MDD is a mental health condition with widely debated pathophysiology, that is still poorly understood and under-studied in the context of female individuals [7]. A defining factor that influences the development of MDD is a dysregulation of the ECS and its endogenous ligands, eCBs and NAEs [10]. The evidence identified through this systematic review suggest altered levels of eCBs and NAEs are present in depressed females. However, no cohesive pattern was identified, with studies observing both reduced and elevated levels of eCBs and NAEs. The variability in study designs from screening tools used for depression, eligibility criteria and analysis methods of eCBs unfortunately lacked the consistency needed to attain a reliable and valid understanding of eCB levels in females with depression.

The included studies used a wide range of assessment tools to confirm and quantify depression, including both self-report screening measures and structured clinical interviews. Difficulties arise when comparing results across studies that use different depression screening measures, as each have a unique set of questions and cut-off bands to quantify depression. Concerns have also been raised about the difference in cut off bands used by self-report questionnaires to measure severity of depressive symptoms. When comparing use of HADS and PHQ-9 in a clinical setting, 74% of individuals assessed by PHQ-9 would be offered an anti-depressant, while only 37% would be written a prescription when assessed by HADS [33]. Self-reported measures were also used to sort participants into groups, with levels of eCBs being compared amongst groups with minor or “less severe” depression [27, 28]. Mild depression is only a less severe variant of major depression, and comparing groups with the same illness may be limit the ability to accurately determine the difference in eCB levels between groups. In contrast, structured clinical interviews using the DSM-IV criteria is currently considered the gold standard for assessing depression [34] and is the suggested tool that researchers should use to assess their study populations [35]. However, it should also be recognised the influence that sex and gender plays in MDD, with concern’s raised about the DSM-V classification of MDD, particularly as the criteria does not distinguish between men and women. Experience of MDD between sexes has been established, with females more likely to report anxiety, somatic symptoms and guilt compared to males. Lack of inclusion from female-sex perspectives and social determinants in the DSM-V classification of MDD, raises the question if diagnostic criteria should be further evaluated to incorporate sex-based differences to improve identification and treatment of MDD. Future studies with the objective to compare levels between healthy populations and those with a mental health conditions should also aim to ensure appropriate screening of participants with standardised diagnostic procedures and allocate a true healthy population as the comparator.

Enrolment of participants with other co-morbid diseases, namely FM and BMS presented another methodological confounder. Literature has found altered levels of eCB in both of these conditions; eCB deficiency has been proposed as a cause of FM due to its role in pain processing and inflammation [36, 37]. Individuals with BMS have altered TRPV1 and CB1 receptors expression on tongue epithelia, suggesting that the ECS plays a role in the pain and severity of symptoms in BMS [38]. Enrolling participants with diagnoses that have underlying pathophysiology linked to ECS dysfunction hinders the ability to determine if variations in eCB levels are due to depression or to another disease. Furthermore, Ho et al. [30] enrolled participants with concurrent depression and heart disease, finding that AEA and 2-AG levels in females with depression was positively correlated with diastolic and mean arterial pressure. The authors suggested that the biochemical changes in depression, such as increased blood pressure, lends itself towards eCBs playing a key faciliatory role in both diseases to reduce the extent of hypertension and it’s associated inflammation [30]. Again, this study was not able to determine if the altered eCB levels was a result of heart disease or presence of depression. Thus, it is essential that future research aimed at assessing the role of the eCB in depressive disorders excludes participants with concurrent diseases in which the ECS is already implicated. To gain further understanding of the ECS in other conditions, a review may be warranted to identify these to understand the mechanisms of the ECS and the potential variations in eCB levels.

Methods of measuring eCBs and storing samples also varied between studies. LC-MS is most commonly used to determine levels of eCBs in blood samples due to its high sensitivity, high validity and short sample preparation time [39]. While variations of LC-MS, including using atmospheric pressure chemical ionization or electrospray ionisation, are routinely used for measuring compounds found in bodily fluids, there is no known research comparing the efficacy and validity of these testing methods in context of eCBs. It is essential that appropriate storage of collected blood samples are used for eCBs measurements [39]. For clinical studies, it is highly recommended that anti-coagulated blood is immediately placed in an ice bath, and then centrifuged before being frozen and stored at -80^oC until analysis [40]. eCBs are highly sensitive to temperature variation; it has been found that whole blood storage at 4^oC can increase AEA concentrations in the sample [40]. Only one study in this review reported storing blood samples at -80^oC [30], with four storing samples at -70^oC [28, 29, 31, 32] and one not disclosing any blood sample handling or storage procedures [26]. When collating results that assess eCB levels in blood samples, researchers should critically analyse blood sample collection and handling techniques in order to draw appropriate conclusions and limit confounding factors from data provided. It is also essential that future studies that assess eCB levels use standardised procedures of blood sampling and to ensure upmost accuracy of eCB measurement and data reporting [40]. There is less data on the validity of measurement of eCBs in hair samples, however hair samples have been suggested to be a highly sensitive indicator of retrospective quantification of eCB levels in those with depression, further research is still needed to fully attain their reliability [27]. Refer to Fig. 2 for an outlines of suggestions to improve methodology in future studies.

The ECS has a documented role in regulating homeostasis of physiological, cognitive and emotional processes in the body, suggesting a possible link between ECS dysfunction and the development of mood disorders in females [41]. Proposed aetiologies of MDD such as genetic influences and inflammation, are both intrinsically linked with the ECS. Firstly, a genetic study of predominantly female participants concluded that CB1 receptor haplotypes increased the individual’s risk of depression after experiencing adverse life events, which may explain the higher ratio of females who are diagnosed with depression [42]. Secondly, increased levels of inflammatory cytokines IL-8, IFN-y, TNF-α and CRP have been positively correlated to depressive symptoms in females [43, 44]. Interestingly, 2-AG and AEA have shown anti-inflammatory activity by down regulating inflammatory cytokines such as TNF-α, suggesting that the ECS may play a role in modulating neuroinflammation associated with depression [18]. Finally, preclinical reports have suggested that modulation of the ECS displays similar actions to pharmaceutical anti-depressants by enhancing serotonergic and noradrenergic transmission resulting in anti-depressant effects [23].

Although beyond the scope of this review, the included studies did highlight the potential influence of stress and anxiety on eCB levels, both conditions which are closely interlinked with the severity of depression in females [28]. It is important not to overlook that mental health disorders often accompany each other and it is essential to understand the relationships between depression and anxiety/stress in order to provide appropriate therapeutic treatment. Approximately 41.6% of individuals with MDD have been found to have a concurrent anxiety disorder over the same 12-month period [45]. The presence of significant anxiety with MDD has often been found to be a predictor of more severe depression and increased suicidal ideation [45]. The involvement of AEA in anxiety has been shown in both human and animal studies in which a significant reduction in anxiety was correlated with increased AEA in female-sexed individuals with clinically relevant anxiety. Interestingly, animal studies have shown that AEA metabolism elicits anxiolytic effects, perhaps providing a novel target for anxiolytic in human individuals [28]. Further exploration into the relationship between anxiety & depression in females and the ECS is warranted, which could provide insight into screening and treatment for females with comorbid anxiety and depression. Although out of scope for this review, future research examining the function of the ECS in females with other psychiatric disorders like bipolar or schizophrenia would be of interest.