Answer quickly: are your legs strong because you run often, or do you run often because your legs are strong? Do you like maths because you are good with numbers, or are you good with numbers because you like maths? In people with post-traumatic stress disorder (PTSD), is the hippocampus small because of the disease, or is PTSD present because the hippocampus is small?

The difference between cause and effect is a constant pain in the neck of neuroscientists. Demonstrations over the last few years that the brain can change significantly over time because of what it experiences have only complicated matters. Braille readers have enlarged cortical representations of their fingertips; London taxi drivers, forced to know by heart their way around town, have an enlarged anterior hippocampus; women with chronic depression have a smaller hippocampus. But which comes first, the altered size of a brain area or the condition?

The usual way out is to look for a relationship between changes in the brain and the duration of practice or exposure. For example, the longer the time as a London cabbie, the larger the anterior part of the hippocampus; the longer the duration of chronic depression, the smaller the female hippocampus. Put these facts together with the discovery of continuous production of new neurons in the human hippocampus, and it seems perfectly plausible that this brain structure undergoes remodelling as it sees fit.

And this seemed to be the case with PTSD. In the 1990s, PTSD triggered by intense trauma and stress — two factors that cause neuronal atrophy and even death — was found to correlate with reduced hippocampal size in veterans of the Vietnam War. The severity of symptoms even correlated with the degree of hippocampal shrinkage relative to controls, which seemed to fit perfectly with the hypothesis of neuronal atrophy caused by stress.

Enter Mark Gilbertson and his data from 40 pairs of identical twins, published in the November issue of Nature Neuroscience. According to magnetic resonance imaging scans, the worse the symptoms of the disorder in combat veterans, the smaller the volume of the hippocampus in the twins who stayed at home.

Reduced hippocampal volume in the healthy twins correlated, of course, with a similar reduction in the veteran brother. The latter, however, was not related to combat severity — although those that developed PTSD had taken part in more severe combat than veterans whom had not. The most parsimonious explanation is that, in the case of PTSD, reduced hippocampal size results neither from trauma nor from the disorder, but is instead a pre-existing condition that predisposes subjects to the development of PTSD once the brain is exposed to trauma. In this case, the reduced size of the hippocampus is probably conducive to an exaggerated hormonal and behavioural response to stress. This scenario would also explain why only a few people develop the disorder, despite the fact that many are exposed to the same trauma.

In the end, the relationship between stress and hippocampal size is probably a two-way street. In other words, the egg (stress) often comes first, but sometimes the chicken (a reduced hippocampus) has precedence. This should be taken as a sign that it is time to drop rhetorical conundrums about chickens and eggs in the face of evidence that nature and environment, brain and society, can often not be dissociated. After all, chicken and egg are one and the same animal.