On 2014 Nov 22, IRWIN FEINBERG commented:

Giedd et al suggest that the changes they measured longitudinally in MRI-estimated cortical volume (and in cortical thickness in previous reports) might index synaptic proliferation and elimination. This seems implausible. As the authors of this study likely recognize, changes in dendritic spines are well-below the limits of current MRI resolution. In addition, the trajectories and magnitudes of the MRI changes deviate from those of the best available data for synaptic density (Huttenlocher 1979). Synaptic density is falling steeply while MRI cortical measures are changing only slightly across adolescence. Although Huttenlocher’s synaptic density data do not contain measurements during adolescence, they do show a 40 to 50% drop off from late childhood to early adulthood, much larger than the 10-15% decline in gray matter volume reported by Giedd et al. There are other brain measures (cortical metabolic rate and NREM delta amplitude) with cross-sectionally measured trajectories that visually and statistically more closely parallel the synaptic density data and might therefore better index this developmentally critical brain variable (Feinberg, Thode et al. 1990). The NREM delta trajectories have recently been confirmed in a longitudinal study which also identified ages 12-16.5 yrs as a period of accelerated brain maturation (Campbell and Feinberg 2009). In contrast to the sleep EEG, MRI measures do not reveal any period of accelerated change, nor would one expect this in post-natal measures of brain structure.

While I do not believe the MRI studies index synaptic elimination, I strongly agree with Giedd et al that it is critical to establish the cellular basis of the developmental MRI changes. (Our laboratory has also emphasized this point (Campbell and Feinberg 2009)). However, it is unlikely that animal studies of the sort proposed by Giedd et al could provide conclusive answers because of species differences. For example, cortical synaptic development in monkeys is isochronic rather than heterochronic as in humans ((Rakic, Bourgeois et al. 1986)). What is urgently needed is a large scale multisite replication of Huttenlocher’s heroic, post-mortem study of human brains. New investigations could include biochemical as well as ultramicroscopic measures of dendritic components and morphology. Such data could also fill in the age gaps in Huttenlocher’s dataset, providing more detailed maturational trajectories for synaptic connectivity. In addition, when ante-mortem MRI studies are available, they could be related to the post-mortem findings, so that the anatomic basis of developmental MRI changes could be determined. Replications lack research glamor. In this instance, however, a replication (and major extension) is urgently needed to understand existing findings and to provide a firm foundation for interpreting MRI and other studies of post-natal brain development.

Campbell, I. G. and I. Feinberg (2009). "Longitudinal trajectories of non-rapid eye movement delta and theta EEG as indicators of adolescent brain maturation." Proceedings of the National Academy of Sciences of the United States of America 106(13): 5177-5180, PMID 19307577.

Feinberg, I., et al. (1990). "Gamma distribution model describes maturational curves for delta wave amplitude, cortical metabolic rate and synaptic density." Journal of Theoretical Biology 142(2): 149-161, PMID 2161971.

Huttenlocher, P. R. (1979). "Synaptic density in human frontal cortex - developmental changes and effects of aging." Brain Research 163(2): 195-205, PMID 427544.

Rakic, P., et al. (1986). "Concurrent overproduction of synapses in diverse regions of the primate cerebral cortex." Science 232(4747): 232-235, 3952506.

On 2014 Nov 22, IRWIN FEINBERG commented:

Giedd et al suggest that the changes they measured longitudinally in MRI-estimated cortical volume (and in cortical thickness in previous reports) might index synaptic proliferation and elimination. This seems implausible. As the authors of this study likely recognize, changes in dendritic spines are well-below the limits of current MRI resolution. In addition, the trajectories and magnitudes of the MRI changes deviate from those of the best available data for synaptic density (Huttenlocher 1979). Synaptic density is falling steeply while MRI cortical measures are changing only slightly across adolescence. Although Huttenlocher’s synaptic density data do not contain measurements during adolescence, they do show a 40 to 50% drop off from late childhood to early adulthood, much larger than the 10-15% decline in gray matter volume reported by Giedd et al. There are other brain measures (cortical metabolic rate and NREM delta amplitude) with cross-sectionally measured trajectories that visually and statistically more closely parallel the synaptic density data and might therefore better index this developmentally critical brain variable (Feinberg, Thode et al. 1990). The NREM delta trajectories have recently been confirmed in a longitudinal study which also identified ages 12-16.5 yrs as a period of accelerated brain maturation (Campbell and Feinberg 2009). In contrast to the sleep EEG, MRI measures do not reveal any period of accelerated change, nor would one expect this in post-natal measures of brain structure.

While I do not believe the MRI studies index synaptic elimination, I strongly agree with Giedd et al that it is critical to establish the cellular basis of the developmental MRI changes. (Our laboratory has also emphasized this point (Campbell and Feinberg 2009)). However, it is unlikely that animal studies of the sort proposed by Giedd et al could provide conclusive answers because of species differences. For example, cortical synaptic development in monkeys is isochronic rather than heterochronic as in humans ((Rakic, Bourgeois et al. 1986)). What is urgently needed is a large scale multisite replication of Huttenlocher’s heroic, post-mortem study of human brains. New investigations could include biochemical as well as ultramicroscopic measures of dendritic components and morphology. Such data could also fill in the age gaps in Huttenlocher’s dataset, providing more detailed maturational trajectories for synaptic connectivity. In addition, when ante-mortem MRI studies are available, they could be related to the post-mortem findings, so that the anatomic basis of developmental MRI changes could be determined. Replications lack research glamor. In this instance, however, a replication (and major extension) is urgently needed to understand existing findings and to provide a firm foundation for interpreting MRI and other studies of post-natal brain development.

Campbell, I. G. and I. Feinberg (2009). "Longitudinal trajectories of non-rapid eye movement delta and theta EEG as indicators of adolescent brain maturation." Proceedings of the National Academy of Sciences of the United States of America 106(13): 5177-5180, PMID 19307577.

Feinberg, I., et al. (1990). "Gamma distribution model describes maturational curves for delta wave amplitude, cortical metabolic rate and synaptic density." Journal of Theoretical Biology 142(2): 149-161, PMID 2161971.

Huttenlocher, P. R. (1979). "Synaptic density in human frontal cortex - developmental changes and effects of aging." Brain Research 163(2): 195-205, PMID 427544.

Rakic, P., et al. (1986). "Concurrent overproduction of synapses in diverse regions of the primate cerebral cortex." Science 232(4747): 232-235, 3952506.