MAJOR DEPRESSION is a serious illness, but also an elusive one. It wrecks lives and may drive people to suicide. It sometimes, though not always, alternates with periods of mania in a condition called bipolar disorder. And it is disturbingly common. Reliable figures are hard to come by, but in some parts of the world as many as one person in four experiences major depression at some point during their life.

Depression’s diagnosis has, though, a worryingly arbitrary quality to it, depending as it does on a doctor’s assessment of a patient’s mood against a checklist of symptoms which may be present in different combinations and are often, in any case, subjective. This has led to a search for reliable biochemical markers of the illness. Not only might these assist diagnosis, they may also improve assessments of prognosis and point towards the most effective treatment in a particular case. Now, a group of neuroscientists at Indiana University, in Indianapolis, led by Alexander Niculescu, think they have found a set of markers that can do all this.

As they write in Molecular Psychiatry, Dr Niculescu and his colleagues have been working with data and blood samples collected over the course of 15 years from hundreds of patients at the Indianapolis Veterans Administration Medical Centre. The target of their investigation was small pieces of RNA, a molecule transcribed from the DNA of genes as part of the process by which the information encoded in those genes is used by cells to make proteins.

Tracking levels in the blood of relevant RNA molecules shows the activity of the underlying genes. That let the researchers identify, in an initial sample of 44 patients’ records, which genes were becoming more and less active as people’s mood disorders waxed and waned. Initially, they found thousands, but they first narrowed these down to the ones that seemed to show the best prediction of mood and then, by turning to the corpus of published research on genes associated with depression, narrowed the selection still further to 26 that had previously been suspected of involvement in the illness. They then followed this clutch up in eight groups of patients, ranging in size from 97-226, to see which best predicted the course and details of a patient’s illness.

Thirteen markers survived this final winnowing. The genes they represent are involved in a range of activities, including running circadian rhythms (the endogenous clocks which keep bodily activities synchronised with each other and with the daily cycle of light and darkness); regulating levels in the brain of a messenger molecule called serotonin, the activity of which is well known to get out of kilter in depression; responding to stress; metabolising glucose to release energy; and signalling within cells.

Together, these 13 RNA markers form the basis of a blood test that can not only diagnose depression, but also predict who will go on to develop bipolar disorder, who is likely to need hospital treatment in the future, and which drugs will be most effective in particular cases. Six of the RNAs were good predictors of depression alone. Another six predicted both depression and mania. One predicted mania alone.

On top of their potential role in diagnosis, three of the genes identified are known from previous work to be affected by lithium carbonate, an established treatment for bipolar disorder, and two other are affected by a class of antidepressant drugs called selective serotonin reuptake inhibitors, of which Prozac is probably the best-known example. It is for this reason that Dr Niculescu thinks his blood test may help to pick appropriate treatments.

The results even indicate some non-psychiatric drugs that might be worth trying, since the analysis showed they had characteristics which could affect some of the biomarkers. A beta blocker called Pindolol, for example, is currently used to treat high blood pressure. But this drug is also known to affect serotonin activity, and Dr Niculescu and his colleagues found, from a search of the published literature, that it has been seen to affect levels of all six of the depression-only biomarkers. That, he thinks, might make it a good candidate for treating depression.

Dr Niculescu and his colleague and co-author Anantha Shekhar have founded a company called MindX Sciences and are seeking regulatory approval for the test’s medical use. If all goes well, future versions might incorporate additional biomarkers, and might use saliva, rather than blood as the fluid sampled. Biomarker-diagnosis of depression is unlikely to replace assessment by checklist, which would, presumably, be used to see who should be sent for screening in the first place. But as a way of confirming and refining diagnoses, and also of suggesting treatment in what is both an uncertain and a sensitive area, it seems an important advance. ■