained pain reduction and improved tolerability are prioritized in clinical decision‐making, but they are hypothesis‐generating and should not be interpreted as establishing superiority. Further adequately powered, randomized studies are warranted to validate these comparative findings and guide optimal clinical decision‐making.
Author Contributions
Sabrina Masrufa Khanam: conceptualization, methodology, investigation, data curation, formal analysis, writing – original draft. Sujit Kumar Sarker: conceptualization, methodology, supervision, writing – review and editing, investigation. Ashekur Rahman Mullick: conceptualization, methodology, writing – review and editing. Shah Niaz MD Rubaid Anwar: conceptualization, methodology, writing – review and editing, formal analysis.
Funding
The authors have nothing to report.
Ethics Statement
Ethical approval for this study was obtained from the Ethical Review Committee of Dhaka Medical College, Dhaka, Bangladesh (Memo No: ERC‐DMC/ECC/2024/211). The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki.
Consent
All participants were provided with detailed information regarding the objectives and procedures of the study prior to enrollment. Written informed consent was obtained from each participant. Participation was entirely voluntary, and strict confidentiality and anonymity of all collected data were maintained throughout the study.
Conflicts of Interest
The authors declare no conflicts of interest.
Declaration of AI Use
An artificial intelligence‐based language tool (ChatGPT, OpenAI) was used to improve the clarity and language of the manuscript. It was not used for study design, data collection, or data analysis, and no AI tool was used to generate or edit any figure or visual content. The authors have carefully reviewed the content and take full responsibility for its accuracy and integrity.
Transparency Statement
The lead author and corresponding author, Sabrina Masrufa Khanam, affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.
Supporting information
Table S1: Change in NPRS and ODI scores from baseline across follow‐up visits.
Acknowledgments
The authors sincerely acknowledge the support of the respective institutions for facilitating this study. The authors are also grateful to all participants for their cooperation and valuable contributions during data collection. An artificial intelligence‐based language tool (ChatGPT, OpenAI) was used solely to improve the clarity and language of the manuscript; the authors reviewed and edited all content and take full responsibility for it.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. The data supporting the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Ferreira M. L., De Luca K., Haile L. M., et al., “Global, Regional, and National Burden of Low Back Pain, 1990–2020, Its Attributable Risk Factors, and Projections to 2050: A Systematic Analysis of the Global Burden of Disease Study 2021,” Lancet Rheumatology 5, no. 6 (June 2023): e316–e329, 10.1016/S2665-9913(23)00098-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Yang N., Di J., Wang W., and Feng H., “Global Burden of Low Back Pain From 1990 to 2021: A Comprehensive Analysis of Risk Factors and Trends Using the Global Burden of Disease Study 2021,” BMC Public Health 25, no. 1 (May 2025): 1886, 10.1186/s12889-025-23178-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Zahid‐Al‐Quadir A., Zaman M. M., Ahmed S., et al., “Prevalence of Musculoskeletal Conditions and Related Disabilities in Bangladeshi Adults: A Cross‐Sectional National Survey,” BMC Rheumatology 4, no. 1 (2020): 69, 10.1186/s41927-020-00169-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Momen Majumder M. S., Hakim F., Bandhan I. H., et al., “Low Back Pain in the Bangladeshi Adult Population: A Cross‐Sectional National Survey,” BMJ Open 12, no. 9 (2022): e059192, 10.1136/bmjopen-2021-059192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dydyk A. M., Khan M. Z., and Singh P., Radicular Back Pain (StatPearls, 2026). [PubMed] [Google Scholar]
- 6. Soar H., Comer C., Wilby M. J., and Baranidharan G., “Lumbar Radicular Pain,” BJA Education 22, no. 9 (2022): 343–349, 10.1016/j.bjae.2022.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Berry J. A., Elia C., Saini H. S., and Miulli D. E., “A Review of Lumbar Radiculopathy, Diagnosis, and Treatment,” Cureus 11 (October 2019): e5934, 10.7759/cureus.5934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Migliorini F., Vaishya R., Pappalardo G., Schneider M., Bell A., and Maffulli N., “Between Guidelines and Clinical Trials: Evidence‐Based Advice on the Pharmacological Management of Non‐Specific Chronic Low Back Pain,” BMC Musculoskeletal Disorders 24, no. 1 (May 2023): 432, 10.1186/s12891-023-06537-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Deyo R. A., Von Korff M., and Duhrkoop D., “Opioids for Low Back Pain,” BMJ (Clinical Research ed.) 350, (January 2015): g6380, 10.1136/bmj.g6380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Knezevic N. N., Candido K. D., Vlaeyen J. W. S., Van Zundert J., and Cohen S. P., “Low Back Pain,” Lancet 398, no. 10294 (2021): 78–92, 10.1016/S0140-6736(21)00733-9. [DOI] [PubMed] [Google Scholar]
- 11. Verma V., Singh N., and Singh Jaggi A., “Pregabalin in Neuropathic Pain: Evidences and Possible Mechanisms,” Current Neuropharmacology 12, no. 1 (2014): 44–56, 10.2174/1570159X1201140117162802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Kwon D. Y., Kim K., Kim D. H., et al., “Comparing the Effectiveness of Pregabalin and Gabapentin in Patients With Lumbar Radiculopathy: A Systematic Review and Meta‐Analysis,” Pain Practice 25, no. 1 (January 2025): e13424, 10.1111/papr.13424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Mathieson S., Maher C. G., McLachlan A. J., et al., “Trial of Pregabalin for Acute and Chronic Sciatica,” New England Journal of Medicine 376, no. 12 (March 2017): 1111–1120, 10.1056/NEJMoa1614292. [DOI] [PubMed] [Google Scholar]
- 14. Hirase T., Hirase J., Ling J., et al., “Duloxetine for the Treatment of Chronic Low Back Pain: A Systematic Review of Randomized Placebo‐Controlled Trials,” Cureus 13 (May 2021): e15169, 10.7759/cureus.15169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Schukro R. P., Oehmke M. J., Geroldinger A., Heinze G., Kress H. G., and Pramhas S., “Efficacy of Duloxetine in Chronic Low Back Pain With a Neuropathic Component: A Randomized, Double‐Blind, Placebo‐Controlled Crossover Trial,” Anesthesiology 124, no. 1 (2016): 150–158, 10.1097/ALN.0000000000000902. [DOI] [PubMed] [Google Scholar]
- 16. Demir C., Özçelep Ö. F., Okumuş E. C., Canlı M., and Alkan H., “Home‐Based Postural Exercise as an Adjunct to Duloxetine Improves Sleep Quality, Physical Quality of Life, and Trunk Mobility in Fibromyalgia: A Randomized Controlled Trial,” Rheumatology International 46, no. 6 (May 2026): 96, 10.1007/s00296-026-06137-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Birkinshaw H., Friedrich C. M., Cole P., et al., “Antidepressants for Pain Management in Adults With Chronic Pain: A Network Meta‐Analysis,” Cochrane Database of Systematic Reviews 2023, no. 5 (2023): CD014682, 10.1002/14651858.CD014682.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Sardar N., Swarnakar R., Ghosh S., et al., “Comparative Effectiveness of Amitriptyline vs Duloxetine in the Treatment of Chronic Low Back Pain: An Observational Study,” World Journal of Methodology 16, no. 1 (March 2026), 10.5662/wjm.v16.i1.107203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Tesfaye S., Sloan G., Petrie J., et al., “Comparison of Amitriptyline Supplemented With Pregabalin, Pregabalin Supplemented With Amitriptyline, and Duloxetine Supplemented With Pregabalin for the Treatment of Diabetic Peripheral Neuropathic Pain (OPTION‐DM): A Multicentre, Double‐Blind, Randomised Crossover Trial,” Lancet 400, no. 10353 (2022): 680–690, 10.1016/S0140-6736(22)01472-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Sany S. A., Tanjim T., and Hossain M. I., “Low Back Pain and Associated Risk Factors Among Medical Students in Bangladesh: A Cross‐Sectional Study,” F1000Research 10 (July 2022): 698, 10.12688/f1000research.55151.3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. De Williams A. C., Davies H. T. O., and Chadury Y., “Simple Pain Rating Scales Hide Complex Idiosyncratic Meanings,” Pain 85, no. 3 (2000): 457–463, 10.1016/S0304-3959(99)00299-7. [DOI] [PubMed] [Google Scholar]
- 22. Kim K. S., An J., Kim J. O., Lee M. Y., and Lee B. H., “Effects of Pain Neuroscience Education Combined With Lumbar Stabilization Exercise on Strength and Pain in Patients With Chronic Low Back Pain: Randomized Controlled Trial,” Journal of Personalized Medicine 12, no. 2 (February 2022): 303, 10.3390/jpm12020303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Mintken P. E., Glynn P., and Cleland J. A., “Psychometric Properties of the Shortened Disabilities of the Arm, Shoulder, and Hand Questionnaire (QuickDASH) and Numeric Pain Rating Scale in Patients With Shoulder Pain,” Journal of Shoulder and Elbow Surgery 18, no. 6 (2009): 920–926, 10.1016/j.jse.2008.12.015. [DOI] [PubMed] [Google Scholar]
- 24. Fairbank J. C. and Pynsent P. B., “The Oswestry Disability Index,” Spine 25, no. 22 (November 2000): 2940–2952. [DOI] [PubMed] [Google Scholar]
- 25. Assel M., Sjoberg D., Elders A., et al., “Guidelines for Reporting of Statistics for Clinical Research in Urology,” European Urology 75, no. 3 (2019): 358–367, 10.1016/j.eururo.2018.12.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Brueckle M. S., Thomas E. T., Seide S. E., et al., “Amitriptyline's Anticholinergic Adverse Drug Reactions—A Systematic Multiple‐Indication Review and Meta‐Analysis,” PLoS One 18, no. 4 (April 2023): e0284168, 10.1371/journal.pone.0284168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Bauer C. S., Nieto‐Rostro M., Rahman W., et al., “The Increased Trafficking of the Calcium Channel Subunit α2δ−1 to Presynaptic Terminals in Neuropathic Pain Is Inhibited by the α2δ Ligand Pregabalin,” Journal of Neuroscience 29, no. 13 (April 2009): 4076–4088, 10.1523/JNEUROSCI.0356-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Alles S. R. A., Cain S. M., and Snutch T. P., “Pregabalin as a Pain Therapeutic: Beyond Calcium Channels,” Frontiers in Cellular Neuroscience 14 (April 2020): 83, 10.3389/fncel.2020.00083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Obata H., “Analgesic Mechanisms of Antidepressants for Neuropathic Pain,” International Journal of Molecular Sciences 18, no. 11 (November 2017): 2483, 10.3390/ijms18112483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Onuţu A. H., “Duloxetine, an Antidepressant With Analgesic Properties—A Preliminary Analysis,” Romanian Journal of Anaesthesia and Intensive Care 22, no. 2 (2015): 123–128. [PMC free article] [PubMed] [Google Scholar]
- 31. Marcianò G., Evangelista M., Vocca C., et al., “Pregabalin and Duloxetine in Patients With Non‐Nociceptive Pain: A Narrative Review Exploring the Pharmacological Effects of This Combination,” Pharmaceuticals 18, no. 10 (2025): 1434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Gatchel R. J. and Mayer T. G., “Evidence‐Informed Management of Chronic Low Back Pain With Functional Restoration,” Spine Journal 8, no. 1 (2008): 65–69, 10.1016/j.spinee.2007.10.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Skljarevski V., Ossanna M., Liu‐Seifert H., et al., “A Double‐Blind, Randomized Trial of Duloxetine Versus Placebo in the Management of Chronic Low Back Pain,” European Journal of Neurology 16, no. 9 (2009): 1041–1048, 10.1111/j.1468-1331.2009.02648.x. [DOI] [PubMed] [Google Scholar]
- 34. Ma X., Zhou S., Sun W., et al., “Efficacy and Safety of Duloxetine in Chronic Musculoskeletal Pain: A Systematic Review and Meta‐Analysis,” BMC Musculoskeletal Disorders 24, no. 1 (May 2023): 394, 10.1186/s12891-023-06488-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Ostelo R. W. J. G., Deyo R. A., Stratford P., et al., “Interpreting Change Scores for Pain and Functional Status in Low Back Pain: Towards International Consensus Regarding Minimal Important Change,” Spine 33, no. 1 (2008): 90–94, 10.1097/BRS.0b013e31815e3a10. [DOI] [PubMed] [Google Scholar]
- 36. Davis S., “Mixed Models for Repeated Measures Using Categorical Time Effects (MMRM),” in Clinical Trials With Missing Data: A Guide for Practitioners (Wiley, 2014), 130–184. [Google Scholar]
- 37. Gillman P. K., “Tricyclic Antidepressant Pharmacology and Therapeutic Drug Interactions Updated,” British Journal of Pharmacology 151, no. 6 (2007): 737–748, 10.1038/sj.bjp.0707253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Reinert J. P., Veronin M. A., and Medina C., “Tricyclic Antidepressants in Nociceptive and Neuropathic Pain: A Review of Their Analgesic Properties in Combination With Opioids,” Journal of Pharmacy Technology 39, no. 1 (2023): 35–40, 10.1177/87551225221139699. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Change in NPRS and ODI scores from baseline across follow‐up visits.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. The data supporting the findings of this study are available from the corresponding author upon reasonable request.