Lack of hippocampal volume differences in primary insomnia and good sleeper controls: An MRI volumetric study at 3 Tesla

John W. Winkelman, Kathleen L. Benson, Orfeu M. Buxton, In Kyoon Lyoo, Sujung Yoon, Shawn O'Connor, Perry F. Renshaw

Research output: Contribution to journalArticlepeer-review

85 Scopus citations


Background: A recent pilot study reported that hippocampal volume (HV) was reduced in patients with primary insomnia (PI) relative to normal sleepers. Loss of HV in PI might be due to chronic hyperarousal and/or chronic sleep debt. The aim of this study was to replicate the earlier pilot report while employing a larger sample, more rigorous screening criteria, and objective sleep data. Methods: This cross-sectional design included community recruits meeting DSM-IV criteria for PI (n= 20, 10 males, mean age 39.3 ± 8.7) or good sleeper controls (n= 15, 9 males, mean age 38.8 ± 5.3). All subjects were unmedicated and rigorously screened to exclude comorbid psychiatric and medical illness. PI subjects underwent overnight polysomnography to screen for sleep-related breathing and movement disorders. HV and total brain volumes were derived by MRI employing a Siemens/Trio scanner operating at 3. Tesla. Data also included 2. weeks of sleep diaries and wrist actigraphy. Results: Mean HV was 4322.0±299.7mm3 for the good sleeper controls and 4601.55±537.4mm3 for the PI group. The dependent variable, HV, was analyzed by ANCOVA. Main effects were diagnosis and gender; whole brain volume served as the covariate. Although the overall model was significant (F=6.3, p=0.001), the main effects of diagnosis (F=2.14) and gender (F=0.04) were not significant. The covariate of whole brain volume was significant (F=5.74, p=0.023) as was the interaction of diagnosis with gender (F=10.22, p=0.003), with male insomniacs having larger HVs than male controls. Conclusions: This study did not replicate a previously published report of HV loss in primary insomnia. Differences between our finding and the previous report might be due to sample composition and method of MRI assessment. Furthermore, we demonstrated no objective differences between the controls and PIs in actigraphic measures of sleep maintenance. Within the PIs, however, actigraphic measures of poor sleep maintenance were associated with smaller HV.

Original languageEnglish
Pages (from-to)576-582
Number of pages7
JournalSleep Medicine
Issue number6
StatePublished - Jun 2010


  • Brain imaging
  • Hippocampal volume
  • Insomnia
  • Magnetic resonance imaging
  • Sleep
  • Sleep disorders
  • Sleep efficiency


Dive into the research topics of 'Lack of hippocampal volume differences in primary insomnia and good sleeper controls: An MRI volumetric study at 3 Tesla'. Together they form a unique fingerprint.

Cite this