On the Trail of White Matter

On a warm day in 1987, I had a eureka insight. Intrigued by the idea of disconnection from my recent behavioral neurology fellowship in Boston (1), and then encountering the ravages of toluene leukoencephalopathy in Denver (2), it flashed within my mind that a dementia syndrome related to diffuse white matter involvement could be as legitimate as focal neurobehavioral syndromes from localized disease. This idea—that I decided to call white matter dementia—fit nicely with enthusiasm for the cortical-subcortical dementia dichotomy (3) and further refined this topic by focusing exclusively on the subcortical white matter. At first, having been counseled that few ideas are truly new, I imagined that I could hardly be so fortunate as to have stumbled onto something novel, but as I cautiously developed the idea and searched the literature, it seemed that indeed no one had thought about white matter in quite this way. So, with trusted colleagues, I introduced the concept to stimulate interest in an area I thought important (4). More generally, I was keenly interested in advancing investigation of how myelinated tracts in the healthy brain contribute to our humanity. In my mind was a conversation I recalled from medical school with the venerable neurophysiologist Vernon Mountcastle, the “father of neuroscience” (5); when I asked where he thought advances in the understanding of behavior would next occur, he replied that they would likely come from the study of neuroanatomy.

Criticism came from some who contended that a truly isolated white matter disorder is neuropathologically rare if not inconceivable. Others argued that the term dementia was overly dramatic in its implications, as only some—perhaps few—people with diffuse white matter disease can be said to have dementia as a result of that pathology. Most of all, however, it was an abiding skepticism of the notion that white matter merited neurobehavioral study that proved stubbornly resistant (6).

Over the years that followed, my work led me to consider the entire range of white matter disorders, even in the context of the degenerative dementias so widely regarded as cortical in origin. As I pursued these studies, it became clear that neurobehavioral features first observed in toxic leukoencephalopathy (2) could be detected in all disorders, notwithstanding the fact that in many cases white and gray matter pathologies are indeed commingled (7). As for the criticism of dementia being too strong a term, the idea of mild cognitive dysfunction (7) proved useful to highlight the point that white matter disease can often produce mild clinical manifestations that, when detected at an early stage, can be addressed while recovery is more feasible than when dementia appears.

What has been called corticocentric myopia (6) has proven far more daunting, as there persists among many neuroscientists and neurologists an almost irresistible urge to associate cognitive function with gray matter, particularly of the cerebral cortex. Many a conversation with trusted colleagues has started with my questioning the assumption that cognitive impairment must mean cortical dysfunction. Yet the selective evolutionary expansion of white compared with gray matter, combined with remarkable neuroimaging advances that enable in vivo depiction of affected tracts, supports the growing recognition that white matter really does matter for behavioral neurology (7). Currently, images of white matter connectivity worthy of art museum display are now commonplace, elegant and detailed atlases of white matter anatomy are available, and the publicly funded Human Connectome Project is characterizing the entirety of white matter in the brain. All of this progress improves understanding of the tissue in which pathology often begins so that clinicians can better describe the resultant neurobehavioral deficits and researchers can develop more effective treatments (7).

Does the idea of white matter dementia inform work on brain-behavior relationships? It is doubtless true that no diagnostic code exists for this entity, and grant funding is scarce if one looks across disease categories to seek unifying neurobehavioral themes based on neuroanatomy. Yet, the 135,000 km of myelinated fibers in each human brain—enough to encircle the earth three times—cannot be ignored in thinking about how the brain works and breaks down. Many of humanity’s most disabling and perplexing medical problems, including Alzheimer’s disease (8) and traumatic brain injury (9), may well be related to white matter pathology that is only beginning to be appreciated. White matter dementia is most fundamentally a call to consider a leukocentric approach to diseases that have defied complete understanding when viewed through the lens of corticocentrism.

More broadly, the study of white matter and behavior serves to balance the familiar reductionism of science and medicine with the scholarship of integration (10). Traditional disciplinary boundaries are transcended by appraising any information relevant to normal and disrupted connectivity within distributed neural networks from which cognition and emotion emerge. New vistas of brain-behavior relationships can be revealed with many potential implications for disease prevention and treatment. While such an approach may seem less relevant than laboratory investigation to many in neurobiology, a clinical perspective reminds us that the human brain must be studied in its aggregate as well as in its details (7–9). As crucial as the analysis of cellular and molecular phenomena has proven to be, the understanding of human behavior and its many afflictions demands equal attention to the brain as a whole.

Reflecting on the trail I came upon, it is one that, despite many obstacles, I am most privileged to have traveled. The half of the brain where myelinated axons reside continues to offer a rich source of endless fascination. Neuroscience of course begins with the knowledge that the neuron is the fundamental unit of the nervous system and that human behavior arises from the concerted activity of billions of neurons working together. But the components of neurons are not trivial, and how these various segments break down is critical to better understanding of etiopathogenesis and treatment (7–9). My focus on white matter has never been intended to deny the importance of its gray counterpart, but rather to redress an imbalance in long-standing assumptions of how the brain mediates behavior. Much remains to be done, and I earnestly hope that the road ahead will beckon investigators to travel farther ahead.