Dr Reynolds cogently argues that a trial of methylfolate treatment should be used before using a label of treatment-resistant depression. He reminds us that he has previously identified a subgroup of patients with depression with a biochemical profile suggestive of a dysregulated one-carbon metabolism and that therefore metabolic profiling may support treatment decisions.

One-carbon metabolism refers to a complex network of biochemical reactions, including the folate cycle, which makes methyl groups and other one-carbon moieties available for cellular processes including synthesis of proteins, genomic maintenance and epigenetic methylation.^{Reference Clare, Brassington, Kwong and Sinclair1} Folate (vitamin B9) cannot be synthesised by animals and is derived entirely from the diet, thus reduced dietary intake and impaired absorption (as can occur in coeliac disease) contribute to folate deficiency. Folate undergoes reduction to the biologically active tetrahydrofolate before entering the folate cycle.^{Reference Ducker and Rabinowitz2}

Methylenetetrahydrofolate reductase (MHTFR) is required in the reduction of folate. Functional polymorphisms in the MHTFR gene have been associated with major depressive disorder^{Reference Lewis, Lawlor, Davey Smith, Araya, Timpson and Day3–Reference Wu, Ding, Sun, Yang, Chen and Zhao5} and resistance to selective serotonin reuptake inhibitor (SSRI) treatment.^{Reference Cho, Amin, An, Rambaran, Johnson and Alzghari6} (10). Similarly, folate and tetrahydrofolate deficiency has been inconsistently^{Reference Frankenburg7} implicated in several physical, neurological and psychiatric disorders including depression^{Reference Clare, Brassington, Kwong and Sinclair1,Reference Reynolds8} and is associated with low serotonin, dopamine and norepinephrine metabolites in the cerebrospinal fluid.^{Reference Bottiglieri, Laundy, Crellin, Toone, Carney and Reynolds9}

Several clinical studies have been published regarding monotherapy or supplementation of antidepressants with different folate formulations. Two recent meta-analyses^{Reference Scheft, Kilarski, Bschor and Kohler10,Reference Sarris, Murphy, Mischoulon, Papakostas, Fava and Berk11} incorporating a range of folate formulations did not support their use as monotherapy or augmenting agents in major depressive disorder. Restricting the analysis to just l-methylfolate was more positive.^{Reference Scheft, Kilarski, Bschor and Kohler10} This formulation is interesting because it does not require the action of MHTFR to enter the folate cycle. This analysis included only a two-phase trial by Papakostas et al,^{Reference Papakostas, Shelton, Zajecka, Etemad, Rickels and Clain12} which showed a statistically significant improvement in SSRI antidepressant effects with l-methylfolate at a dose of 15 mg and an underpowered trial by Reynolds et al ^{Reference Reynold, Crellin, Bottiglieri, Laundy, Toone and Carney13} that compared amitriptyline with l-methylfolate as monotherapy.

To date, folate-related or other neuromodulatory agents have not been examined in randomised controlled trials meeting the most commonly accepted minimal clinical criteria for treatment-resistant depression (non- response to >2 treatments in the current episode) and therefore were not included in the augmentation meta-analysis conducted by Strawbridge et al. ^{Reference Strawbridge, Carter, Marwood, Bandelow, Tsapekos and Nikolova14} Nevertheless, the authors are conducting a new meta-analysis using the same methods for major depression non-responsive to >1 antidepressant in the current episode (protocol registered via Prospero),^{Reference Scott, Strawbridge, Marwood, Gnanapragasam, Bandelow, Cleare and Young15} in which it appears likely that methylfolate, at least, may be included.

It seems therefore, that although there are some studies and a plausible rationale suggesting that folate could be clinically useful in a subgroup of patients with major depressive disorder and/or treatment-resistant depression, further work needs to be done in this area before this can become a standard recommended treatment option.