Over the past decade, about half a million veterans have received diagnoses of post-traumatic stress disorder or traumatic brain injury. Thousands have received both. Yet underlying the growing numbers lies a disconcerting question: How many of those diagnoses are definitive? And how many more have been missed?

No one can say. Although PTSD is hardly new, diagnoses still largely rely on self-reported symptoms. And while severe brain injuries are often clearly diagnosable, finding evidence of mild TBIs, particularly older ones, can be all but impossible. It means that for a soldier who, five years after duty in Iraq, still feels "not right," with symptoms from headaches to sleeping problems to irritability, doctors can only guess at the cause.

Now, in one of the largest studies of its kind, researchers based out of New York University's medical school have begun a five-year study to find biological signals, known as biomarkers, that could provide objective evidence of those so-called invisible injuries of war.

'Beyond subjective reports'

"We want to elevate mental health to standard physical health," said Dr. Charles Marmar, chairman of the psychiatry department at NYU Langone Medical Center and the project's lead investigator. "You don't go from having shortness of breath to having cardiac surgery; you have a series of objective lab tests first. We would like to do the same thing with PTSD and TBI. That is, go beyond subjective reports."

Marmar faces skepticism that biological signals can be found for the disorders like PTSD or depression, which his group is also studying, or even a less severe brain injury.

His project is significant both because of its size -- researchers hope to recruit 1,500 subjects -- but also because much of its financing is already guaranteed through a $17 million grant from the Steven A. and Alexandra M. Cohen Foundation, founded by the billionaire hedge-fund manager. Marmar hopes to match that with federal grants.

Biomarkers are physiological road signs that can tell doctors whether a person has a disease or injury or is likely to contract a particular ailment. Tissue damaged by a heart attack releases chemicals into the blood that can be detected. Abnormal levels of the proteins amyloid and tau, as well as shrinkage of certain areas of the brain, are considered markers of Alzheimer's disease.

The lack of reliable markers for PTSD and mild TBI has had significant consequences, experts say. Without clear-cut tests to spot them early, the disorders can go undetected until symptoms become disabling. Misdiagnoses readily occur, leading to ineffective or even damaging treatments. And weeding out fraudulent claims can be difficult and contentious.

Five groups to be studied

The study plans to look at five groups, each consisting of 300 veterans: people with PTSD; with TBI; with both, and with depression but without PTSD or TBI. The fifth group will be of veterans with combat deployments who show no symptoms of any of those disorders.

One team will use magnetic resonance imaging to compare the brain structures of healthy people with those of people with PTSD. Another will use brain imaging to test a theory that abnormalities in the thalamus are an indication of head trauma, said Dr. Robert Grossman, dean and chief executive of the Langone Medical Center and a senior adviser on the study. And other researchers will look for biomarkers in genes, blood and hormones. One group will even analyze audio recordings of speech to see whether evidence of PTSD can be found in the pitch, timber and tone of voices.