Macey · NeuroImage. Clinical 2018 · case-control study · n=1045

Sex-specific hippocampus volume changes in obstructive sleep apnea.

Cited 100 times in the scientific literature.

Level 4 - case-series / case-control

Case-control neuroimaging study

PubMed 30101062 · doi:10.1016/j.nicl.2018.07.027 · record verified 2026-08-31

What was done

High-resolution T1-weighted MRI scans were analyzed to assess hippocampal sub-regional morphology in 66 newly diagnosed, untreated obstructive sleep apnea (OSA) patients (mean age 46.3 ± 8.8 years; AHI 34.1 ± 21.5 events/h; 50 males, 16 females) compared with 979 age-matched controls (426 males, mean age 46.5 ± 9.9 years; 59 recruited controls plus controls from IXI and OASIS-MRI datasets). Hippocampal surfaces were segmented with FSL FIRST, scaled for total intracranial volume, and compared between groups and sexes using ANCOVA (corrected P < .05).

What was found

OSA patients overall showed small areas of greater bilateral volume in CA1 (surface displacement ≤ 0.56 mm), subiculum, and uncus, and smaller volume in right posterior CA3/dentate (≥ -0.23 mm). Males with OSA showed bilateral volume increases throughout CA1 and subiculum (≤ 0.61 mm, larger on the right) and bilateral reductions in mid- and posterior-CA3/dentate (≥ -0.45 mm). Females with OSA showed only right-sided differences: increased CA1 and subiculum/uncus volume (≤ 0.67 mm), decreased posterior CA3/dentate volume (≥ -0.52 mm), and right hippocampal head and tail volume reductions not seen in males.

Why it matters

This study demonstrates that untreated OSA is associated with distinct, sex-specific patterns of hippocampal sub-regional structural alterations. These morphological variations may help explain sex differences in autonomic, memory, and affective symptoms observed clinically in OSA.

Limits

The study is cross-sectional, precluding causal inference. The female OSA sample was small (n = 16), limiting statistical power. Controls were pooled from disparate external MRI databases, introducing potential heterogeneity in scanner hardware and acquisition protocols. Proposed cellular mechanisms (inflammation versus neuronal injury) were speculative and not directly measured.

Cited by