Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

L-methylfolate Shows Promise Against Negative Symptoms in Schizophrenia

23 Mar 2017

by Lesley McCollumFolate Structure

Supplementation with l-methylfolate (l-MF), the active form of folate, may help reduce negative symptoms in schizophrenia, according to a new randomized, double-blind trial published online March 14 in Molecular Psychiatry. The study was spearheaded by Donald Goff, now at the New York University School of Medicine and Nathan Kline Institute, and led by Joshua Roffman of Massachusetts General Hospital and Harvard Medical School in Boston.

In the study of 55 medicated patients with schizophrenia, 29 patients receiving 15 mg of oral l-MF for 12 weeks had high blood levels of l-MF, the primary outcome measure of the study, and also showed improved negative symptom scores compared with 26 patients receiving a placebo. l-MF treatment also generated structural and functional brain changes that may provide clues to the processes behind the supplement’s effects on negative symptoms.

Though preliminary because of the study’s small sample size, the findings add to the growing evidence supporting the idea that raising low folate levels present in schizophrenia may help treat negative symptoms. The ease of its potential implementation as an add-on therapy for schizophrenia makes it an enticing approach—as a "medical food" it’s readily available and doesn’t require specific indications for its use.

Large scale, placebo-controlled, randomized clinical trials are still needed before its implementation, said Robert Buchanan of the University of Maryland, US, who was not involved in the study, in an email interview with SRF. Specifically, he continued, the trials should be designed to allow interpretation of clinical outcomes, which would be a critical issue for negative symptoms.

Previous studies using folic acid have shown promising outcomes for negative symptoms, but variation in genes that regulate the body’s use of folic acid in the schizophrenia population have led to inconsistent responses (see SRF related news story). One gene in particular is MTHFR, which plays a key role in the conversion of folic acid to l-MF (Hill et al., 2011; see SRF related Q&A with Roffman). So in this study, Roffman and colleagues gave l-MF directly to circumvent the genetic variants that may interfere with thebioavailability or efficacy of folic acid.

Folate Foods2

“Unlike the previous studies of folic acid, there was no dependence on MTHFR or other genotype effects on negative symptom improvement,” said Roffman. “So for that reason we feel that this intervention may have a better chance of working in a larger component of the population with schizophrenia.”

The difference in the change in Positive and Negative Syndrome Scale (PANSS) negative score between l-MF and placebo was about three points—not a huge change, noted Roffman—which reflected a larger response in some patients and a more subtle response in others. “I wouldn’t say that this is a cure for negative symptoms, but it does appear to help,” he said.

Patients taking l-MF also showed improvement in PANSS total and general psychopathology scores, but these differences were dependent on specific genotypes. The treatment did not have any effect on positive symptoms or cognition.

“[T]he fact that changes in blood l-MF levels did not correlate with changes in clinical outcome measures is problematic,” said Buchanan, though he agreed with the authors that peripheral blood levels may not reflect levels in the central nervous system, as constraints on the transfer of l-MF across the blood-brain barrier cause levels in the brain to plateau.

A unique feature of this trial was the use of magnetic resonance imaging scans before and after treatment to observe potential brain changes. After treatment with l-MF, patients had increased thickness of the medial prefrontal cortex (mPFC), and were better able to deactivate the region during a working memory task, a process that is impaired in schizophrenia. Patients taking l-MF also showed reduced connectivity between the mPFC and limbic regions, suggesting potential restoration of the hyperconnectivity between these regions that is associated with schizophrenia. However, Buchanan cautions that the lack of a healthy control group also treated with l-MF makes it difficult to conclude if the changes reflect normalization of limbic dysfunction.

Though brain imaging measures did not correlate with symptom changes, a relationship that’s hard to draw for any study of this small size, said Roffman, “… all of these changes were interesting in that they were convergent around medial prefrontal structure and physiology in the context of significant genotype-independent effects on negative symptoms.”



Submitted by Robert W. Buchanan on

I think that the examination of the efficacy of L-methylfolate (LMF) in people with schizophrenia is a logical extension of the previous work of Roffman and colleagues with folate supplementation. However, I am not quite sure of the proposed rationale for the use of LMF (i.e., bypass potentially low-functioning enzymes of the metabolic pathway that regulates the conversion of folate to LMF), since in their previous study with folate and vitamin B12 supplementation (Roffman et al., 2013), the genetic variant that was related to treatment response was the high-functioning variant of FOLH1, which appears to regulate intestinal folate absorption. I also was surprised that the primary outcome measure was not negative symptoms, since there is such a pressing need for new, effective treatments for this symptom component in schizophrenia, and their previous study observed a negative symptom effect in people with the FOLH1 variant.

I assume that the authors chose plasma LMF for their primary outcome measure to demonstrate target engagement, but the fact that changes in plasma LMF levels did not correlate with changes in clinical outcome measures is problematic, though I agree with the authors that peripheral levels may not reflect CNS levels of LMF.

In the previous two studies from this group, folate supplementation led to negative symptom improvement, as measured by the SANS total score. In the current study, they did not observe a significant change in SANS total score in the experimental treatment group. Although there was a change in the PANSS negative symptom scale, there was also a significant and greater reduction in the PANSS general psychopathology score, which raises the question of pseudo-specificity for the change in PANSS negative symptom scale. Such a concern is heightened in that there was no attempt to recruit participants enriched for the presence of negative symptoms or to control potential secondary factors of negative symptom change by limiting the severity of other symptom dimensions. Thus, I would consider their negative symptom results to be hypothesis generating rather than hypothesis supporting. However, in defense of the authors, their primary goal in the current study was to examine the effect of LMF supplementation on plasma LMF levels and not to assess clinical efficacy of LMF supplementation.

Since they did not include a healthy control group that was treated with LMF, I think that it would be difficult to call the observed changes a "normalization" of limbic dysfunction. Also, none of the functional MRI changes were associated with any of the cognitive or symptom measures, which leaves you, as the authors note, with a disassociation between the MRI and clinical results that is difficult to interpret.

If LMF is to be considered for future therapeutic use in people with schizophrenia, then large-scale, placebo-controlled, randomized clinical trials, with a pre-specified primary clinical outcome, will need to be conducted. The trials would need to be designed appropriately to allow interpretation of the study results (i.e., control for pseudo-specificity). This would be a critical issue for negative symptoms and/or cognitive impairments.

Submitted by Joshua Roffman on

Thanks to Dr. Buchanan for his insightful comments, which I will do my best to address here.

1. Regarding the rationale for methylfolate versus folic acid: I appreciate the opportunity to clarify this important point. In our previous studies, there were actually two genetic variants that stratified response to folic acid supplementation: FOLH1 T484C and MTHFR C677T. FOLH1 facilitates intestinal absorption of naturally occurring folate by cleaving its polyglutamate tail. Notably, neither folic acid (i.e., what one would find in a multivitamin) nor methylfolate has a polyglutamate tail―and therefore we would not expect absorption of either of these synthetic folates to be affected by FOLH1 genotype. In that regard, the independence of methylfolate effects on negative symptoms is not surprising―but the dependence of folic acid in our previous study (Roffman et al., 2013) was unexpected. In that study, we speculated that patients who carried the putatively low-functioning FOLH1 484C allele entered the study at a relative disadvantage compared to 484T homozygotes (as the 484C carriers had lower blood folate levels at baseline), potentially explaining their diminished response to treatment after 16 weeks.

On the other hand, for the MTHFR variant, we would anticipate a priori that folic acid response should depend less on C677T genotype than would response to methylfolate. Within the folate pathway, MTHFR is situated downstream of where folic acid enters, but upstream of where methylfolate enters (see Supplemental Figures in the paper for an illustration). This is the primary rationale for why methylfolate would conceivably work in more individuals than would folic acid. Indeed, while we saw in two previous RCTs of folic acid that 677T allele carriers showed more negative symptom improvement than 677C homozygotes, in the recent methylfolate study, negative symptom improvement did not depend on MTHFR genotype. However, even with the present results supporting this notion, there remain two major caveats. First, it would take a head-to-head trial of folic acid versus methylfolate, ideally with stratification based on MTHFR genotype, to establish such an advantage of methylfolate. Second, even though the variants in question are coding, missense polymorphisms that were previously associated with negative symptom severity in previous studies (e.g., Roffman et al., 2013), the N’s are fairly small for pharmacogenomic analyses. For both of these reasons, I wholeheartedly agree with Dr. Buchanan that the findings should be replicated in larger RCTs, with negative symptom improvement as the primary endpoint.

2. Dr. Buchanan is correct that we saw changes in both PANSS negative and PANSS general psychopathology. (Notably, the latter but not the former was dependent on a common genetic variant in MTR that occurs downstream of where methylfolate enters into the folate pathway.) We did explicitly determine whether methylfolate effects on negative symptoms might reflect its effects on depression―which not only shows some phenotypic overlap with negative symptoms but also has been shown to improve after methylfolate treatment (Papakostas et al., 2012). In our recent study, the lack of effect of methylfolate on depression scores (Calgary scale) provides some reassurance about pseudo-specificity, but I agree that this merits continued careful consideration in future RCTs.

3. Regarding “normalization” of limbic dysfunction: While we did not image (or treat) healthy controls in this study, I would refer Dr. Buchanan and other readers to our recent paper on network dysfunction in schizophrenia (Eryilmaz et al., 2016), where we observed increased limbic connectivity in schizophrenia patients compared to matched controls. For our pre-post treatment comparisons in the methylfolate paper, we conducted a highly focal analysis that used exactly the same seed regions that had demonstrated patient-control differences in our previous paper. The fact that methylfolate treatment was associated with decreased connectivity suggests that limbic network function shifted toward a more normal level.

4. Finally, about the primary endpoint for the methylfolate study: It is important to note that our multisite RCT of folic acid (Roffman et al., 2013) had not yet been analyzed at the time that the methylfolate study was designed. Had we known at that time that folic acid would show MTHFR-dependent effects on negative symptom improvement, this likely would have motivated a clinical rather than physiologic endpoint in the methylfolate trial. That being said―when considering in aggregate 1) the data from (now) three RCTs of folic acid or methylfolate showing an effect on negative symptoms and two additional RCTs showing improvement on general symptom ratings, 2) replicated cross-sectional data associating deficient folate metabolism specifically with negative symptom impairment, and 3) the MRI data from our recent study demonstrating a convergence of structural and functional effects of methylfolate on the medial prefrontal cortex, I would argue that a fairly convincing case can be made at this point for the relationship between folate metabolism and negative symptoms. When also taking into account the safety profile of folic acid/methylfolate and the treatment-refractory nature and tremendous burden of negative symptoms, the possibility of a uniquely favorable risk-to-benefit ratio should motivate additional study.

Submitted by Joseph Firth on

This comment was co-written by Alison Yung.

The pioneering study by Joshua Roffman and colleagues (2017) represents a positive step forward toward demonstrating and understanding the efficacy of folate supplementation in schizophrenia. Our recent meta-analysis in Psychological Medicine (Firth et al., 2017) showed a significant effect of B vitamins (including folate) on overall symptomology in schizophrenia. However, we were unable to establish from where these effects arose due to insufficient reporting on specific symptom domains across the included studies. Nonetheless, the new study by Roffman et al. (2017) finds that daily supplementation with 15 mg of methylated folic acid has specific effects on negative (but not positive) symptoms of schizophrenia.

Furthermore, Roffman and colleagues built upon the findings of their previous work in this area by demonstrating that methylated forms of folic acid (i.e., the 15 mg of methylfolate used in this study) can benefit a broader range of patients than typical folic acid supplements. This may be because certain genetic variants which interfere with folate absorption are prevalent among people with schizophrenia and reduce the bioavailability of standard folic acid supplements (Roffman et al., 2013).

The benefits of methylfolate for schizophrenia were first found over 27 years ago in a small but important study conducted by Godfrey and colleagues in London, UK, and published in The Lancet back in 1990 (Godfrey et al., 1990). However, this very early study only examined improvements in total symptom scores―and their experiment has not been replicated since, until now.

Roffman successfully replicates the result, rules out the possibility of the effect being driven by antidepressant effects of folate, and identifies methylfolate as a promising new medication for tackling one of the most damaging and difficult-to-treat aspects of schizophrenia: negative symptoms. Despite no change in neurocognitive task performance, significant increases in cortical thickness were also observed from methylfolate treatment. This indicates that folate may also be capable of inducing neuroplasticity, and perhaps having long-term effects on cognitive dysfunction in schizophrenia―another highly detrimental but often untreated aspect of the disorder.

Overall, the robust methodology and promising findings from this single-site study now call for wide-scale replication, particularly given the tolerability of this water soluble B vitamin. Continuing down the path laid out by Roffman and colleagues here will certainly increase our understanding of the role of folate in schizophrenia, and may even provide a safe and novel adjunctive treatment for improving the typically poor recovery rates in people with this condition.

Submitted by Maureen Martin on

I wonder if response to methylfolate is associated with the presence of folate receptor autoantibodies in schizophrenia (Ramaekers et al., 2014), and if methylfolate response and folate receptor autoantibodies are markers of immune dysregulation similar to what is seen in PANDAS. If so, we could easily speculate on additional treatments that might be beneficial to these patients based on what is shown to help in PANDAS.

Submitted by Joshua Roffman on

One additional update: While change in negative symptoms did not correlate with change in mPFC thickness in the l-methylfolate study (n = 35), some new results from Walton, Ehrlich, and colleagues provide important additional context. In a paper published online in Psychological Medicine on May 26 (Walton et al., 2017), data from the ENIGMA consortium (n = 1,985) demonstrate a significant inverse relationship between negative symptom severity and cortical thickness in the very same medial prefrontal region where l-methylfolate treatment associated with cortical thickening. This additional evidence strengthens the case that LMF-induced changes in brain structure are meaningful for symptom improvement―a question that we will hopefully be able to address in larger, multisite studies.