EEG Delta Power Decline Can Begin Before Age 11: A Reply to Campbell and Feinberg

Abstract

CAMPBELL AND FEINBERG1 QUESTION OUR CONCLUSION THAT THE ADOLESCENT SLEEP EEG POWER DECLINE CAN OCCUR BEFORE AGE 11 YEARS BY presenting averaged data from children whose sleep was studied on unspecified “usual” schedules. They state that delta power decline begins after age 11 and hypothesize that this age “…signals the onset of brain adolescence” characterized by “…massive, parallel declines in cerebral metabolic rate, synaptic density and delta wave amplitude.”1 Findings from neuroanatomical studies suggest that defining the onset of cortical reorganization in the healthy maturing brain is not so straightforward. By stating that this process does not begin until after age 11, Campbell and Feinberg overlook such important factors as (1) brain maturation occurs at different times in different cortical regions and (2) the timing of maturation is associated with significant individual variability (in part related to sex2). Studies using MRI,2,3 postmortem synaptic density,4 waking EEG,5 and PET6 indicate developmental differences among brain regions. For example, an MRI study showed peak cortical thickness at different ages in different regions3; a corollary is that the age-related decline began at different ages for different regions. Decline in gray matter volume is thought to reflect the elimination of synapses, seen as reduced EEG amplitude.5 Furthermore, Feinberg et al.7 have modeled the maturational curves for delta wave amplitude in the sleep EEG showing the age of peak delta wave amplitude at 7.4 years. The recent Campbell and Feinberg8 data derive from a single EEG derivation C3/A2 or C4/A1 (used interchangeably), which lies over the postcentral gyrus.9 Shaw et al.3 estimated 8.5 years as the age of peak cortical thickness in this region. Our data,10 collected under conditions where prior sleep and in-lab time in bed were held constant, support the notion that the sleep EEG can manifest cortical maturation at an early age. For example, we found a 55% reduction in delta power and a 52% decline in overall power (PΔ of −44%/year, Figure 1)10 at C3/A2 for one boy from age 9.4 to 11.2 years. Furthermore, the decline in sleep EEG power occurred for frequencies up to 16 Hz (Figure 4).10 Our data also showed that the sleep EEG power changes differed regionally (see Figures 2 and 4).10 Our findings show that the maturational timing varies, cannot be pinpointed to a specific age, and can begin before age 11. The individual variability shown in Figure 1 and Table 110 emphasizes this point. Indeed, large individual differences appear in Campbell and Feinberg's8 Figure 2, where several participants show decreases of delta power before age 11 and others large increases in delta power after age 11. We conclude by congratulating Campbell and Feinberg on their longitudinal studies and Feinberg for his pioneering work. Our comments were meant to highlight methodological differences in control of such variables as prior sleep and sleeping environment that impact the sleep EEG and may account for the observed differences. Finally, neither our data set nor that of Campbell and Feinberg8 can provide a definitive statement about the onset of “brain adolescence” without including younger children in whom synaptic pruning has not yet begun.