Neuromelanin levels in individuals with substance use disorders: A systematic review and meta-analysis

Dopamine ’ s role in addiction has been extensively studied, revealing disruptions in its functioning throughout all addiction stages. Neuromelanin in the substantia nigra (SN) may reflect dopamine auto-oxidation, and can be quantified using neuromelaninsensitive magnetic resonance imaging (neuromelanin-MRI) in a non-invasive manner.In this pre-registered systematic review, we assess the current body of evidence related to neuro-melanin levels in substance use disorders, using both post-mortem and MRI examinations. The systematic search identified 10 relevant articles, primarily focusing on the substantia nigra. An early-stage meta-analysis (n = 6) revealed varied observations ranging from standardized mean differences of (cid:0) 3.55 to + 0.62, with a pooled estimate of (cid:0) 0.44 (95 % CI = (cid:0) 1.52, 0.65), but there was insufficient power to detect differences in neuro-melanin content among individuals with substance use disorders. Our gap analysis highlights the lack of sufficient replication studies, with existing studies lacking the power to detect a true difference, and a complete lack of neuromelanin studies on certain substances of clinical interest. We provide recommendations for future studies of dopaminergic neurobiology in addictions and related psychiatric comorbidities.


Introduction
The role of dopamine in addiction has been well studied for the past 40 years (Solinas et al., 2019).Research has confirmed that dopamine functioning is disrupted in all three stages of addiction (Volkow and Morales, 2015;Wise and Koob, 2014) -binge/intoxication, withdrawal/negative affect, and preoccupation/anticipation -each of which is governed by distinct neurobiological mechanisms (Wise and Koob, 2014).With initial use of the drug, dopamine release at the ventral striatal regions produces a rewarding response, and as drug use becomes more habitual, dopamine aberration shifts from the ventral to the dorsal striatal regions, which are associated with habit formation (Volkow and Morales, 2015).Withdrawal is associated with negative affect and decreased dopamine activity, as well as increased activity of the stress system (Koob, 2008;Melis et al., 2005).Drug cravings involve dopamine release occurring in key limbic regions associated with emotions and memory (Koob and Volkow, 2010).
The ventral tegmental area (VTA) and the substantia nigra (SN) may be promising candidates to reflect how substance use affects the dopamine system across key subcortical regions, as they are the primary dopamine-producing nuclei in the brain (Swanson, 1982;Björklund and Dunnett, 2007).The VTA is a dopaminergic structure of the midbrain that plays a key role in substance use disorders (SUDs).Research has shown that the VTA is necessary in the stress-, cue-, and drug-primed reinstatement in rodents self-administering both cocaine (Mahler et al., 2013;McFarland and Kalivas, 2001;McFarland et al., 2004;Kalivas and McFarland, 2003) and heroin (Stewart, 1984;Bossert et al., 2004;Wang et al., 2012).Excitation of VTA neurons induces conditioned place preference in mice (Tsai et al., 2009) and reinstates reward-seeking behaviours (Adamantidis et al., 2011;Kim et al., 2012).The dopaminergic neurons of the SN are also implicated in substance use.With excitation of the dopaminergic neurons in the SN, mice learn intracranial self-stimulation as well (Rossi et al., 2013).Hence, dopaminergic neurons in the midbrain play an important role in substance use.
Neuromelanin is a pigment found in the dopaminergic neurons of the SN, VTA, and locus coeruleus (LC) (Zucca et al., 2014).In brief, neuromelanin is generated from excess catecholamines within the cytosol and this process is promoted by iron as the two molecules bind together to form neuromelanin (Zucca et al., 2014).Neuromelanin accumulates in the human brain throughout the lifespan and is believed to have neuroprotective functions; however, in disorders in which dopamine signaling is impacted, this accumulation likely crosses a critical threshold for cellular damage and can result in negative consequences for the dopamine pathway (Sulzer et al., 2000).In individuals with Parkinson's disease, a condition characterized by SN dopamine neuronal death, the degree of dopamine neuron degeneration corresponds to the loss of neuromelanin-pigmented neurons, suggesting that neuromelanin serves as a proxy for midbrain dopaminergic tone (Hirsch et al., 1988).Direct comparisons between post-mortem neuromelanin-sensitive magnetic resonance imaging (neuromelanin-MRI) and neuropathological findings have demonstrated a close association between neuromelanin-MRI signal intensity in the SN and the quantity of neuromelanin-containing neurons, affirming the link between neuromelanin-MRI hyperintensity and the presence of neuromelanin (Kitao et al., 2013).Recent research with neuromelanin-MRI has provided further evidence for the use of neuromelanin as a proxy measure for dopaminergic pathway abnormalities, even in the absence of neurodegeneration.For example, neuromelanin-MRI signal is correlated with levels of dopamine in the striatum, and blood flow of the SN in healthy controls (HCs) and individuals with schizophrenia (Cassidy et al., 2019).
Correlations have been found between striatal dopamine release and the neuromelanin signal in the SN (Cassidy et al., 2019) and research has demonstrated that striatal dopamine transporter binding is highly correlated with postmortem SN cell counts, confirming the validity of dopamine transporter imaging as reflective of nigrostriatal dopaminergic degeneration (Kraemmer et al., 2014).Additionally, in individuals with Parkinson's disease, there is a positive correlation between neuromelanin contrast-to-noise ratio and striatal dopamine transporter density (Isaias et al., 2016).Therefore, the integrity of the dopamine signal in the SN/VTA complex may serve as an index of subcortical dopaminergic changes due to substance use, such that in some conditions it may index alterations in dopamine function or alternatively, degeneration of dopamine neurons.
Studies show that SUDs may relate to aberrations in dopamine transporters throughout the brain, including in the SN.For example, in persons who use methamphetamine, a significant loss of dopamine terminal markers (dopamine, tyrosine hydroxylase, and dopamine transporters) have been found in the post-mortem striatum (Wilson et al., 1996).Positron emission tomography studies have further shown persistently decreased dopamine transporter density in those with a history of methamphetamine and methcathinone use, further suggesting loss of dopamine transporters or dopamine terminals (McCann et al., 1998).In addition to the striatum, many studies indicate that methamphetamine use results in a loss of dopaminergic neurons in the SN (Ares-Santos et al., 2013, 2014;Granado et al., 2011Granado et al., , 2013)).Due to dopaminergic cell death and damage to dopamine transporters, neuromelanin levels will decrease, as is seen in Parkinson's disease (Martin-Bastida et al., 2017).Research on alcohol use disorder also shows dopaminergic abnormalities.Acute alcohol consumption increases striatal dopamine release (Heinz et al., 2004) and chronic alcohol use has been found to result in reduced striatal dopamine receptor availability (Heinz, 2002).Kamp and colleagues showed a reduction in dopamine D2/D3 striatal receptors in alcohol use, indicating a key role of the dopamine system in individuals with an alcohol use disorder (Kamp et al., 2019).Similarly, administration of tetrahydrocannabinol, the primary addictive ingredient in cannabis, induces dopamine release in the striatum (Bossong et al., 2015).Individuals who use cannabis present with a blunting of dopamine release in the striatum (Bloomfield et al., 2014) and a reduction of dopamine synthesis capacity (Murray et al., 2014).Thus, commonly used substances such as alcohol and cannabis, are associated with changes in dopamine receptors and signaling; however, we do not know how these substances may affect neuromelanin.
In individuals who use cocaine, a significantly increased neuromelanin-MRI signal was seen in the SN compared with HCs (Cassidy et al., 2020).However, Jarcho et al. (2022) found lower neuromelanin-MRI signal within the SN with more severe substance use disorder scores (Jarcho et al., 2022).Studies also examine neuromelanin in other dopaminergic areas of the brain.Wang et al. (2021) found that individuals with cocaine use disorder expressed a higher neuromelanin-MRI signal within the LC compared to those without this disorder, though no differences were seen between groups in the VTA/SN.Other substances have been included in neuromelanin studies as well.Kaiya (1980) examined the effects of neuroleptics on post-mortem neuromelanin levels, including propericiazine, levomepromazine, carpipramine, haloperidol, chlorpromazine, perphenazine, and perazine.This study found no significant difference in neuromelanin content between individuals who used neuroleptics and those who did not (Kaiya, 1980).To date, no systematic reviews or meta-analyses have been conducted to examine neuromelanin levels in those with a SUD or a history of substance use, and due to the variable results found in these studies, this is needed to determine what future research needs to examine within this field.
The objective of this paper was to synthesize previous studies examining neuromelanin levels in those with a SUD with varying levels of severity, and to examine the differences in neuromelanin levels among different substances and measurement approaches through a systematic review and meta-analysis.Based on prior observations reported above, we expected neuromelanin levels to be reduced in individuals with a SUD compared to those without a past or current SUD and decrease with higher substance use, especially stimulants.We also examined the influence of age and sex on differences in neuromelanin levels or neuromelanin-MRI signals.Note that we use the term 'individuals with substance use disorder' as per the recommendations on non-handicapping language from the National Institute of Drug Abuse; this term may differ from the original descriptions provided by authors of various primary manuscripts when describing the respective study populations.

Methods
The protocol for this study was uploaded to the International Prospective Register of Systematic Reviews website (CRD42023394990).

Study search
A systematic search of the literature was conducted to identify studies examining neuromelanin in individuals with a past or present substance use disorder (both formally diagnosed or self-reported) (Fig. 1).We conducted the search using Google Scholar (1960-November 20th, 2023), PubMed (1951-November 20th, 2023), and Ovid (1947to November 20th, 2023).Search terms were: (neuromelanin OR melanin) AND (substance use OR substance use disorder OR substance abuse OR drug abuse OR drug dependence OR addiction).Authors JA and FZ conducted the literature search and used Covidence to exclude articles according to the a priori criteria.Articles were included if they: (1) measured brain neuromelanin levels (postmortem or in vivo using MRI); (2) included human subjects reporting a past and/ or present SUD irrespective of the severity; and (3) were published in English.

Statistical analyses
All statistical analyses were carried out in RStudio Version 2023.12.1.402(Posit team, 2024).We compared our outcome of interest, neuromelanin levels, between those with a past / current SUD and those without a past / current SUD using the standardized mean difference (SMD).We included studies where SUD was identified using diagnostic criteria (e.g., DSM) as well as rating scale-based cut-offs (e.g., self-report of substance use).SMD and 95 % confidence intervals (CIs) were used as the meta-analysis summary statistic.The regions of interest (ROIs) were the SN, LC, and VTA.If an ROI was included in fewer than four studies, it was not included in the meta-analysis.Effect sizes > 0.2 were interpreted as small, > 0.5 were moderate, and > 0.8 were interpreted as large (Cohen, 1988).To examine heterogeneity between studies, the I 2 statistic was used; an I 2 statistic greater than 50 % was the threshold used to identify significant heterogeneity, based upon the Cochrane guidelines (Higgins et al., 2003).When the heterogeneity threshold was met, the random effects model was used, as compared to the fixed effects model that was used when the heterogeneity threshold was not met.To examine whether neuromelanin-MRI and ex-vivo methods are just as likely to detect levels of neuromelanin, we performed subgroup analyses based on the study method.Subgroup analyses were also conducted to compare drugs of use.Power analyses were conducted using the RStudio package, metapower, on the main meta-analysis and the subgroup analyses (Cuijpers et al., 2021).Meta-regressions were conducted to assess the relationship between deviated neuromelanin levels and (1) average participant age; (2) male: female ratio; and (3) publication year.Studies did not include information regarding amount of substance use; therefore, we were unable to test associations between substance dose and neuromelanin outcomes.
Newcastle-Ottawa Quality Assessment Scale scores ranged from 1 to 5, with an average score of 3.3 out of 6, suggesting that the quality of studies were average overall (Table 1).

Neuromelanin-MRI studies
Tavares and colleagues found no significant differences in the contrast ratio of neuromelanin in the SN/VTA of first episode psychosis patients who consumed illicit substances (mean = 1.11 (standard deviation (SD) = 0.04) versus 1.12 (0.03) for patients who did not use illicit substances; p = 0.331) (Tavares et al., 2018).Cassidy and colleagues found that persons with a cocaine use disorder showed significantly increased neuromelanin-MRI signal across the SN compared with persons without a cocaine use disorder (t = 2.07, df = 49, p = 0.044, Cohen's d = 0.62, 95 % CI = 0.19, 1.12) but the signal was not correlated with duration of use (ρ = -0.33,p = 0.18) or dollars spent per week on cocaine (ρ = -0.08,p = 0.74) (Cassidy et al., 2020).In Jalles et al., first episode psychosis patients with a SUD had a significant increase in SN/VTA neuromelanin-MRI signal and although not statistically significant, the LC had greater neuromelanin-MRI signal variability in these psychosis patients, particularly those with a SUD (Jalles et al., 2020).In Wang et al. (2021), persons with a cocaine use disorder had a higher neuromelanin-MRI signal in the LC relative to those without a cocaine use disorder; however, there was no correlation between the signal and years of cocaine use or with life-time amount of cocaine use (r = 0.01, p = 0.969).The neuromelanin signal of the VTA/SN pars compacta did not show significant differences between persons with a cocaine use disorder and those without a cocaine use disorder (Wang et al., 2021).Finally, in Jarcho et al. (2022), voxelwise analysis revealed SUDs were negatively correlated with neuromelanin-MRI signal in a significant number of voxels within the SN/VTA complex (416 of 1807 SN voxels, P corrected < 0.007).Kariks (1978) found that people with chronic alcohol use often had neuronal degeneration of the SN but not in the LC.In four cases where the damage was widespread, complete necrosis of the SN neurons was noted; however, no control group was used to compare if this SN degradation or necrosis was present in those who did not have chronic alcohol use as well.Additionally, this was reported qualitatively, indicating, for example, that "in each of the 40 cases examined, either mild, moderate, or severe neuronal depigmentation was present" (Kariks, 1978).Therefore, this study was not included in quantitative synthesis.Kaiya (1980) found no significant differences in neuromelanin content between brains of individuals who took neuroleptics versus individuals who did not, but neuromelanin content in the SN and the LC of the medicated brains tended to decrease in the sixth decade (Kaiya, 1980).There was no significant correlation between neuromelanin content and total dosage or administered duration in the SN or LC (Kaiya, 1980).Reyes et al. (1991) found that individuals with an intravenous SUD had lower neuromelanin volume density than individuals without a SUD.In Arango et al. (1994), a 23 % reduction in the total number of LC neurons was found in individuals with alcohol use disorder compared to controls (p = 0.0002).Little et al. (2009) found the estimated total number of melanized dopamine cells in the anterior midbrain was significantly reduced in persons with a cocaine use disorder versus HCs by 16 % (t = 2.600, df = 17, P = 0.02, two-tailed).There were trends toward decreases of both dorsal tier and ventral tier melanin-containing cells; however, these comparisons were not statistically significant (ventral tier melanin cells: t = 1.775, df = 17, P = 0.09, two-tailed; dorsal tier melanin cells: t = 1.781, df = 17, P = 0.09, two-tailed) (Little et al., 2009).
The total sample was composed of 105 individuals with a past or current SUD and 134 individuals with no past or current SUD.There were no significant increases or decreases of SN/VTA neuromelanin in those with a SUD compared to those without a SUD (SMD = -0.44;95 %

Table 1
The Newcastle-Ottawa Quality Assessment Scale of included studies.

Author (Year)
Selection Comparability Total Kariks et al. (1978) (Kariks, 1978)  CIs = − 1.52, 0.65; p > 0.05).Between-study heterogeneity was notable (I 2 = 90.31,p < 0.001), therefore a random effects model was used.A post hoc power analysis found the power of finding a significant difference here to be 0.66; thus, there was a moderate likelihood of detecting a true effect.Four studies examined neuromelanin levels in the LC (Wang et al., 2021;Kaiya, 1980;Arango et al., 1994;Jalles et al., 2020); however, 1 study (Jalles et al., 2020) lacked quantitative neuromelanin values for the LC.Thus, the LC was not included in the meta-analysis.Table 2

Subgroup analyses
The subgroup analysis of neuromelanin-MRI studies (n = 4) revealed no significant deviations of neuromelanin signal in those with a SUD compared to those without a SUD in the SN/VTA (SMD = 0.11; 95 % CIs = − 0.20, 0.42; p > 0.205) (Cassidy et al., 2020;Wang et al., 2021;Jalles et al., 2020;Tavares et al., 2018).A power analysis revealed that the estimated power for detecting a significant subgroup effect here was only 0.09.A subgroup analysis of studies in which participants used cocaine (n = 3) did not indicate any significant deviation in neuromelanin levels in individuals with a cocaine use disorder compared to those without (SMD = − 0.84; 95 % CIs = − 3.33, 1.65; p > 0.05) (Jarcho et al., 2022;Wang et al., 2021;Little et al., 2009).Here, there was moderate power (0.71) to detect any subgroup effects if they were present.Thus, we cannot rule out type-II errors in the subgroup analyses.Table 3

Discussion
To our knowledge, this is the first systematic review and meta-analysis to examine neuromelanin deviations in individuals with SUDs or with varying levels of substance use.Our results indicate that: 1) overall, the studies investigating this question are limited in number, with limited sample size; 2) substances studied to date do not exert a differential effect on SN/VTA neuromelanin signal; and 3) the effect of cannabis and nicotine are yet to be systematically studied.While both the pooled results of post-mortem studies and MRI studies were not statistically significant in the meta-analysis, the direction of the pooled results were opposite for the two modalities, with postmortem studies indicating a possible reduction of neuromelanin in persons with a substance use disorder, while MRI studies finding a numerically higher neuromelanin-MRI signal in individuals with a SUD than in those without.Based on the largest reported effect sizes in each modality, postmortem studies need n = 6 subjects to detect the same effect 80 % of the time, while MRI studies need n = 84 subjects; thus, most individual MRI studies were underpowered.
Our systematic review highlights several key gaps.It shows that more research needs to be conducted on how substance use affects neuromelanin levels; specifically, no studies looked at neuromelanin in relation only to cannabis or nicotine/tobacco use.Jarcho and colleagues (Jarcho et al., 2022) had a subset of participants that used cannabis (17 of 33), but these individuals were grouped with those with an alcohol use disorder and those that had used hallucinogens; therefore, any neuromelanin differences due to cannabis consumption cannot be deducted.Based on studies examining the dopaminergic changes after cannabis use, we can expect neuromelanin levels to be altered as acute tetrahydrocannabinol increases dopamine release, and long-term use has an association with reduced dopaminergic functioning (Bloomfield et al., 2016).Similarly, there is conflicting evidence on the dopaminergic changes implicated with nicotine use.Imaging studies show that striatal dopamine release occurs with nicotine administration in individuals who are dependent on nicotine (Barrett et al., 2004;Montgomery et al., 2007); however, this is not always reproducible (Montgomery et al., 2007).Further, decreased dopamine D1 receptor density in the ventral striatum of individuals who smoke cigarettes has been reported, which suggests that smoking could lead to a chronically underactive mesolimbic dopamine system (Dagher et al., 2001; Yasuno Fig. 2. Study effect sizes and 95 % confidence intervals of neuromelanin level differences in the substantia nigra/ventral tegmental area complex between those with a past or current substance use disorder and those without.The size of each study's marker is proportional to the sample size.Studies with a circle marker used postmortem techniques and those with a square marker used neuromelanin- MRI. et al., 2007).
Our review also found that the LC neuromelanin differences in individuals with a SUD were not consistent.Arango and colleagues (1994) found that patients with alcohol use disorder had significantly fewer LC neurons than controls (Arango et al., 1994).Kaiya and colleagues (1980) did not find significant differences in LC neuromelanin content in those taking neuroleptics compared to controls (Kaiya, 1980).This study did include individuals with alcohol use disorder; however, no results pertaining to differences in neuromelanin between individuals with versus without alcohol use disorder were reported (Kaiya, 1980).Although not statistically significant, Jalles et al. (2020) found that the LC had greater signal intensity variability in patients with psychosis, even more so in patients with a substance use disorder, and Wang et al. (2021) found that participants with cocaine use disorders had a significantly higher neuromelanin signal in the LC compared to those without.Focus on LC neuromelanin in the future could allow us to examine how drug use impacts noradrenergic turnover, in addition to dopaminergic changes.Little et al. (2009) examined the anterior midbrain, separating the brain regions into ventral tier, dorsal tier, and medial, non-compacta.For the purposes of our review, we included all of these areas as the SN/VTA [Findings] ↑: neuromelanin increase in those with a substance use disorder compared to those without, ↓: neuromelanin decrease in those with a substance use disorder compared to those without, n.s.: no significant difference between those with a substance use disorder and those without.n/a: not applicable.Abbreviations: neuromelanin-MRI, neuromelanin-sensitive magnetic resonance imaging; GRE, gradient echo; SMD, standardized mean difference.* Letter to the editor.a substance use severity was notably lower than reported in clinical populations.
complex (Fig. 4).Additionally, Cassidy et al. (2020), Wang et al. (2021) included both the SN and VTA as one area.Jalles et al. (2020); Tavares et al. (2018) did not specify whether the region they looked at included only the SN or both regions as one; therefore, for simplicity, we assumed a grouped approach.This is because papers using a SN mask often include the VTA but do not specify this and these areas are often grouped particularly because the VTA lacks clear demarcation along the midline (Eapen et al., 2011).In future studies using neuromelanin-MRI, a 3D-sequence may overcome this limitation, as it is possible to obtain high-resolution volume images and 3D-GRE images specifically have been able to properly visualize the VTA (Sasaki et al., 2013).
Of the five studies utilizing neuromelanin-MRI, three used a 2D-GRE sequence (Cassidy et al., 2020;Jarcho et al., 2022;Tavares et al., 2018), with the other two studies (Wang et al., 2021;Jalles et al., 2020) not specifying the method used.The other sequence that neuromelanin-MRI studies may employ is a first spin echo (FSE) sequence, rather than the GRE sequence.The benefit to using a FSE sequence is that it offers better resolution; though GRE sequences offer speed, which is often preferred with clinical patient groups (Chavhan, 2016).The emerging use of other sequences may further shorten acquisition (Kusama et al., 2020), and enable microstructural properties to be studied in conjunction, to understand the contribution of neurodegeneration to the observed signal change (Al Haddad et al., 2023).
Contrary to our expectation of an overall reduction in neuromelanin signal in SUDs (especially cocaine), no significant changes were found.Any higher neuromelanin signals in those with a current SUD may indicate higher dopamine turnover, but over time, this may lead to higher degeneration and lower signals as noted in the other studies.However, based on the heterogeneity of substances that this review includes, no conclusive results on this hypothesis can be made.Longitudinal studies tracking changes in neuromelanin levels over time in individuals with SUDs could provide valuable insights into the progression of addiction and potential targets for intervention.
It is important to note the limitations of our review.Firstly, the number of studies included in the study and the subgroup analysis was limited, which affected the statistical power and may have impacted the generalizability of the findings.Secondly, there was considerable heterogeneity among the included studies in terms of study design, sample characteristics, dopamine measurement techniques, and substances used, which may have contributed to the overall lack of significant findings in the main analysis.In particular, of the six studies included in the meta-analysis (and the four that used neuromelanin-MRI), two of the studies included patients with psychosis, of which the symptoms are associated with neuromelanin-MRI signal (Cassidy et al., 2019); therefore, further studies are especially needed to validate the null findings presented here.It is possible that psychosis and substances have a contrasting impact on neuromelanin levels which could not be elucidated from any studies to date.Lastly, the studies included in our analysis predominantly focused on alcohol and cocaine use disorders, limiting the generalizability of our findings to other addictive substances.
In conclusion, this study contributes to the existing body of literature on the dopamine neurobiology in individuals with SUD by comprehensively synthesizing previous research on neuromelanin levels.Our pre-registered meta-analysis did not find a consistent increase or decrease in SN neuromelanin levels among individuals with SUDs compared to those without.Nevertheless, the absence of significant results may be equally informative, highlighting the gaps in our research approaches towards understanding the relationship between neuromelanin levels and substance use.In particular, we need sufficiently powered neuromelanin-MRI studies, to focus on several commonly used substances, refine measurement approaches and standardizing analytical techniques and reporting methods to enable data-pooling, and explore additional factors that may influence neuromelanin levels in individuals with SUDs.Given the emerging relevance of neuromelanin measures to psychotic and neurodegenerative disorders, future research should focus on expanding the body of literature to related substances of  et al., 2020;Tavares et al., 2018).**Means and standard deviations estimated from Fig. 1 using 90 % intensity in (Wang et al., 2021), with use of Web-PlotDigitizer (Rohatgi, 2024).Abbreviations: SD, standard deviation; SMD, standardized mean difference; CIs, confidence intervals.interest.

Conflicts of interest
LP reports personal fees from Janssen Canada, Otsuka Canada, SPMM Course Limited, UK, Canadian Psychiatric Association; book royalties from Oxford University Press; investigator-initiated educational grants from Sunovion, Janssen Canada, Otsuka Canada outside the submitted work.CMC is inventor on a patent using neuromelaninsensitive MRI, licensed to Terran Biosciences, but has received no royalties.4. Cross-section of the midbrain displaying the substantia nigra, ventral tegmental area, and substantia nigra/ventral tegmental area complex, with segmentation based on Trutti et al., 2019(Trutti et al., 2019).

Fig. 3 .
Fig. 3. Funnel plot depicting publication bias in the meta-analysis.Each point represents a study included in the analysis, with the horizontal axis representing the standardized mean difference and the vertical axis representing the standard error or precision of the effect size estimate.

J
Ahrens reports funding from the Canadian Institutes of Health Research, Schizophrenia Society of Canada Foundation and the Canadian Consortium for Early Intervention in Psychosis, and Quebec Bio-Imaging Network (QBIN).R A Rabin received funding from the Fonds de Recherche du Quebec-Santé.L Palaniyappan's research is supported by the Canada First Research Excellence Fund, awarded to the Healthy Brains, Healthy Lives initiative at McGill University (through New Investigator Supplement to LP) and Monique H. Bourgeois Chair in Developmental Disorders.He receives a salary award from the Fonds de recherche du Quebec-Santé.

Table 2
Methods and neuromelanin findings of included studies.

Table 3
Substantia nigra/ ventral tegmental area values used in meta-analysis and the resulting standardized mean differences.
* Means and standard deviations given from corresponding authors (Jalles