Disease of Development?
In "White Matter Matters," R. Douglas Fields reports that myelination problems are implicated in schizophrenia. Certain studies have found that the long-term outcome for schizophrenia patients is better in developing countries than in developed ones. Because myelin formation continues into one's 20s and is affected by experience, is it possible that an enriched living situation could help such patients recover some impaired myelin? Or that psychiatric drugs (less available in developing countries) blunt neuronal activity that could assist further myelination?

Phil Thompson
Los Altos, Calif.

FIELDS REPLIES: The parallel between schizophrenia recovery rates and reliance on medication in developing and developed countries is interesting and may offer practical insights. But schizophrenia is a complex disorder that has both genetic and environmental components, making it difficult to conclude that either one's drug therapy or living situation provides a superior approach to treatment.

Both approaches may in part affect schizophrenia via effects on myelin formation. In addition to causing myelin-forming cells to respond to altered functional activity in the brain, antipsychotics may have direct effects on oligodendrocytes. Until recently, it was not appreciated that oligodendrocytes have many of the same neurotransmitter receptors and transporters that are in neurons, including those that are implicated in schizophrenia. Antipsychotics such as haloperidol can affect the survival of oligodendrocytes.

In animal studies and postmortem analyses of schizophrenia patients, loss of oligodendrocytes is typically seen, but some studies report that dopamine protects oligodendrocytes from injury and that antipsychotic drugs boost synthesis of lipids needed to form myelin.

Most likely myelin is involved in many aspects of schizophrenia. In ignoring the role of myelinating glia in mental illness until now, researchers were playing with only half a deck. Novel treatments may come not only from recognizing treatments' ancillary effects on myelin formation but also from using the rest of that deck. Drugs might be devised to target myelinating glia specifically, offering new possibilities for this tragic disease.

Market Correction
In "When Markets Beat the Polls," Gary Stix details studies that suggest markets using securities for candidates "predict" the results of elections better than polls do. But polls do not make predictions; they simply report the state of opinion at a specific time. The assertion that a market was more accurate than a poll 100 days before an election is only meaningful if one assumes that the distribution of opinion was frozen over those 100 days. What method can provide the "actual" opinion 100 days in advance to serve as a comparative measure of accuracy?

Furthermore, all polls have a sampling error. An election-day poll reporting that a candidate has 52 percent of the vote is accurate if that candidate receives anywhere from 49 to 55 percent. Markets do not have to contend with a sampling error, so they do provide more precise results, but that does not make them more accurate. A market result may be closer to what a candidate received, but if that candidate's share is within a poll's band of sampling error, the poll is just as accurate.

Jorgen Rasmussen
Distinguished Professor Emeritus
Iowa State University

Growing Up Fast
"The End of Cosmology?" by Lawrence M. Krauss and Robert J. Scherrer, states that the quickening expansion of the universe will eventually pull galaxies apart faster than light, causing them to drop out of view. It was my understanding, however, that according to relativity theory, objects cannot move faster than the speed of light relative to an observer.

Jonathan Nichols
San Francisco

According to Krauss and Scherrer, a direct implication of the accelerating expansion of the universe they describe is that three quarters of the energy in the universe is "dark"—in other words, a mystery. I have to question whether the authors have extrapolated the consequences of accelerating expansion beyond the point where it is reasonable to be so authoritative.

Steve Didcott
Great Missenden, England

KRAUSS REPLIES: Regarding Nichols's letter, in general relativity superluminal travel is possible, but one has to be careful about what one means. Special relativity implies that nothing can travel through space faster than light speed. General relativity, however, allows for the possibility that objects that are locally at rest in their background inertial frame can be separating from distant objects at greater than light speed because the space between these objects is expanding faster than light. There is no limit on the expansion rate of space. Even in a universe of non-accelerating expansion (that is big enough), there will always be objects receding from one another at greater than light speed. A Scientific American article by Charles H. Lineweaver and Tamara M. Davis addresses the subject at www.SciAm.com/jul2008.

As to Didcott's letter, it is true that one must be wary of extrapolations into new domains in physics. I should point out, however, that it is not simply the acceleration of the universe that implies dark energy. All the other observations in cosmology, particularly the estimates of big bang nucleosynthesis, the measurements of galaxy clustering, and the inference that the universe is flat (confirmed by observations of the cosmic microwave background), appear to be consistent only if we assume dark energy's existence. Because we do not know the nature of dark energy, we cannot say with any real certainty that the rest of the universe will disappear in the far future. But the possibility is fascinating, not least because it points out how dependent we are on our observations in framing conceptions of the cosmic past and future.