Efficacy of curcumin in depression: A grade-assessed systematic review and meta-analysis of randomized controlled trials

Abstract

Backgrounds

Curcumin is a natural polyphenolic compound that is believed to have the potential to treat various ailments, including depressive mood. However, studies on the anti-depressive properties of curcumin have yielded conflicting results. This meta-analysis seeks to assess the clinical benefits of curcumin for patients with depression or depressive symptoms.

Methods

We systematically reviewed Scopus, PubMed, Web of Science, and Embase, inception February 15, 2025, to identify studies evaluating the impact of curcumin supplementation on depression in adults. The overall effects calculated using random effects model.

Results

Nineteen RCTs were included in meta-analysis. A pooled analysis of eight studies showed a significant effect for curcumin supplementation in term of improving depression (SMD: -0.76; 95% CI: -1.22, -0.30, p = 0.001; I²= 94.5%, p-heterogeneity < 0.001). Similarly, combined effect of eleven studies investigating the effect of curcumin on depressive symptoms revealed an significant improvement (SMD: -0.53; 95% CI: -0.97, -0.09, p = 0.019; I²= 85.7%, p-heterogeneity < 0.001). No evidence of publication bias was observed for depression and depressive symptoms respectively (Begg’s test = 0.428, and 0.533).

Conclusions

This meta-analysis suggests that curcumin supplementation may improve depression and depressive symptoms significantly; however, due to sensitivity analysis the observed effect was fragile. Therefore, these findings should be interpreted with caution, and further high-quality trials are needed to determine its potential clinical efficacy.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12991-026-00676-z.

Backgrounds

Depression is a common and debilitating mental health disorder that affects more than 280 million of people worldwide [1]. Characterized by persistent feelings of sadness, hopelessness, and a lack of interest in activities, depression can significantly impair an individual’s quality of life and functioning. Depression is a result of a complex interaction between biological, psychological, and social factors. In this regard, bullying has been known as psychosocial stressor. Both being a victim of bullying and perpetrating bullying have been associated with an increased risk of depressive symptoms and suicidal behavior [2]. Accordingly, most of the patients need complementary therapies to alleviate their symptoms. Moreover, COVID-19 pandemic had affected the quality of life among depressed individuals, which in turn increases vulnerability to stressful life [3]. Almost half of the major depressive patients do not response to first line treatments such as such as selective serotonin reuptake inhibitors (SSRIs) [4]. Therefore, many individuals seek alternative or complementary therapies to manage their symptoms.

One such alternative remedy that has gained attention in recent years is curcumin, a compound found in the popular spice turmeric [5]. Traditionally used in Ayurvedic medicine for its antioxidant and anti-inflammatory properties [6], curcumin has been the subject of extensive research for its potential role in managing various health conditions, including depression. As the prevalence of depression continues to rise, there is a growing interest in exploring natural remedies such as curcumin as potential treatment options. Several trials have investigated the antidepressant properties of curcumin, shedding light on its potential mechanisms of action and efficacy in managing depressive symptoms [7–10]. Besides, several meta-analyses have investigated the effects of curcumin supplementation on depression. First, in a Mini Meta-Analysis of Clinical Trials in 2015, Al-Karawi et al. showed that curcumin administration resulted in a significant reduction of major depression symptoms. Subgroup analyses indicated that curcumin had the highest effect when given to middle-aged patients, for longer durations, and at higher doses. However, the number of included trials and curcumin supplementation dosage and duration were limited [11]. Later, a 2017 meta-analysis of six clinical trials with a total of 377 patients showed that curcumin had significant clinical efficacy in improving depressive symptoms. The analysis also noted significant anti-anxiety effects in three of the trials. No adverse events were reported in any of the trials. However, the small number of included trials did not allow for a funnel plot or sensitivity analysis. Further, the duration of the included studies was limited and the generalizability of the meta-analysis was affected by the inclusion of patients with unipolar depression and those with high risk of suicide [12]. Another meta-analysis in 2019 found that curcumin, when added to standard care, can improve depressive and anxiety symptoms in individuals with depression. Small number of included studies was one of the limitations of this study [13]. More recent meta-analysis included ten trials with a total of 594 patients showed that curcumin had a significant effect in reducing depression or depressive symptoms and improved response rates. Subgroup analysis revealed that curcumin significantly reduced depression, while its effect on depressive symptoms was not statistically significant. Despite these findings, the overall quality of evidence was low [14].

While the exact mechanisms through which curcumin exerts its antidepressant effects are not fully understood, it is believed to modulate various neurotransmitters such as dopamine [15] and neurotrophic factors [16] as well as inflammatory pathways [17] implicated in the pathophysiology of depression. These findings have sparked interest in exploring curcumin as a novel therapeutic agent for depression. However, there is insufficient clinical evidence to support the use of curcumin in difficult-to-treat depression. In addition to its potential antidepressant effects, curcumin is also known for its favorable safety profile and low risk of adverse effects, making it an attractive candidate for individuals seeking natural remedies for depression [12]. However, it is important to note that the bioavailability of curcumin is relatively low, and its efficacy as a standalone treatment for depression remains to be fully elucidated [18]. Therefore, further research is needed to determine the optimal dosage, formulation, and duration of curcumin supplementation for managing depressive symptoms effectively.

In conclusion, the exploration of curcumin as a potential natural remedy for depression holds promise for expanding the treatment options available to individuals struggling with this debilitating mental health disorder. Several meta-analyses have been performed to investigate the impact of curcumin on depression [12–20]. However, there are some recent published clinical trial studies that have not pooled by any meta-analysis. Therefore, performing an updated meta-analysis seem necessary. Ultimately, this paper seeks to contribute to the growing body of knowledge on natural remedies for depression and stimulate further research to better understand the potential role of curcumin in managing depressive symptoms.

Methods

This study’s protocol has undergone approval and submission to the international prospective register of systematic reviews database (PROSPERO), assigned the registration number CRD42024512648.

Search strategies

To find relevant studies on the effects of curcumin, on depression symptoms, we conducted a comprehensive literature search in electronic databases, including PubMed, Scopus, ISI Web of Science, and Embase, until February 15, 2025. were retrieved to collect relevant RCTs. The search strategy was developed following the PICOS principle: Population - adult participants (aged 18 years and older); Intervention - curcumin or placebo; Comparator - RCT. The detailed search strategies are reported in Table Supplemental. 1. In addition to searching the database, we also checked the reference lists of previous reviews in the field to identify potentially relevant literature. Two main reviewers (VM and ESH) reviewed the titles and abstracts of articles according to the eligibility criteria. Any disagreements during the initial screening were resolved through discussion, with a final consensus agreement facilitated by a third reviewer (AHF).

Study selection

Studies with RCT design (parallel or crossover) and patients with a diagnosis of depression or exhibiting depressive symptoms are eligible. Animal studies, retrospective, prospective-cohort, cross-sectional, and pretest-post-test investigations, and studies that did not primarily or secondarily evaluate any of the previously mentioned outcomes of interest were also excluded.

Data extraction

Two researchers independently screened the literature and cross-checked their findings. When they disagreed, they consulted with each other and asked a third researcher to make the final decision. The data included information such as author, publication year, treatment details, and patient characteristics (age, sex, number of participants, diagnostic criteria, mean depression score, and the scale used).

Assessment of risk of Bias

The evaluation of risk of bias in individual studies was conducted using the revised Cochrane risk-of-bias tool (RoB2). The final judgments and the overall risk of bias were categorized as either “low,” “high,” or expressed as having “some concerns. In addition, the certainty of evidence was assessed using the GRADE standards working group. This assessment evaluated evidence quality across four categories: high, moderate, low, and very low [21].

Statistical analysis

All statistical analyses were carried out using Stata (version 14.2; Stata Corp). A random-effects model was used to calculate the pooled effect sizes, considering the expected heterogeneity across studies. To calculate the pooled effects, all data were gathered as means ± standard deviation (SD) for each variable. The standardized mean difference (SMD) and 95% confidence intervals (CI) were either extracted or computed based on the mean differences between the intervention and control groups in each study. Heterogeneity among the included studies was assessed using Cochran’s Q test and the I-squared (I²) statistic [22]. We performed subgroup analyses based on factors such as study design, curcumin dosage, duration of supplementation, and participant characteristics to explore potential sources of heterogeneity. A sensitivity analysis was conducted by removing one primary study at a time to evaluate the robustness and consistency of the results. Publication bias was not assessed, as none of the outcomes included at least ten RCTs, which is the minimum threshold recommended for funnel plot analysis and Egger’s test [23].

Results

Flow of studies

Initially, we identified 1121 records, of which 371 duplicates were removed. After reviewing the titles and abstracts, 28 records were selected for full-text assessment. Among these, nine studies were excluded due to their lack of relevance to our outcome measures or same data. After a thorough screening process, 19 trials met al.l the inclusion criteria and were deemed eligible. (Fig. 1). Table 1 presents the baseline characteristics of the population. The trials included participants aged 20 to 66, and the intervention periods varied from 4 to 24 weeks. Studies have used various measurement such as: Beck depression inventory (BDI), CDSS, Greene climacteric scale (GCS), Inventory of depressive symptomatology–self report (IDS-SR), Hamilton depression rating scale-17 (HAM-D17), Patient health questionnaire-9 (PHQ-9), Montreal cognitive assessment (MoCA-Indonesian version), Montgomery–Asberg depression rating scale (MADRS) and Depression Anxiety Stress Scales-21 (DASS-21). The dose of curcumin varied from 0.8 to 3 g/day. The health condition of the most of the included RCTs were depression [7, 8, 24–28].

Fig. 1.

PRISMA flow diagram of the study selection process

Table 1.

Characteristics of the included studies

Study	Participants, n	Health condition	Gender	Age, year	Intervention		Duration (week)	Scale
					Treatment group	Control group
Bergman et al. 2013 [25]	Int: 19, Con: 20	Depression	Both	65.8	500 mg/d curcumin	Placebo	5	HAM-D
Kanchanatawan et al. 2018 [26]	Int: 33, Con: 32	Depression	Both	42.6	500-1500 mg/d curcumin	Cellulose	12	MADRS
Lopresti et al. 2017 [27]	Int: 28, Con: 36 Int: 33, Con: 36	Depression	Both	40.39 46.53	500 mg/d curcumin 1000 mg/d curcumin	Cellulose	12	IDS-SR30
Lopresti et al. 2014 [8]	Int: 28, Con: 28	Depression	Both	44	1000 mg/d curcumin	Cellulose	8	IDS-SR30
Sanmukhani et al. 2014 [7]	Int: 18, Con: 17	Depression	Both	40	1020 mg/d curcumin + fluoxetine	Fluoxetine	6	HAM-D17
Yu et al. 2015 [24]	Int: 50, Con: 50 Int: 50, Con: 50	Depression	Men	44	1000 mg/d curcumin	Soybean powder	6	HDRS-17 MADRS
Panahi et al. 2015 [28]	Int: 61, Con: 50 Int: 61, Con: 50	Depression	Both	40.69	1000 mg/d curcumin	Placebo	6	BDI HADS
Asadi al. 2020 [33]	Int: 35, Con: 37	diabetic patients with peripheral neuropathy	Both	30-60	80 mg/d nano-curcumin	Placebo	8	DASS-21
Farshbaf-Khalili et al. 2022 [31]	Int: 26, Con: 28	Healthy Postmenopausal Women	Women	52	1000 mg/d curcumin	Microcrystalline cellulose	8	GCS
Lopresti et al. 2021 [30]	Int: 38, Con: 39	Adults with self-reported digestive complaints	Both	40.69	500 mg/d curcumin extract (Curcugen)	Microcrystalline cellulose	8	DASS-21
Mamsharif et al. 2023 [29]	Int: 35, Con: 35	Smokers	Both	33	80 mg/d nano-curcumin	Starch	12	BDI
Esmaily et al. 2015 [32]	Int: 15, Con: 15	Obese individuals	Both	38	1000 mg/d curcumin	Placebo	4	BDI
Miodownik et al. 2019 [51]	Int: 20, Con: 18	Chronic schizophrenia	Both	54	3000 mg/d curcumin	Placebo	24	CDSS
Setiawati et al. 2017 [39]	Int: 10, Con: 4	Systemic lupus erythematosus	Women	29.7	150 mg/d curcuminoids	Placebo	4	BDI
Hosseini et al. 2023 [34]	Int: 33, Con: 32	T2DM and hypertriglyceridemia	Both	55	500 mg of curcuminoids plus 5 mg of piperine	Placebo	12	DASS-21
Yaikwawong et al. 2024 [36]	Int: 113, Con: 114 Int: 113, Con: 114 Int: 113, Con: 114 Int: 113, Con: 114	Obese patients with T2DM	Both	62.26	250 mg of curcuminoids	Placebo	12, 24, 36, 48	PHQ-9
Bahrami et al. 2024 [37]	Int: 62, Con: 61	Patients with premenstrual syndrome and dysmenorrhea	Women	20.7	curcuminoid+piperine	placebo+piperine	3 successive menstrual cycles	DASS-21
Mojtahedi et al. 2024 [35]	Int: 22, Con: 20	Migraine patients	Both	39.27	nano-curcumin	placebo	8	DASS-21
Laksmidewi et al. 2024 [38]	Int: 39, Con: 39	Patients with cervical carcinoma	Women	46.74	curcumin extract	placebo	14 days on and 7 days off) between chemotherapy cycles	MoCA-Ina

Risk of Bias

Out of the 19 RCTs reviewed, twelve were deemed high quality [8, 27, 29–38]. The results of the risk of bias of included studies are shown in Table 2. Using the GRADE methodology, both outcomes were rated low for evidence quality (Table 3).

Table 2.

Results of risk of bias assessment for randomized clinical trials

Study	Randomization Process	Deviation from Intended Interventions	Selection of the Reported Result	Measurement of the Outcome	Missing Outcome Data	Overall
Bergman et al. 2013 [25]	Low	Low	Low	Low	High	High
Kanchanatawan et al. 2018 [26]	Low	Low	Low	Low	High	High
Lopresti et al. 2017 [27]	Low	Low	Low	Low	Low	Low
Lopresti et al. 2014 [8]	Low	Low	Low	Low	Low	Low
Sanmukhani et al. 2014 [7]	Low	Low	Low	High	High	High
Yu et al. 2015 [24]	Low	Low	Low	Low	High	High
Panahi et al. 2015 [28]	Low	Some concerns	Low	Some concerns	Low	High
Asadi al. 2020 [33]	Low	Low	Low	Low	Low	Low
Farshbaf-Khalili et al. 2022 [31]	Low	Low	Low	Low	Low	Low
Lopresti et al. 2021 [30]	Low	Low	Low	Low	Low	Low
Mamsharif et al. 2023 [29]	Low	Low	Low	Low	Low	Low
Esmaily et al. 2015 [32]	Low	Low	Some concerns	Low	Low	Low
Miodownik et al. 2019 [51]	Low	Low	Low	Low	High	High
Setiawati et al. 2017 [39]	Low	Some concerns	Some concerns	Low	High	High
Hosseini et al. 2023 [34]	Low	Low	Low	Low	Low	Low
Yaikwawong et al. 2024 [36]	Low	Low	Low	Low	Low	Low
Bahrami et al. 2024 [37]	Low	Low	Low	Low	Low	Low
Mojtahedi et al. 2024 [35]	Low	Low	Low	Low	Low	Low
Laksmidewi et al. 2024 [38]	Low	Low	Low	Low	Low	Low

Table 3.

Summary of findings and quality of evidence assessment using the GRADE approach

	No of patients (meta-analysis)	SMD (95% CI)	Risk of bias1	Inconsistency2	Indirectness3	Imprecision4	Publication bias5	Quality of evidence6
Depression	1658 (13)	-0.76 (-1.22, -0.30)	Not serious	Serious	Not serious	Serious	Not serious	Low
Depressive syndrome	663 (11)	-0.53 (-0.97, -0.09)	Not serious	Serious	Not serious	Serious	Not serious	Low

BMI, Body mass index; FBS, Fasting blood sugar; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance; TC, Total cholesterol; LDL-c, Low-density lipoprotein; HDL-c, High-density lipoprotein; TG, Triglyceride

¹ Risk of bias based on the Cochrane risk of bias tool

² Downgraded if there was a substantial unexplained heterogeneity (I² > 50%, P < 0.10) that was not explained by subgroup analyses

³ Downgraded if there were factors present relating to the participants, interventions, or outcomes which restrict the overall generalizability of the results

⁴ Downgraded if the 95% confidence interval (95% CI) crossed the minimally important difference (MID)

⁵ Downgraded if there was an evidence of publication bias using funnel plot

⁶ Since all included studies were randomized controlled trials, the certainty of the evidence was graded as high for all outcomes by default and then downgraded based on prespecified criteria

Curcumin on patients with depression

The finding of eight RCTs with 14 arms revealed a significant reduction on depression following curcumin administration (SMD: -0.76; 95% CI: -1.22, -0.30, p = 0.001; I²= 94.5%, p < 0.001) (Fig. 2). Sensitivity analysis using leave-one-out approach indicated that no study may influence the overall effect. Visual inspection of funnel plot showed an asymmetry (Fig. 3). However, Egger’s and Begg’s regression test suggesting no small-study effect (p = 0.333, 0.428, respectively).

Fig. 2.

Forest plot with mean difference and 95% confidence intervals (CIs), the effects of curcumin supplementation on depression

Fig. 3.

Funnel plot assessing publication bias for studies on curcumin and depression

Curcumin on patients with depressive symptoms

Combining the data from eleven trials showed the significant reducing effect of curcumin supplementation on depressive symptoms (SMD: -0.53; 95% CI: -0.97, -0.09, p = 0.019; I²= 85.7%, p < 0.001) (Fig. 4). Sensitivity analysis showed that excluding two studies change the results of curcumin on depressive symptoms to non-significant model [35, 39]. Visual inspection of funnel plot showed an asymmetry (Fig. 5). While, Begg’s test indicated to evidence of publication bias (p = 0.533).

Fig. 4.

Forest plot detailing mean difference and 95% confidence intervals (CIs) of the effects of curcumin supplementation on depressive symptoms

Fig. 5.

Funnel plot assessing publication bias for studies on curcumin and depressive symptoms

Subgroup analysis

According to the subgroup analysis results for the effects of curcumin on depression, in participants aged > 45 years, curcumin significantly improved depression (SMD: -1.01; 95% CI: -1.27, -0.75; p < 0.001; I² = 71.6%, p = 0.003). Besides, studies with ≥ 8 weeks of supplementation showed a significant reduction in depression (SMD: -0.97; 95% CI: -1.22, -0.72; p < 0.001; I² = 73.8%, p = < 0.001). Curcumin supplementation at doses < 1000 mg/day significantly reduced depression (SMD: -1.07; 95% CI: -1.29, -0.84; p < 0.001; I² = 63.0%, p = 0.013). Moreover, subgroup analysis based on sample size indicated that both lower and larger sample size curcumin can show its beneficial reducing effect on depression. According to the depression diagnosis criteria, studies using the IDS-SR30 (SMD: -0.39; 95% CI: -0.68, -0.10; p = 0.008) and PHQ-9 (SMD: -1.22; 95% CI: -1.36, -1.08; p < 0.001) showed a significant reduction in depression value. Also, both low- and high-quality studies showed a significant effect on depression outcome (Table 4) (Supplementary figures).

Table 4.

The summary of the Subgroup analysis for included outcomes

Group	No. of comparisons	SMD (95% CI)	P-value	I2 (%)	P-heterogeneity
Curcumin on depression
Total	13	-0.76 (-1.22, -0.30)	0.101	94.5	<0.001
Age(year)
≤45	7	-0.60 (-1.52, 0.32)	0.156	96.4	<0.001
>45	6	-1.01 (-1.27, -0.75)	<0.001	71.6	0.003
Duration (week)
<8	5	-0.48 (-1.75, 0.79)	0.339	97.4	<0.001
≥8	8	-0.97 (-1.22, -0.72)	<0.001	73.8	<0.001
Dose(mg/day)
<1000	7	-1.07 (-0.82, 0.13)	<0.001	63.0	0.013
≥1000	6	-0.47 (-1.49, 0.55)	0.231	96.7	<0.001
Sample size
≤65	4	-0.64 (-1.11, -0.16)	<0.001	66.4	0.030
˃65	9	-0.81 (-1.40, -0.22)	0.317	96.1	<0.001
Outcome Measures
HDRS	3	-0.71 (-2.79, 1.36)	0.358	98.0	<0.001
IDS-SR	3	-0.39 (-0.68, -0.10)	<0.001	0.0	0.970
MADRS	2	-0.21 (-2.43, 2.00)	0.829	97.6	<0.001
BDI	1	-1.17 (-1.58, -0.77)	<0.001	-	-
PHQ-9	4	-1.22 (-1.36, -1.08)		0.0	1.0
Quality studies
Low	10	-0.86 (-1.42, -0.31)	0.248	95.6	<0.001
High	3	-0.39 (-0.68, -0.10)	<0.001	0.0	0.970
Curcumin on depressive symptoms
Total	11	-0.53 (-0.97, -0.09)	0.039	85.7	<0.001
Age(year)
≤45	6	-1.02 (-1.86, -0.17)	0.190	91.6	<0.001
>45	5	-0.15 (-0.48, 0.17)	0.081	50.8	0.087
Duration (week)
<8	2	-1.10 (-4.09, 1.89)	0.498	91.5	0.001
≥8	7	-0.69 (-1.28, -0.09)	0.016	88.0	<0.001
Intermittent intervals	2	0.01 (-0.52, 0.54)		71.2	0.063
Dose(mg/day)
<1000	8	-0.79 (-1.36, -0.21)	<0.001	89.3	<0.001
≥1000	3	0.03 (-0.33, 0.38)	0.231	0.0	0.608
Sample size
≤65	6	-1.03 (-2.05, -0.01)	0.335	91.8	<0.001
˃65	5	-0.21 (-0.49, 0.06)	0.012	50.5	0.088
Outcome Measures
DASS	5	-0.96 (-1.72, -0.19)	0.032	91.5	<0.001
BDI	3	-0.59 (-1.72, 0.54)	0.220	83.0	0.003
GCS	1	-0.06 (-0.59, 0.48)	0.581	-	-
CDSS	1	-0.11 (-0.74, 0.53)	0.417	-	-
MoCA-Ina	1	0.30 (-0.15, 0.74)		-	-
Quality studies
Low	5	-1.01 (-1.87, -0.14)	0.297	92.2	<0.001

Subgroup analysis was conducted based on age, supplementation dosage and duration, sample size, and type of diagnosis tool. Accordingly, it has been demonstrated that younger adults (≤ 45 years) (SMD: -1.02; 95% CI: -1.86, -0.17; p = 0.018; I² = 85.7%, p < 0.001), supplemented by lower doses (< 1000 mg/day) (SMD: -0.79; 95% CI: -1.36, -0.21; p = 0.007; I² = 89.3%, p < 0.001), for continues long-term (< 8 weeks) (SMD: -0.69; 95% CI: -1.28, -0.09; p = 0.024; I² = 88.0%, p < 0.001) which were assessed by DASS-21 scale (SMD: -0.96; 95% CI: -1.72, -0.19; p = 0.014) had reduced level of depressive symptoms. However, studies with low quality tended to show significant effect of curcumin supplementation (SMD: -1.01; 95% CI: -1.87, -0.14; p = 0.047) (Table 4) (Supplementary figures).

Discussion

The current updated meta-analysis summarized the results of nineteen studies investigating the effect of curcumin on depression and depression-related symptoms. The pooled analyses of eight studies on patients with depression and obese adults with T2DM, revealed that curcumin supplementation has significant reducing effect on depression. Similarly, it was accompanied with favorable effects on depression symptoms too. Funnel plot inspection suggested asymmetry in both main analyses, whereas Egger’s and Begg’s tests did not indicate statistically significant small-study effects. However, these tests are known to have low statistical power when fewer than 10–20 studies are included, and therefore the absence of statistical significance does not exclude publication bias. The larger effect sizes observed in smaller trials further raise the possibility of small-study effects or selective reporting. Consequently, the pooled estimates should be interpreted with caution, and future large-scale, adequately powered RCTs are required to minimize the influence of publication bias and imprecision.

Among previous meta-analysis studies, only Wang et al.‘s study [14] had separated the analysis for major depressive disorder and depressive symptoms in subjects with various health status. Also, our study has 3 additional studies compared to previous studies [29, 40, 41], which caused significant results in previous meta-analyses to be violated. The study by Xiang Ng et al. [12] and Wang et al. [14] did not perform subgroup analysis based on demographic and intervention characteristics. The results of subgroup analysis in the rest of the studies led to almost similar results to those obtained by our study. The study of Al-Karawi et al. [11] showed that curcumin supplementation in a long-term duration (≥ 6 weeks vs. <6 weeks) led to the more ameliorating effect on depression. However, in terms of mean age and dose, due to the existence of only one study in some subgroups, their results cannot be certain. Similarly, in the study of Fusar-Poli et al., long-term supplementation (> 6 weeks vs. ≤6 weeks) in low doses (≤ 500 mg/day vs. >500 mg/day) had better effects. In our meta-analysis, in the studies on depression symptoms, two types of nano-curcumin and conventional curcumin have been prescribed. However, only conventional curcumin was used on major depressive disorder. Karthikeyan et al. showed that nano-curcumin led to a significant improvement in depression symptoms due to better bioavailability. It has about 9 times more bioavailability than conventional curcumin [42]. In patients with depression, the duration of supplementation was less than 8 weeks in most of the studies where curcumin was prescribed at ≥ 1000 mg/day. Consequently, the non-significant effect of doses higher than 1000 mg per day can be attributed to the short duration of supplementation. It should be noted that curcumin is registered as Generally Recognized as Safe (GRAS) by the U.S. Food and Drug Administration (FDA), even at doses up to 8 g/day [43]. As a result, the higher doses and long-term use of the supplement cannot have side effects. Included studies reported mild-to-moderate digestive, dermatological, visual, auditory, oral, and neurological symptoms in some cases. However, no serious side effects were reported in any of the included studies. By combining our results and previous studies, we can conclude that a lower dose of curcumin and a longer duration of supplementation in nano form may have a better effect on depression. However, if higher doses are administered over a longer period of time, they may have a significant effect.

Various scales have been used to assess depression in the included studies, including HDRS, IDS-SR, MADRS, BDI, DASS, and CDSS. These measurements represents different purposes, such as screening, diagnosis, differential diagnosis, evaluating severity, monitoring treatment effects, and observational assessment [44]. There are inconsistencies in the emphasis placed on specific symptom domains across these scales. MADRS and DASS allocate over 15% of their items to the “Mood & Outlook” domain, whereas BDI assigns less than 10% to this category. In contrast, BDI places the greatest emphasis on “Confidence & Self-Judgment” [45]. In our analysis, curcumin led to a significant improvement in IDS-SR and BDI. Both have reasonable internal validity and are sensitive in measuring treatment outcome compared to the HDRS [46].

The precise mechanism behind the antidepressant effect of curcumin is not completely understood, but it is believed to involve several different pathways in the brain. Curcumin has been shown to modulate levels of key neurotransmitters, including serotonin, dopamine, and norepinephrine, which play critical roles in mood regulation [47]. In addition, curcumin exhibits potent anti-inflammatory properties [48], and chronic inflammation has been implicated in the development of depression [49], its anti-inflammatory effects may contribute to alleviating depressive symptoms. Curcumin also promotes the production of brain-derived neurotrophic factor (BDNF), a protein that supports neuronal growth, survival, and synaptic plasticity; low BDNF levels are associated with depression, and enhancing BDNF may improve both mood and cognitive function [50]. All these together, these mechanisms suggest that curcumin exerts multifaceted effects on neural pathways involved in mood regulation and may provide therapeutic benefits in depression.

We conducted a sensitivity analysis to assess the robustness of our findings, and the results indicated that the exclusion of specific studies significantly influenced the overall effect of curcumin on depressive symptoms. Despite extensive subgroup and sensitivity analyses, substantial heterogeneity remained across pooled estimates. This residual heterogeneity likely reflects clinically meaningful differences that could not be fully explored due to limited and inconsistent reporting across trials. Potential contributors include variability in baseline depression severity, the use of concomitant antidepressant medications, differences in comorbidity burden (e.g., obesity or metabolic disorders), heterogeneity in curcumin formulations and bioavailability (including the co-administration of piperine), as well as differences in intervention duration and outcome assessment instruments. Collectively, these factors may influence both the magnitude and consistency of treatment effects and underscore the need for future trials with standardized reporting and more homogeneous study designs.

The current study had some strengths. The current meta-analysis adds novelty by: Including newly published clinical trials that were not assessed in previous meta-analyses; Performing subgroup analyses based on age, dosage, intervention duration, and type of curcumin formulation to explore potential sources of heterogeneity; And applying rigorous statistical methods to synthesize the evidence and provide a more comprehensive evaluation of curcumin’s effects on depression and depressive symptoms. Thus, while curcumin has been studied in the context of depression, the novelty of our work lies in providing an updated, systematic, and methodologically robust analysis that clarifies the clinical significance of curcumin supplementation.

Several limitations of the present meta-analysis should be acknowledged. First, substantial heterogeneity was observed across pooled analyses, likely due to differences in study populations, baseline health conditions, curcumin dosage, formulation, intervention duration, and outcome assessment tools. Second, subgroup analyses indicated that the significant effects of curcumin on both depression and depressive symptoms were mainly driven by low-quality studies, which limits the strength and certainty of the conclusions. Third, studies enrolling patients with major depressive disorder exclusively used conventional curcumin formulations; therefore, the potential efficacy of high-bioavailability forms such as nano-curcumin in this population remains unclear. Several recent studies on depressive symptoms in other populations have employed nano-curcumin formulations with markedly higher bioavailability [29, 33]. Although findings from these trials suggest potential superior efficacy of nano-curcumin, no RCTs have directly compared conventional and nano-curcumin in patients with MDD. This represents a critical knowledge gap that limits clinical translation, highlighting the need for future trials to evaluate the relative efficacy and optimal formulation of curcumin in patients with formally diagnosed major depressive disorder. Fourth, considerable variability existed in the measurement instruments used to assess depression and depressive symptoms, each emphasizing different symptom domains, which may have influenced effect size estimates. Fifth, several subgroup analyses were based on a small number of trials, reducing statistical power and limiting generalizability. Finally, asymmetry in funnel plots suggests a possible risk of publication bias, although formal statistical tests did not consistently confirm small-study effects. Taken together, these limitations highlight the need for future well-designed, large-scale RCTs using standardized diagnostic criteria, validated outcome measures, and high-bioavailability curcumin formulations administered over longer durations. These methodological limitations, combined with the low GRADE certainty, mean that the pooled estimates must be interpreted with caution.

Conclusion

Curcumin supplementation was associated with a significant reduction in both clinically diagnosed depression and depressive symptoms; however, the overall pooled effect was fragile, becoming non-significant after the removal of two trials using sensitivity analysis. Subgroup analyses indicated that beneficial effects on depression were more pronounced among participants aged over 45 years, with doses below 1000 mg/day and intervention durations of at least 8 weeks. Similarly, reductions in depressive symptoms were more evident in studies with longer supplementation periods. Overall, these findings suggest that the antidepressant effects of curcumin may be influenced by age, dosage, duration of intervention, curcumin formulation, and study quality.

From a clinical perspective, the present findings support curcumin as a potentially safe and well-tolerated adjunctive strategy for managing depressive symptoms, rather than a stand-alone treatment. However, due to variability in study design, outcome measures, and participant characteristics, these results should be generalized with caution. Future well-designed randomized controlled trials using standardized diagnostic criteria, validated outcome measures, and optimized curcumin formulations are required to establish definitive clinical recommendations.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Abbreviations

SSRIs

Selective serotonin reuptake inhibitors

SD

Standard deviation

SMD

Standardized mean difference

CI

Confidence intervals

BDI

Beck depression inventory

GCS

Greene climacteric scale

IDS-SR

Inventory of depressive symptomatology–self report

HAM-D17

Hamilton depression rating scale-17

PHQ-9

Patient health questionnaire-9

MoCA-Indonesian version

Montreal cognitive assessment

MADRS

Montgomery–Asberg depression rating scale

DASS-21

Depression Anxiety Stress Scales-21

GRAS

Generally recognized as safe

FDA

Food and Drug Administration

BDNF

Brain-derived neurotrophic factor

Author contributions

V.M. designed the study and corporate in the interpretation of results. A.H.F. and M.F. contributed to data collection. M.M. conducted data analysis. M.A. assisted in methodology development, supervised the project and provided critical revisions. F.S. supervised the project and provided critical revisions. All authors contributed to the writing, reviewed, and approved the final manuscript.

Funding

None.

Data availability

The original data used during the current study can be obtained by contacting the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.