The Epidemiology of Idiopathic Normal Pressure Hydrocephalus Prevalence, Incidence, and Radiological Features in 70-Year-Olds

Abstract

Background: Idiopathic normal pressure hydrocephalus (iNPH) is a potentially reversible but underdiagnosed condition in older adults. Population-based data on disease incidence and progression remain limited. This thesis aimed to estimate the prevalence, incidence, and six- year progression of iNPH in a population-based cohort of 70-year-olds; evaluate the predictive value of vascular risk factors (VrFs), clinical symptoms, and MRI findings for iNPH development; and propose reference values for morphological and volumetric imaging markers. Methods: Papers I and II were cross-sectional studies including 791 participants from the Gothenburg H70 Studies (examined 2014–2016, mean (SD) age 71 (0.3), who underwent clinical assessments and brain MRI. Gait, balance, cognition, and urinary continence were assessed, and iNPH was diagnosed using modified International Guidelines (IG). Paper I estimated iNPH prevalence and Paper II assessed the diagnostic performance of 12 imaging markers, including Evans’ index (EI), z-EI, brain per ventricle ratios (BVRAC, BVRPC), sulcal compression, Sylvian fissure enlargement, callosal angle, and diameters of the temporal horns, third and fourth ventricles, midbrain, and pons. Corresponding volumetric measurements were obtained using FreeSurfer. Paper III was a longitudinal follow-up (2019–2022) at age 77 (0.8), with repeated clinical and MRI assessments in 467 participants from baseline and 90 newly examined (n=557 at age 77). Participants were classified as possible iNPH, asymptomatic iNPH, radiologic iNPH with isolated symptoms, or non-iNPH. Analyses included prevalence, incidence, and predictive modelling. Results: At age 71, the prevalence of possible and asymptomatic iNPH was 1.5% each; by age 77, this increased to 2.9% and 2.3%, respectively. Paper II proposed reference values and found z-EI and BVR to be more strongly associated with lateral ventricular volume than EI. The six- year incidence of possible and asymptomatic iNPH was 4.8 and 1.9 cases per 1,000 person- years, respectively. z-EI, BVRAC, and callosal angle were the strongest radiologic predictors of iNPH. Most VrFs were not associated with iNPH, though higher BMI increased iNPH risk. Radiological changes often preceded symptoms; during follow-up, 43% of asymptomatic iNPH and 67% of radiologic iNPH with isolated symptoms progressed to symptomatic iNPH. Conclusions: iNPH is more common than previously recognized in this age group, with radiological signs often preceding symptoms, underscoring the need for early detection. Population-based reference values were established, and novel markers such as z-EI and BVR showed potential for improving diagnostic accuracy. These findings offer new insights into iNPH epidemiology and support refinement of diagnostic criteria.