Discussion

To the best of our knowledge, this is the first study to explore the relationship between degenerative spinal stenosis and hydrocephalus-related cranial imaging characteristics in NPH patients. Although our data cannot provide measures of craniospinal compliance or CSF flow dynamics, the associations found between spinal stenosis and NPH-related imaging features may support the hypothesis, that spinal canal narrowing promotes hydrocephalus formation through alterations of CSF hydrodynamics. Our data therefore seem consistent with findings of in vitro studies (Benninghaus et al., 2026), which recently demonstrated that cervical stenosis and lumbar stenosis increase CSF-flow resistance, resulting in a significant reduction in dynamic compliance and spinal CSF-flow. This led to an increase in intracranial pressure amplitudes (of up to 7.85 mmHg), which is believed to promote hydrocephalus formation (Eide, 2006).

In addition, a relationship between spinal canal measurements and ventricular size has recently been described even independent of NPH. Grosu et al. (2024) analyzed potential associations between the cervical spinal canal width and scoliosis, with white matter, grey matter and ventricular volumes. They found that a smaller width of the cervical spinal canal at the vertebrae level C2/3 was correlated with lower grey and white matter volumes, and a larger ventricular size. Based on these findings, authors suggested that spinal characteristics constitute independent risk factors for neurodegeneration, which is consistent with the results of our present study. Moreover, narrowing of the lumbosacral canal causing prestenotic slowing and reduced spinal CSF-flow (Kim et al., 2021) has previously been suggested as causative for an increase in CSF albumin quotient and protein concentration reducing CSF-reabsorption (Seyfert et al., 2002), which may also contribute to hydrocephalus formation.

In our cohort, the mean CA value (67° ± 16) was smaller than that reported for healthy subjects or patients with other neurodegenerative diseases in the literature. This value is therefore well within the expected range for NPH patients (Ishii et al., 2008). Such narrowing of the CA is considered a reliable biomarker for NPH, with a diagnostic accuracy of up to 93%, depending on the chosen threshold angle (Ishii et al., 2008; Baroncini et al., 2018; Miskin et al., 2017). Unlike healthy subjects, who typically exhibit larger callosal angles (mean 112° ± 11) than NPH patients, larger callosal angles observed in other neurodegenerative diseases (as e.g. in Alzheimer's disease: mean 104° ± 15) are indicative of e vacuo widening resulting from brain atrophy (Ishii et al., 2008). With CA having been described to be significantly smaller in NPH patients as compared to other neurodegenerative conditions (Ishii et al., 2008), those NPH-patients with CA in the expected range, but with relatively larger CA have previously been reported to be less responsive to CSF diversion treatment than those with relatively smaller CA. A cutoff value of 63° has been reported to have the best prognostic accuracy (Virhammar et al., 2014). This may suggests, that relatively larger callosal angles in NPH patients may be associated with secondary neurodegeneration and brain atrophy, and that symptoms are less likely to be reversible under CSF diversion. In our study, we found a trend towards larger CA in CS patients with smaller AP, as well as in LS patients with smaller DSCSA. As our analysis accounted for age, purely age-related effects were excluded. These findings therefore might support the hypothesis, that degenerative spinal stenoses chronically promote NPH by increasing CSF flow-resistance, reducing craniospinal compliance, and leading to higher intracranial CSF pressure amplitudes, which may promote secondary neurodegeneration in the long term.

However, some NPH patients may not exhibit an association between spinal canal narrowing and callosal angle, if neurodegeneration has not yet progressed, so that dynamic measures as such as CSF flow-velocity or CSF pulsatility would be required to actually capture the effects of spinal stenosis on intraventricular CSF hydrodynamics. It therefore might be, that NPH patients are more heterogenous in their associations between spinal pathology and static NPH imaging parameters, depending on the disease stage and the degree of secondary neurodegeneration, which could have contributed to the only borderline findings in our study. To validate this hypothesis, future studies are required that include larger patient samples and prospectively acquired imaging data such as dynamic CSF-flow measures (e.g. heavily T2-weighted constructive interference in steady state (CISS) sequences or cardiac-gated phase-contrast cinemode MRI of CSF), rather than only static imaging parameters.

NPH-grading scores significantly improved after shunt surgery, confirming a relevant degree of hydrocephalus-related symptoms responsive to CSF-diversion. However, it cannot be excluded, that spinal stenosis-related symptoms impacted the clinical presentation. A recent study by Tominaga et al. (2023) compared NPH patients with and without lumbar stenosis after shunt surgery and found, that those patients with coexisting LS experienced worse outcomes regarding gait functions. Therefore, coexisting spinal stenosis should be considered when assessing NPH patients and evaluating treatment outcomes. The prevalence of lumbar spinal stenosis in this cohort of 224 NPH patients, as previously reported by Tominaga et al. (2023), was 32.6%. However, based on the retrospective nature of our study and the incomplete spinal imaging data available, we were unable to determine the prevalence here. The prevalence of LS in the general population has been reported as ranging from 11 to 38% in the literature, but much higher prevalence have been found in older age groups (Jensen et al., 2020). It should also be noted, that the reported prevalence rate varies in the literature, as different studies have used different definitions for spinal stenosis. While some studies used only clinically defined criteria, others defined stenoses solely on the basis of radiological findings, whereas still others included only patients with symptomatic spinal canal stenoses (Kato et al., 2015; Jensen et al., 2020, Ishimoto et al., 2012). Future studies analyzing the prevalence of degenerative spine disease in NPH patients should use standardized spinal stenosis definitions based on radiological criteria. Determining the prevalence and characteristics of spinal stenosis in NPH patients could help to elucidate potential associations between degenerative spine disease and NPH. Therefore, larger epidemiological studies and the availability of nation-wide patient registries are needed. Taking into account concomittant degenerative spine disease in patients with NPH may also contribute to a better understanding of the substantial interindividual variability in responsiveness to the spinal tap test. Furthermore, those NPH patients undergoing surgical treatment for spinal stenosis should be longitudinally assessed to evaluate, whether spinal treatment might in the long term also positively impact on NPH disease dynamics.

Limitations

The major limitation of the present study is its retrospective nature, with spinal imaging data available for only a fraction of NPH patients screened. Therefore, selection bias must be taken into account, as those patients with spinal MRI scans may differ systematically from those without, which prevents generalizability of the present findings. Furthermore, statistical power was limited due to the heterogenous data structure with missing data, and only small sample sizes, which might have contributed to the only borderline findings of the present study. Although uncorrected analyses suggested significant effects, these were no longer significant after adjustment for multiple comparisons. Nevertheless, the corresponding effect sizes were consistently large, suggesting potentially robust underlying effects that warrant replication in larger samples. The limited sample size also prevented further analysis of the potential impact of specific spinal stenosis characteristics such as the level and number of segments affected or single versus multiregional involvement. A more detailed, systematic analysis of the potential impact of these characteristics on hydrocephalus, could therefore not be conducted.

Prospective acquisition of clinical and standardized total spine magnetic resonance imaging (including spinal canal volumetry with dynamic examinations using advanced MRI techniques to analyse cerebrospinal fluid pulsations and flow), as well as long-term follow-up with comparison to age-matched controls is therefore needed to investigate whether degenerative spine disease has an influence on NPH formation and disease dynamics.

While we recognise that the current study is only exploratory and rather hypothesis-generating than confirmative, we believe that focusing on the spinal compartment as a potentially relevant pathophysiological factor for NPH disease dynamics provides a novel and clinically highly relevant perspective that should be further pursued.