Comparison of decline in different cognitive domain in patients with normal pressure hydrocephalus

Idiopathic normal pressure hydrocephalus (iNPH) is a progressive syndrome manifested usually by the Hakim's triad: incontinence, gait disturbance and cognitive decline [1]. Gait disturbance is considered a primary symptom of iNPH and is present in 94–100% of patients. Cognitive decline is considered to be the second most prominent symptom reaching 78–98% frequency of onset [1].

In 1965, Adams et al. [2] described widening of lateral ventricles while maintaining unelevated intracranial pressure as a result of flow malfunction in cerebrospinal fluid (CSF). Hypoperfusion, glymphatic impairment, disturbance of metabolism, astrogliosis, neuroinflammation and blood-brain barrier disruption are speculated to cause white matter and gray matter lesions leading to typical iNPH symptoms [3]. The prevalence of iNPH is still discussed as there is a lack of homogeneity across the prevalence studies [4]. However, the recent review estimated prevalence ranging from 10/100 000 to 22/100 000 for probable iNPH and 29/100 000 for possible iNPH in the elderly population, increasing with higher age [4‐7].

Despite the importance of cognitive decline in a manifestation of iNPH, a more detailed description is only now starting to be researched. Generally, cognitive deficits in iNPH patients are described as frontal and subcortical syndrome [5]. Symptoms of frontal impairment include slowing of psychomotor speed, deficits in attention, working memory and executive functions [8]. Subcortical dementias include a very similar profile and are characterized in addition by visuoconstructional deficits and memory impairment that manifests itself in the domain of retrieval rather than learning. This is often referred to as a fronto-subcortical deficit. Compared to cortical dementias, subcortical dementias are characterized by the absence of aphasia, apraxia and agnosia. Furthermore, they are characterized by a non-hippocampal type of memory deficit and preserved awareness of their illness [9]. Of note is that the division of subcortical and cortical impairments is not universally accepted since the two types considerably overlap each other [10, 11].

One of the most affected cognitive areas in patients with iNPH is executive function [12‐14]. Executive functions contribute to planning, decision-making, judgment, inhibiting actions and other mental systems that lead to purposeful behavior and its regulation [9]. Furthermore, psychomotor pace is usually slowed in iNPH patients [15]. Psychomotor pace refers to the ability to perceive and respond to changes in the environment, such as the presentation of a specific stimulus [16]. Psychomotor pace can also be clinically assessed as the ability to respond verbally or motorically at an appropriate speed to a given situation. Some authors define psychomotor pace more broadly as the processing speed at which the brain is capable of working [17]. Although normal pressure hydrocephalus is not a typical cortical dementia like Alzheimer's disease, it is shown to also affect the patient's memory functions, however, it seems to affect retrieval rather than storing process of memory. Additionally, visuoconstructional abilities seem to be disrupted in patients with iNPH, especially when assessed by more complex figures [18]. This article aims to evaluate characteristics of the neuropsychological cognitive profile in iNPH patients and neuropsychological interdomain correlations.

The resulting sample consisted of 53 iNPH patients who were indicated for VP shunt implantation and underwent neuropsychological evaluations with the mean age of 73.21 ± 5.48 years, 64% (n=34) were men. The mean pre-LD in Dutch Gait Scale was 26.4 ± 9.9, while post-ELD Dutch Gait Score was 18.8 ± 10.8 (p<0.001). Moreover, the median Charlson’s comorbidity index was 6 points. All patients had at least two symptoms from the Hakim’s triad. The majority of patients harbored all three (77.7%). For the analysis, we attempted to use all the iNPH confirmed patients, however, not all of them accomplished all the required tests embraced in the neuropsychological battery. Hence, for the analysis of executive functions, we used data of 38 patients (36.6 ± 5.15); 37 patients (39.5 ± 10.12) for the analysis of non-verbal memory; 37 patients (34.1 ± 4.72) for the analysis of verbal memory; 40 patients (37.4 ± 8.14) for the analysis of attention/psychomotor speed. For the analysis of visuoconstruction, we used data of 38 patients (41.3 ± 8.72). The description of each sample for the analysis can be found in Table 1.

We found strong correlation between executive functions and verbal memory (r=0.829; p<0.001) or attention and visuoconstructional abilities (r=0.917; p<0.001). Moderate correlation was found between all neuropsychological domains and MoCA test, suggesting that cognitive decline progresses gradually across the domains. See Figure 2 for more details.

We found significant results for all of the observed cognitive domains with iNPH group scoring significantly lower (see Table 2). A statistically significant result was found for executive functions domain (t=-16.01, p<0.001) with a very large effect size (d=-2.6). Analogous results were obtained for (1) the verbal memory (t=-20.48, p<0.001, d=-3.36); (2) the visuoconstruction (t=-6.16, p<0.001, d=-0.99); and (3) the non-verbal memory (t=55.5, p<0.001, d=-1.04). Finally, also for (4), the attention/psychomotor speed, we found a statistically significant difference (t=18, p<0.001) with a large effect size (d=-1.542).

Our results confirm that iNPH patients are severely affected in all cognitive domains measured. Moreover, we found that verbal memory and executive functions are one of the most affected cognitive domains in this diagnosis. These findings concur with previous research [12, 20]. We found attention/psychomotor speed to be less impaired than executive functions and verbal memory. Additionally, we found non-verbal memory and visuoconstructional abilities to be the least impaired of the measured cognitive domains in iNPH patients. However, it is important to note that all of the measured cognitive domains were considered in regard to educational and age correlated norms and all of them were severely impaired in that respect.

As pointed out in the introduction of this article, all of these cognitive domains are associated with the iNPH diagnosis. There is no common understanding however, of how much the individual domains are impaired when viewed in relation to each other. Thus, we offer our results as an invitation to further scrutiny. We found that verbal memory is much more impaired than non-verbal. These results suggest that iNPH damages memory unevenly. It is an interesting result, nevertheless, it is important to note that we measured non-verbal memory with only one neuropsychological test (ROCFT), and thus we recommend using more neuropsychological tests in the future which measure non-verbal memory in more detail, such as for example Brief Visual Memory Test - Revised. Also, the results of verbal memory in our observed group relies heavily on AVLT, which is a difficult task to perform. Considering the progress of cognitive decline, it can be suggested that in later stages the AVLT test has a smaller differentiating ability and may thus lead to potentially biased results as the floor effect could slightly skew the results. On the other hand, we used Fonemic and Categorical Verbal Fluency tests to assess the same cognitive domain – verbal memory in iNPH patients, which should improve the results and provide a more comprehensive picture of iNPH patient’s memory. The strong correlation appeared between executive functioning and verbal memory, while not between executive functioning and non-verbal memory supporting the idea that memory is impaired unevenly based on the presented material.

Our results additionally confirmed that executive functions are severely impaired in iNPH. This is in accordance with previous research [12, 13, 18]. It seems that inhibition and shifting ability of executive functions are impaired [12, 32, 33], however data on ability to maintain the activity are missing and the data on updating ability measured by Digit span backward test are mixed, impaired [32] and not significant [34]. Detailed description of executive functioning might help better diagnostic quality of cognitive tests in the future.

From our study, it seems that visuoconstructional abilities are the least affected cognitive domain in iNPH patients, even though it is still impaired compared to healthy individuals. Possible explanation might be that visuoconstructional abilities are affected in the later stages of the illness and hence are less severely impaired in general. It might be interesting to observe the progress of visuoconstructional abilities in iNPH with respect to its potential diagnostic value. Lastly, we did not focus on language abilities in this study, however, from the previous research it is possible to assume that verbal fluency is impaired but naming ability mostly remains intact [15].

Compared to Alzheimer's disease (AD), iNPH cognitive profile is characterized by frontal lobe symptoms and disproportionately impaired memory function [8]. Spontaneous recall seems to be impaired in both, AD [18, 35] and iNPH, however, in iNPH the role of clue is crucial and should normalize the memory performance as it is theorized as subcortical type of cognitive decline [5]. It might be due to a fact that iNPH does not typically affect the hippocampus as AD does [36]. Saito et al. [18] highlight the episodic part of memory to be more impaired in AD than in iNPH and also depict visuoconstructional abilities as more damaged in iNPH than AD. This finding is supported by Nerg et al. (2021) [37]. They used a comprehensive CERAD Neuropsychological Battery in iNPH, AD and healthy patients and concluded that 1) iNPH are more affected on all performed tests than healthy patients; 2) iNPH performed worse on verbal fluency (which might reflect either slower psychomotor pace or impaired semantic fluency) and Clock Drawing test (which might reflect worse visuoconstructional abilities); 3) AD patients had more severe episodic memory dysfunction than iNPH (and healthy patients). Moreover, attentional functions seem to be less severely impaired in AD patients rather than memory [38], unlike in iNPH where the attention along with psychomotor pace is theorized to be impaired quite early in the process [15]. On the other hand, in Frontotemporal Lobar Degeneration (FTLD), the cognitive decline is manifested by the impairment in executive functions and slower psychomotor speed much like we observe in iNPH [39]. Nevertheless, for example Diehl and Kurz [38] found in their study that patients with frontotemporal dementia performed better in visuoconstructional abilities not only in comparison to AD group but also performed equally in comparison to healthy control group. This would suggest no visuoconstructional impairment in Frontotemporal Dementia (FTD) and might be a differentiating factor between FTD and iNPH. Also, along with FTD and especially its most common subgroup behavioral variant FTD, there are often psychiatric symptoms and frequent changes in personalities [40, 41]. This is not a well studied topic in iNPH and might be a very interesting field to research in the future. Lastly, compared to Parkinson's disease (PD), iNPH cognitive disabilities arise earlier in the stages of the illness and with more diffuse impairment. Only 25% of PD showed executive dysfunctions compared to 65 % of iNPH [42].

One of the limitations is the comparison of iNPH patients to standardized norms from different neuropsychological tests rather than healthy adult control groups. However, comparison of different cognitive domains in this study might serve as a solid ground to build a profound cognitive profile of iNPH patients. Secondarily, we did not discriminate between different stages of the illness in our cohort, thus reflecting one cognitive profile instead of detailed description of the development of overall iNPH cognitive decline. Also, the results might be skewed due to lack of differentiation between progression of the illness as the cognitive decline in later stages is global and cannot be easily differentiated from other cognitive declines such as the one with AD or PD [15]. Some cognitive domains were not covered with more neuropsychological tests and it might be worth examining them in more depth in the future. Lastly, the interpretation of A/PS domain is limited, as we used ROCFT for assessing this domain, which is not traditional, but it has been assessed in a recent publication [27]. We believe that it can provide useful insight for further scrutiny.

Patients with iNPH are affected in all cognitive domains and the cognitive decline is uneven across these domains. The impairment of memory depends on the presented material. Verbal memory seems to be much more severely affected than non-verbal memory. Along with VM, the EFs are the most affected. Attention/Psychomotor speed is less affected than VM and EFs. The least affected domains are nVM and VA. However, describing the progress of cognitive decline in more detail might be fruitful in future. Nevertheless, overall our study might be used for researchers to focus cognitive functioning in iNPH in a more precise way and for practitioners, it might serve as a cornerstone for deciding which neuropsychological tests to deploy.