Circadian Rhythms Tied to Changes in Brain Morphology in a Densely Sampled Male

Abstract

Circadian, infradian, and seasonal changes in steroid hormone secretion have been tied to changes in brain volume in several mammalian species. However, the relationship between circadian changes in steroid hormone production and rhythmic changes in brain morphology in humans is largely unknown. Here, we examined the relationship between diurnal fluctuations in steroid hormones and multiscale brain morphology in a precision imaging study of a male who completed 40 MRI and serological assessments at 7 A.M. and 8 P.M. over the course of a month, targeting hormone concentrations at their peak and nadir. Diurnal fluctuations in steroid hormones were tied to pronounced changes in global and regional brain morphology. From morning to evening, total brain volume, gray matter volume, and cortical thickness decreased, coincident with decreases in steroid hormone concentrations (testosterone, estradiol, and cortisol). In parallel, cerebrospinal fluid and ventricle size increased from A.M. to P.M. Global changes were driven by decreases within the occipital and parietal cortices. These findings highlight natural rhythms in brain morphology that keep time with the diurnal ebb and flow of steroid hormones.

Keywords: MRI; brain structure; diurnal rhythms; precision imaging; steroid hormones.

Figure 1.

Precision imaging study design. A, For 30 consecutive days, a healthy 26-year-old male underwent brain imaging and serological assessment every 12–24 h for a total of 40 sessions. B, Steroid hormones demonstrated typical diurnal patterns, with testosterone (saliva), estradiol (only assessed with serum), and cortisol (saliva) peaking in the morning and significantly reduced in the evening. Note that saliva sampling was performed at each session (testosterone, cortisol; n = 40); a blood sample was obtained once per day (estradiol; n = 30), following human subjects protocol regulations. C, The subject completed mood assessments (e.g., sleep quality, mood, anxiety, stress) at each session. All measures were within the standard range and remained consistent across the study duration. Abbreviations: PSQI, Pittsburgh Sleep Quality Index; STAI, State-Trait Anxiety Inventory; PSS, Perceived Stress Scale. Steroid hormone concentrations by time of day are reported in Extended Data Table 1-1.

Figure 2.

Neuroanatomical changes throughout the day in a densely sampled male. A, Total brain volume and gray matter volume decreased over the course of the day, while white matter volume did not (at a Bonferroni-corrected threshold). B, Cerebrospinal fluid, lateral ventricle, and fourth ventricle volume all increased from morning to evening. Abbreviations: TBV, total brain volume; GMV, gray matter volume; WMV, white matter volume; CSF, cerebrospinal fluid; ns, not significant; *p < 0.006, **p < 0.001, ***p < 0.0001. Hormone and behavior correlations with brain morphology measures are reported in Extended Data Table 2-1 and Extended Data Figure 2-1, respectively.

Figure 3.

Regional neuroanatomical changes between morning and evening sessions. A, Multivariate regression demonstrates a mostly negative relationship between time of day and gray matter volume (FDR at q < 0.05; nonsignificant test statistics were set to zero for interpretability). B, Gray matter volume in the striate cortex and temporal occipital regions are heightened in the morning and reduced in the evening. Abbreviations: GMV, gray matter volume. For additional regions, see Extended Data Table 3-1 and for regional cortical thickness changes, see Extended Data Table 3-2.

Figure 4.

Subcortical gray matter volume by time of day. A, Multivariate regression for subcortical regions demonstrated decreasing volume throughout the day. Left, Purple tones indicate volumetric reduction while aqua tones indicate volumetric expansion from A.M. to P.M. (FDR at q < 0.05; nonsignificant test statistics were set to zero for interpretability). Gray matter volume was higher in the right cerebellum, brainstem, and right hippocampus in morning compared with evening sessions (see example box plots on the right). For the full list of subcortical results, see Extended Data Table 4-1. B, Diagram illustrates medial temporal lobe subregion segmentation from the ASHS parcellation. Gray matter volume in CA1, dentate gyrus, and subiculum did not change throughout the day. This pattern held for all medial temporal lobe subregions (CA2/3, entorhinal cortex, perirhinal cortex, and parahippocampal cortex, p > 0.007 for all, Extended Data Table 4-2). For a manual segmentation example, see Extended Data Figure 4-3. Abbreviations: GMV, gray matter volume; ns, not significant.