Structured coaching interventions for anxiety and depressive disorders in Primary Care: A randomized controlled trial Authors: María del Pino Pérez García, Juan Francisco García Granado, Jose Luis Hernández Fletas, Noemi María Domínguez Hernández, Pedro Saavedra Santana Published: 2026 ABSTRACT ObjectiveTo evaluate the effectiveness of structured coaching interventions on anxiety, depression, emotional intelligence, and quality of life in adults with anxiety and/or depressive disorders in Primary Care.DesignPilot randomized, controlled, open-label clinical trial.SiteSan Gregorio Health Center, Telde (Gran Canaria, Spain), Primary Care setting.ParticipantsThirty adults aged 18–65 years with anxiety and/or depressive disorders were randomized (1:1:1) to control, face-to-face coaching, or telephone coaching.InterventionsFive weekly structured coaching sessions based on the GROW model were delivered face-to-face or by telephone. The control group received usual care.Main measurementsPrimary outcomes were anxiety (Hamilton Anxiety Scale, HAS) and depression (Montgomery–Åsberg Depression Rating Scale, MADRS). Secondary outcomes included emotional intelligence (Trait Meta-Mood Scale-24) and quality of life (EuroQol-5D), assessed at baseline, one month, and five months.ResultsA significant time × treatment interaction was observed for anxiety (p < 0.001). Both coaching modalities achieved greater reductions than control at one month (p = 0.002) and five months (p = 0.019). Face-to-face coaching showed a stronger short-term effect, whereas telephone coaching demonstrated more sustained improvement. No significant effects were found for depressive symptoms. Emotional regulation improved at one month in the face-to-face group (p = 0.013) but was not sustained. Quality of life improved over time without between-group differences.ConclusionsStructured coaching was associated with reduced anxiety symptoms in Primary Care. These findings should be interpreted cautiously and support the feasibility of larger trials. Introduction Anxiety and depressive disorders are the most prevalent mental health conditions in Primary Care, with estimated rates of 6.7% and 4.1%, exceeding 10% when DSM-5 criteria are applied.1, 2 Globally, prevalence ranges from 20% to 40%, with a marked increase after the COVID-19 pandemic.3 Although psychotherapy is considered first-line treatment, limited access and long waiting times often lead to early pharmacological management. Coaching, defined by the Spanish Association of Coaching (ASESCO) as an individualized and confidential process that promotes personal and professional development without constituting psychotherapy, has emerged as a complementary healthcare strategy. Health coaching focuses on intrinsic motivation, goal setting, and behavioral change, and can be delivered face-to-face or remotely.4, 5 Previous studies suggest that coaching reduces stress and improves well-being, adherence, and clinical outcomes in anxiety and depression,6, 7, 8 as well as in chronic conditions such as heart failure, COPD, chronic pain, and obesity.9, 10, 11, 12 Randomized controlled trials and systematic reviews have reported benefits in patient activation, self-management, and psychological well-being across different populations.13, 14, 15 However, evidence on its effectiveness in affective disorders within real-world Primary Care settings remains limited. Therefore, we conducted a pilot randomized controlled trial to evaluate the feasibility and potential effectiveness of structured coaching interventions delivered face-to-face or by telephone in this setting. The objective of this randomized controlled clinical trial is evaluated the effectiveness of structured coaching as a complementary intervention in adults aged 18–65 years with anxiety and/or depressive disorders in Primary Care at the San Gregorio Health Center (Gran Canaria, Spain). Participants were randomized into three groups: face-to-face coaching based on the GROW model, telephone coaching, or usual care (control). Primary outcomes were anxiety and depression, assessed using the Hamilton Anxiety Scale and the Montgomery–Åsberg Depression Rating Scale. Secondary outcomes included emotional intelligence and quality of life, measured with the Trait Meta-Mood Scale-24 and the EuroQol-5D. Material and methods Participants were consecutively recruited by the Primary Care team at the San Gregorio Health Center (Gran Canaria, Spain). The study included adults aged 18–65 years with a clinical diagnosis of anxiety and/or depressive disorders registered in the DRAGO electronic medical record system of the Canary Islands Health Service. Eligible diagnoses were identified according to ICD-10 and ICD-9 codes and included anxiety disorders (e.g., generalized anxiety disorder, panic disorder, phobias, adjustment disorder, post-traumatic stress disorder, obsessive–compulsive disorder, and somatoform disorders) and depressive disorders (depressive episode and major depressive disorder). Inclusion and exclusion criteria Eligible participants were adults aged 18–65 years meeting diagnostic criteria for anxiety and/or depressive disorders based on the study psychometric assessments, who provided written informed consent. Exclusion criteria included age <18 or >65 years; presence of other major psychiatric disorders, neurodegenerative diseases, significant neurological deficits, or fibromyalgia; substance abuse; terminal illness; dependency status; and inability or refusal to provide informed consent. Study design This pilot randomized, controlled, open-label clinical trial was conducted in a Primary Care setting between October 2023 and May 2025 to evaluate the feasibility, acceptability, and preliminary effectiveness of structured coaching interventions. Thirty adults aged 18–65 years meeting inclusion criteria were recruited from the San Gregorio Health Center (Telde, Spain). Data were obtained from the DRAGO electronic medical record system (Canary Islands Health Service) and recorded in an anonymized, password-protected database. Participants were randomized (1:1:1) to usual care (control), face-to-face coaching, or telephone coaching using a computer-generated sequence, maintained by an independent researcher to ensure allocation concealment. Baseline assessment included sociodemographic and clinical variables, as well as validated measures of anxiety, depression, emotional intelligence, and quality of life. Follow-up assessments were conducted at baseline, one month, and five months. Selection process Participants were screened by Primary Care physicians using validated instruments to confirm eligibility. Moderate-to-severe anxiety was defined as a Hamilton Anxiety Scale score ≥18, and moderate-to-severe depression as a Montgomery–Åsberg Depression Rating Scale score ≥20. When clinically indicated, both scales were administered. Perceived emotional intelligence was assessed using the Trait Meta-Mood Scale-24 (TMMS-24), and health-related quality of life with the EuroQol-5D. Following baseline assessment and confirmation of eligibility, participants provided written informed consent prior to enrollment. The study database included sociodemographic (age, sex, educational level, occupation, marital status, household composition, and income) and clinical variables (Primary Care diagnosis, psychotropic medication use, and psychiatric or psychological follow-up). Anxiety and depression were assessed using the Hamilton Anxiety Scale (HAS) and the Montgomery–Åsberg Depression Rating Scale (MADRS). Emotional intelligence was measured with the Trait Meta-Mood Scale-24 (TMMS-24), and health-related quality of life with the EuroQol-5D. Assessments were conducted at baseline, one month, and five months after the intervention. Coaching intervention The intervention consisted of five individual weekly coaching sessions of approximately 60 min, delivered face-to-face or by telephone according to group allocation. The first session was conducted within 7 days after baseline assessment. All sessions followed a standardized structure based on the GROW model16 (Goal, Reality, Options, Will/Action), guiding goal setting, exploration of the current situation, identification of strategies, and development of an action plan. Sessions were individualized according to participants’ objectives but followed a consistent methodological structure. The first session included goal clarification and agreement, intermediate sessions focused on progress review and action planning, and the final session on consolidation and maintenance strategies. The intervention was delivered by a certified coaching professional with experience in healthcare settings. Statistical analysis Data were analyzed using appropriate statistical methods for qualitative and quantitative variables. Given the exploratory nature of this pilot study and the limited sample size, analyses were considered exploratory and no formal sample size calculation was performed. A total of 30 participants (10 per group) was deemed sufficient to assess feasibility and obtain preliminary effect size estimates for future studies. Quantitative variables included outcome scale scores (HAS, MADRS, TMMS-24, EuroQol-5D) and number of cohabitants; net monthly income was treated as a continuous variable. Qualitative variables included sociodemographic and clinical characteristics. For each participant, outcomes were evaluated at baseline, one month, and five months. Percentage change from baseline was calculated, and clinically relevant improvement was defined as a ≥20% change. The number needed to treat (NNT) was estimated for each outcome. Categorical variables were expressed as frequencies and percentages, and continuous variables as mean ± standard deviation (SD) or median and interquartile range (IQR), depending on distribution. Between-group comparisons were performed using Chi-square or Fisher's exact test for categorical variables, and F-test or Kruskal–Wallis test for continuous variables. When appropriate, post hoc comparisons were conducted using the Nemenyi method.17 Statistical significance was set at p < 0.05. All analyses were performed using R statistical software (version 4.2.1; R Core Team, 2022).18 Outcome variables Primary outcomes were changes in anxiety and depressive symptoms assessed using the Hamilton Anxiety Scale (HAS) and the Montgomery–Åsberg Depression Rating Scale (MADRS). Secondary outcomes included quality of life (EuroQol-5D) and emotional intelligence (Trait Meta-Mood Scale-24, TMMS-24). Outcomes were evaluated at baseline, one month, and five months. This study was conducted in accordance with the Declaration of Helsinki and approved by the Research Ethics Committee of Las Palmas (CEI/CEIm) (approval code: 2024-241-1; Act 5/2024, May 23, 2024). Authorization was obtained from the Primary Care Management of Gran Canaria. All participants provided written informed consent prior to inclusion. Participation was voluntary, and participants could withdraw at any time without affecting their standard care. Data were anonymized and stored in a secure, password-protected database to ensure confidentiality and compliance with applicable regulations. Results Thirty participants were included, with equal allocation across the three groups (n = 10 per group). Overall, 70% were women. The proportion of women was 60% in the control and face-to-face groups and 90% in the telephone coaching group. No statistically significant differences were observed in sex distribution between groups (χ2 = 2.857; p = 0.240). Baseline sociodemographic and clinical characteristics were comparable across treatment arms (Table 1).Table 1Baseline patient characteristics.Table 1VariableOveralln = 30Controln = 10Phonen = 10Face-to-facen = 10p-ValueAge, years142.2 ± 11.742.8 ± 12.341.7 ± 13.642.3 ± 10.40.982Sex male29 (30.0)4 (40.0)1 (10.0)4 (40.0)0.297Psychotropic drugs221 (70.0)9 (90.0)4 (40.0)8 (80.0)0.077Psychiatric follow-up27 (24.1)2 (20.0)1 (11.1)4 (40.0)0.43Follow-up by psychologist212 (41.4)4 (40.0)3 (33.3)5 (50.0)0.893 Educational level20.478 Primary education2 (6.9)01 (11.1)1 (10.0) Secondary education4 (13.8)2 (20.0)1 (11.1)1 (10.0) High school9 (31.0)3 (30.0)1 (11.1)5 (50.0) University education11 (37.9)5 (50.0)4 (44.4)2 (20.0) Postgraduate education3 (10.3)02 (22.2)1 (10.0) Marital status20.676 Single11 (37.9)3 (30.0)3 (33.3)5 (50.0) Married (religious)5 (17.2)2 (20.0)3 (33.3)0 Married (civil)7 (24.1)3 (30.0)1 (11.1)3 (30.0) Widowed1 (3.4)1 (10.0)00 Domestic partnership2 (6.9)01 (11.1)1 (10.0) Separated3 (10.3)1 (10.0)1 (11.1)1 (10.0) Baseline anxiety scale329 (23.8; 35)26 (22; 33)29 (26; 32)32 (25.2; 38)0.593Net monthly income32000(1500; 2500)2150(825; 2500)2000(1600; 3000)1900(1500; 2425)0.759Data are means ± SD1, frequencies (%)2 and medians (IQR)3. Table 1Baseline patient characteristics.Table 1VariableOveralln = 30Controln = 10Phonen = 10Face-to-facen = 10p-ValueAge, years142.2 ± 11.742.8 ± 12.341.7 ± 13.642.3 ± 10.40.982Sex male29 (30.0)4 (40.0)1 (10.0)4 (40.0)0.297Psychotropic drugs221 (70.0)9 (90.0)4 (40.0)8 (80.0)0.077Psychiatric follow-up27 (24.1)2 (20.0)1 (11.1)4 (40.0)0.43Follow-up by psychologist212 (41.4)4 (40.0)3 (33.3)5 (50.0)0.893 Educational level20.478 Primary education2 (6.9)01 (11.1)1 (10.0) Secondary education4 (13.8)2 (20.0)1 (11.1)1 (10.0) High school9 (31.0)3 (30.0)1 (11.1)5 (50.0) University education11 (37.9)5 (50.0)4 (44.4)2 (20.0) Postgraduate education3 (10.3)02 (22.2)1 (10.0) Marital status20.676 Single11 (37.9)3 (30.0)3 (33.3)5 (50.0) Married (religious)5 (17.2)2 (20.0)3 (33.3)0 Married (civil)7 (24.1)3 (30.0)1 (11.1)3 (30.0) Widowed1 (3.4)1 (10.0)00 Domestic partnership2 (6.9)01 (11.1)1 (10.0) Separated3 (10.3)1 (10.0)1 (11.1)1 (10.0) Baseline anxiety scale329 (23.8; 35)26 (22; 33)29 (26; 32)32 (25.2; 38)0.593Net monthly income32000(1500; 2500)2150(825; 2500)2000(1600; 3000)1900(1500; 2425)0.759Data are means ± SD1, frequencies (%)2 and medians (IQR)3. Baseline patient characteristics. Data are means ± SD1, frequencies (%)2 and medians (IQR)3. Baseline scores on anxiety, depression, emotional intelligence, and quality of life Baseline values are shown in Table 2. Participants presented moderate-to-severe anxiety levels (HAS mean 29.4 ± 9.1) and moderate depressive symptoms (MADRS mean 15.8 ± 15.0). The mean TMMS-24 score was 73.4 ± 15.0, indicating moderate perceived emotional intelligence, and the EQ-5D mean score was 52.0 ± 21.4.Table 2Baseline scores on anxiety, depression, emotional intelligence, and quality of life by treatment group.Table 2VariableControln = 10Phonen = 10Face-to-facen = 10p-ValueAnxiety (HAS)26.3 ± 8.828.4 ± 6.832.1 ± 9.50.316Depression (MADRS)13.5 ± 14.39.7 ± 13.213.3 ± 17.80.157Emotional intelligence (TMMS-24)72.7 ± 16.978.8 ± 11.968.7 ± 15.50.327Quality of life (EQ-5D)53.5 ± 17.051.0 ± 29.451.5 ± 18.00.965 Table 2Baseline scores on anxiety, depression, emotional intelligence, and quality of life by treatment group.Table 2VariableControln = 10Phonen = 10Face-to-facen = 10p-ValueAnxiety (HAS)26.3 ± 8.828.4 ± 6.832.1 ± 9.50.316Depression (MADRS)13.5 ± 14.39.7 ± 13.213.3 ± 17.80.157Emotional intelligence (TMMS-24)72.7 ± 16.978.8 ± 11.968.7 ± 15.50.327Quality of life (EQ-5D)53.5 ± 17.051.0 ± 29.451.5 ± 18.00.965 Baseline scores on anxiety, depression, emotional intelligence, and quality of life by treatment group. No significant baseline differences were observed between groups for anxiety, depression, emotional intelligence, or quality of life (all p > 0.05). Changes in anxiety and depression scores over time Repeated-measures ANOVA showed a significant main effect of time on anxiety (F[2,25] = 17.3; p < 0.001) and a significant time × group interaction (F[4,52] = 6.03; p < 0.001; Greenhouse–Geisser corrected p = 0.006), indicating differential trajectories across interventions. No significant main effect of group was observed (F[2,26] = 0.67; p = 0.521). In the control group, anxiety scores remained stable (26.8 baseline; 27.2 at 1 month; 26.1 at 5 months). The telephone coaching group showed a progressive reduction (28.4; 21.4; 19.7), whereas the face-to-face group demonstrated a marked short-term decrease followed by partial rebound (32.1; 21.3; 26.9). Polynomial contrasts revealed significant linear (p = 0.002) and quadratic trends (p = 0.001), consistent with sustained improvement in the telephone group and a rebound pattern in the face-to-face group (Fig. 1).Figure 1Estimated marginal means of Hamilton Anxiety Scale (HAS), Montgomery–Åsberg Depression Rating Scale (MADRS), Trait Meta-Mood Scale-24 (TMMS-24), and EuroQol-5D (EQ-5D) scores across baseline, 1 month, and 5 months by treatment group. Estimated marginal means of Hamilton Anxiety Scale (HAS), Montgomery–Åsberg Depression Rating Scale (MADRS), Trait Meta-Mood Scale-24 (TMMS-24), and EuroQol-5D (EQ-5D) scores across baseline, 1 month, and 5 months by treatment group. For depressive symptoms, no significant main effects of time (F[2,27] = 1.72; p = 0.190) or group (F[2,27] = 0.29; p = 0.749) were detected. The time × group interaction was also non-significant (F[4,54] = 1.09; p = 0.372; Greenhouse–Geisser corrected p = 0.362). Depression scores remained stable in the control group (13.5; 13.8; 13.9). Modest, non-significant reductions were observed in the telephone (9.7; 8.9; 8.4) and face-to-face groups (13.3; 10.2; 10.1). Polynomial contrasts confirmed the absence of significant linear (p = 0.153) or quadratic effects (p = 0.372) (Fig. 1). Changes in emotional intelligence and quality of life Repeated-measures ANOVA showed a significant main effect of time on emotional intelligence (F[2,27] = 6.54; p = 0.003), with no significant main effect of group (F[2,27] = 1.02; p = 0.374) or time × group interaction (F[4,54] = 1.53; p = 0.208). Polynomial contrasts revealed a significant quadratic trend (p = 0.001), indicating initial improvement followed by slight decline at five months. Mean TMMS-24 scores increased at one month in all groups (control: 72.7–76.4; telephone: 78.8–83.4; face-to-face: 68.7–81.8). At five months, scores decreased slightly in the control (72.4) and face-to-face groups (77.8), whereas the telephone group-maintained improvement (83.5) (Fig. 1). For quality of life, a significant main effect of time was observed (F[2,26] = 11.78; p < 0.001), without significant group (F[2,26] = 1.15; p = 0.331) or interaction effects (F[4,52] = 1.19; p = 0.328). Linear (p = 0.007) and quadratic (p < 0.001) components indicated early improvement with partial decline at five months. EQ-5D scores improved in one month in all groups (control: 53.5–61.0; telephone: 54.4–72.8; face-to-face: 51.5–73.0). At five months, the telephone group-maintained improvement (71.7), while slight reductions were observed in the control (56.5) and face-to-face groups (67.0) (Fig. 1). Percentage change from baseline in anxiety and depression scores Results are summarized in Table 3.Table 3Percentage increase of the scales from baseline.Table 3Controln = 10Face-to-facen = 10Phonen = 10p-ValueAnxiety One month3.6 (−12.1; 5.9)a−36.4 (−51.9; −31.1)b−24.7 (−33.5; −10.5)b0.002 Five months0 (−10.2; 4.8)a−22.3 (−30.6; −13)a,b−34.9 (−38.5; −17.5)b0.019 Depression One month0 (−40; 0)0 (−31; 33.3)0 (−3.3; 2.5)0.366 Five months−2.9 (−40; 14.3)0 (−31.5; 18.8)−19.2 (−44.9; 5)0.770 Perception One month−1.6 (−6.5; 0)9.7 (2.0; 38.6)0 (−5.9; 3.4)0.114 Five months0 (−19.7; 0)5 (−20.8; 30.2)0 (−5.9; 0)0.746 Understanding One month1.7 (0; 20.9)22 (11.9; 58.8)8.2 (5; 11.1)0.252 Five months10.7 (1.9; 18.6)15.6 (2.7; 60)9.9 (6.1; 21.5)0.754 Regulation One month0 (0; 0)a21.4 (9.3; 35.1)b6.2 (1.2; 22.4)a,b0.013 Five months0 (−7.5; 5.8)13.1 (2.3; 17.9)6.2 (0; 39.1)0.135 Emotional intelligence One month0 (−1.7; 2.6)a14.4 (5.4; 42.7)b5.4 (1.6; 7.5)a,b0.053 Five months0.6 (−5.9; 4.7)10.1 (0.3; 32.7)8.8 (5.4; 13.2)0.140 Quality of life One month9.2 (7.1; 12.5)34.8 (27.1; 44.1)11.4 (7.2; 80)0.115 Five months0 (−5.4; 15.0)32.9 (17; 71.1)11.8 (6.2; 33.3)0.230Data are medians (IQR).a,b Different superscripts across rows indicate significant different medians. Table 3Percentage increase of the scales from baseline.Table 3Controln = 10Face-to-facen = 10Phonen = 10p-ValueAnxiety One month3.6 (−12.1; 5.9)a−36.4 (−51.9; −31.1)b−24.7 (−33.5; −10.5)b0.002 Five months0 (−10.2; 4.8)a−22.3 (−30.6; −13)a,b−34.9 (−38.5; −17.5)b0.019 Depression One month0 (−40; 0)0 (−31; 33.3)0 (−3.3; 2.5)0.366 Five months−2.9 (−40; 14.3)0 (−31.5; 18.8)−19.2 (−44.9; 5)0.770 Perception One month−1.6 (−6.5; 0)9.7 (2.0; 38.6)0 (−5.9; 3.4)0.114 Five months0 (−19.7; 0)5 (−20.8; 30.2)0 (−5.9; 0)0.746 Understanding One month1.7 (0; 20.9)22 (11.9; 58.8)8.2 (5; 11.1)0.252 Five months10.7 (1.9; 18.6)15.6 (2.7; 60)9.9 (6.1; 21.5)0.754 Regulation One month0 (0; 0)a21.4 (9.3; 35.1)b6.2 (1.2; 22.4)a,b0.013 Five months0 (−7.5; 5.8)13.1 (2.3; 17.9)6.2 (0; 39.1)0.135 Emotional intelligence One month0 (−1.7; 2.6)a14.4 (5.4; 42.7)b5.4 (1.6; 7.5)a,b0.053 Five months0.6 (−5.9; 4.7)10.1 (0.3; 32.7)8.8 (5.4; 13.2)0.140 Quality of life One month9.2 (7.1; 12.5)34.8 (27.1; 44.1)11.4 (7.2; 80)0.115 Five months0 (−5.4; 15.0)32.9 (17; 71.1)11.8 (6.2; 33.3)0.230Data are medians (IQR).a,b Different superscripts across rows indicate significant different medians. Percentage increase of the scales from baseline. Data are medians (IQR). a,b Different superscripts across rows indicate significant different medians. At one month, significant differences between-group were observed for anxiety (p = 0.002). The face-to-face group showed the greatest reduction (median −36.4%; IQR −51.9 to −31.1), followed by the telephone group (−24.7%; IQR −33.5 to −10.5), while the control group showed minimal change (3.6%; IQR −12.1 to 5.9). Post hoc analyses confirmed greater reductions in both intervention groups compared with control. At five months, between-group differences in anxiety remained significant (p = 0.019). The telephone group demonstrated the largest sustained reduction (−34.9%; IQR −38.5 to −17.5), followed by the face-to-face group (−22.3%; IQR −30.6 to −13), whereas the control group showed no meaningful change (0%; IQR −10.2 to 4.8) (Fig. 2).Figure 2Percentage change from baseline in anxiety and depression at one- and five-month follow-up by intervention group. Percentage change from baseline in anxiety and depression at one- and five-month follow-up by intervention group. No significant between-group differences were detected for depressive symptoms at one month (p = 0.366) or five months (p = 0.770), despite a numerically greater reduction in the telephone group at five months (−19.2%; IQR −44.9 to 5). For emotional intelligence, a significant difference was found in the regulation subscale at one month (p = 0.013), with greater improvement in the face-to-face group (21.4%; IQR 9.3–35.1). No differences were observed over the five months. Total emotional intelligence showed a borderline difference at one month (p = 0.053) without sustained effects. No significant between-group differences were observed for perception, understanding, or quality of life at follow-up, although the face-to-face group showed consistently greater numerical improvements (Fig. 3).Figure 3Percentage change from baseline in emotional intelligence dimensions (perception, understanding, and regulation) and quality of life at one- and five-month follow-up by intervention group. Percentage change from baseline in emotional intelligence dimensions (perception, understanding, and regulation) and quality of life at one- and five-month follow-up by intervention group. Discussion This pilot randomized controlled trial evaluated structured coaching interventions in a Primary Care population with affective disorders. Participants continued receiving usual Primary Care management during follow-up, including pharmacological or psychological treatments when clinically indicated. Therefore, the potential influence of cointerventions cannot be completely excluded. Longitudinal and percentage-change analyses yielded consistent findings, particularly for anxiety. A significant time × treatment interaction was observed for anxiety. While symptoms remained stable in the control group, both coaching modalities achieved meaningful reductions. Face-to-face coaching showed a stronger short-term effect, whereas telephone coaching demonstrated more sustained improvement at five months, suggesting modality-specific temporal patterns. These findings are consistent with previous studies suggesting that health coaching interventions may improve psychological well-being and self-management in patients with chronic conditions and mental health problems.13, 14 No significant effects were found for depressive symptoms. Although modest reductions were observed, especially in the telephone group, they did not reach statistical significance. This divergence may reflect a greater influence of coaching on cognitive and behavioral aspects of anxiety, such as worry and emotional dysregulation, than on core depressive symptoms. Emotional intelligence improved over time, with a short-term advantage in emotional regulation favoring face-to-face coaching, though this effect was not sustained. Quality of life increased across follow-up without significant between-group differences. Limitations of this study include the small sample size, limited statistical power, open-label design, and reliance on self-reported measures. The pilot nature of the study further restricts generalizability; therefore, results should be interpreted as preliminary and hypothesis-generating. In addition, the open-label design and the use of heteroapplied psychometric scales may have introduced measurement bias, since outcome assessments were not blinded to treatment allocation. This could have influenced symptom ratings, particularly in subjective measures such as anxiety and depression. Additionally, the number needed to treat (NNT) was calculated as an exploratory measure to estimate the potential clinical impact of the intervention; however, given the small sample size, this estimate should be interpreted with caution. Finally, the control group did not receive an equivalent attention-based intervention, which could have introduced attention-related bias. Conclusions Findings from this pilot randomized controlled trial suggest that structured coaching may represent a feasible complementary strategy for anxiety management in Primary Care. Face-to-face coaching showed a stronger short-term effect, whereas telephone coaching demonstrated more sustained improvement in the five months. No significant effects were observed for depressive symptoms. Emotional regulation improved in the short term but was not maintained, and quality of life increased without significant between-group differences. These findings should be interpreted cautiously given the pilot nature of the study, and larger trials with longer follow-up are required.What is known about the subject•Anxiety and depression disorders are highly prevalent in Primary Care and represent a significant healthcare burden.•Limited access to psychological interventions encourages the early use of pharmacological treatment.•Healthcare coaching has shown benefits in psychological well-being and self-care in different clinical populations.•Evidence regarding its effectiveness in affective disorders in Primary Care is still limited.What this study contributes•This pilot randomized clinical trial evaluates the impact of structured coaching on patients with anxiety and depressive disorders in primary care.•Coaching was associated with a significant reduction in anxiety symptoms compared to usual care.•The in-person format showed a greater short-term effect, while the telephone format showed a more sustained improvement.•No significant effects on depressive symptoms were observed. What is known about the subject •Anxiety and depression disorders are highly prevalent in Primary Care and represent a significant healthcare burden.•Limited access to psychological interventions encourages the early use of pharmacological treatment.•Healthcare coaching has shown benefits in psychological well-being and self-care in different clinical populations.•Evidence regarding its effectiveness in affective disorders in Primary Care is still limited. •Anxiety and depression disorders are highly prevalent in Primary Care and represent a significant healthcare burden. Anxiety and depression disorders are highly prevalent in Primary Care and represent a significant healthcare burden. •Limited access to psychological interventions encourages the early use of pharmacological treatment. Limited access to psychological interventions encourages the early use of pharmacological treatment. •Healthcare coaching has shown benefits in psychological well-being and self-care in different clinical populations. Healthcare coaching has shown benefits in psychological well-being and self-care in different clinical populations. •Evidence regarding its effectiveness in affective disorders in Primary Care is still limited. Evidence regarding its effectiveness in affective disorders in Primary Care is still limited. What this study contributes •This pilot randomized clinical trial evaluates the impact of structured coaching on patients with anxiety and depressive disorders in primary care.•Coaching was associated with a significant reduction in anxiety symptoms compared to usual care.•The in-person format showed a greater short-term effect, while the telephone format showed a more sustained improvement.•No significant effects on depressive symptoms were observed. •This pilot randomized clinical trial evaluates the impact of structured coaching on patients with anxiety and depressive disorders in primary care. This pilot randomized clinical trial evaluates the impact of structured coaching on patients with anxiety and depressive disorders in primary care. •Coaching was associated with a significant reduction in anxiety symptoms compared to usual care. Coaching was associated with a significant reduction in anxiety symptoms compared to usual care. •The in-person format showed a greater short-term effect, while the telephone format showed a more sustained improvement. The in-person format showed a greater short-term effect, while the telephone format showed a more sustained improvement. •No significant effects on depressive symptoms were observed. No significant effects on depressive symptoms were observed. Ethical considerations This study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Research Ethics Committee of Las Palmas (CEI/CEIm de Las Palmas) (approval code: 2024-241-1; Act 5/2024, May 23, 2024). Authorization to conduct the study in the Primary Care setting was granted by the Primary Care Management of Gran Canaria. All participants provided written informed consent prior to inclusion in the study. Participation was voluntary, and participants were informed of their right to withdraw at any time without affecting their standard medical care. All data were anonymized and stored securely in a password-protected database to ensure confidentiality and data protection in accordance with applicable regulations. Funding This study did not receive external commercial or industry funding. Conflict of interest The authors declare that they have no conflicts of interest.