Olfactory Dysfunction as a Marker for Cognitive Impairment in General Paresis of the Insane: A Clinical Study

Introduction

General paresis of the insane (GPI) is one of the most serious and harmful clinical types of neurosyphilis. The incidence of GPI has been on the rise in recent years due to the increase in syphilis cases worldwide, with 10.6 million new cases reported annually according to the WHO [1,2]. GPI is largely reversible if it is diagnosed early and standard treatment is administered [3]. However, the clinical manifestations of GPI, including cognitive impairment, neuropsychiatric symptoms, and brain structural abnormalities, are nonspecific, which pose challenges in diagnosis [4]. Consequently, a marker that can indicate GPI is needed to assist in the early diagnosis of GPI.

Olfactory dysfunction is a common marker for a broad spectrum of neuropsychiatric disorders. A number of studies have reported olfactory dysfunction in individuals with neurodegenerative diseases such as Parkinson’s disease (PD) and Alzheimer’s disease (AD) [5,6]. With respect to psychiatric disorders, deficits in olfactory processes have been extensively reported in major depressive disorders [7,8], bipolar disorders [9,10] and schizophrenia [11,12]. Collectively, these observations highlight the associations of olfactory dysfunction with cognitive impairment and neuropsychiatric symptoms. We assumed that olfactory dysfunction may be a potential marker for GPI that can reflect the clinical symptoms.

Several areas in the basal frontal and medial temporal lobes are collectively referred to as the ‘primary olfactory cortex,’ including the piriform cortex, amygdala, and entorhinal cortex. Within these areas, neurons project to various brain regions in the limbic lobe and frontal cortex, including the orbitofrontal cortex (OFC), insula, and hippocampus. These regions are commonly referred to as the ‘secondary olfactory cortex’ [13]. Three case reports for GPI in the early stage (in 1951, 1958, and 2018) reported postmortem pathological findings of olfactory bulb (OB) atrophy [14–16]. Several MRI studies have revealed that patients with GPI exhibit brain structural abnormalities in olfactory brain regions, including the temporal lobe (59.1%), frontal lobe (30.0%), and bilateral hippocampus [4,17]. Furthermore, previous work has demonstrate that the medial temporal lobe atrophy (MTA) and white matter hyperintensity (WMH) focused in the frontal cortex and bilateral temporal areas are common in patients with GPI and are associated with cognitive function [18].

The assessment of olfactory function consists of several components: (1) the perception of odors at low concentrations (odor threshold, OT), (2) the nonverbal distinction of different smells (odor discrimination, OD), (3) the ability to name or associate an odor (odor identification, OI), (4) the hedonic valence of odors (odor pleasantness), and (5) the familiarity of odors [19]. OT tests and OD or OI tests are relatively independent measures of olfactory function [19]. OT dysfunction is thought to be due more to the damage to peripheral sensory organs of the olfactory system, whereas OD and OI dysfunction may be more attributable to impairments in the central olfactory brain regions, which are involved in higher levels of cognitive processing [20]. Additionally, OI deficits may be the result of a combination of sensory and cognitive impairments [21]. Since diseases can have different impacts across different olfactory functions, it is necessary to explore the pattern of olfactory dysfunction in GPI.

In this study, we investigated the presence and pattern of olfactory dysfunction in patients with GPI. We hypothesized that patients with GPI exhibit general olfactory dysfunction, and that the degree of olfactory dysfunction is associated with clinical manifestations, including cognitive impairment, neuropsychiatric symptoms, blood and CSF biomarkers, and brain structural abnormalities. Furthermore, the present study applied mediation analysis to investigate whether impaired olfactory function in GPI is attributed to peripheral sensation or central processing deficits.

Material and Methods

SUBJECTS:

Forty patients with GPI were recruited from The Affiliated Brain Hospital of Guangzhou Medical University, and 37 age- and sex-matched healthy controls (HCs) were recruited from communities in Guangzhou. This study was approved by the Ethics Committee of the Affiliated Brain Hospital of Guangzhou Medical University. Written informed consent was obtained from the participants or their next of kin.

The inclusion criteria for patients with GPI were: (1) positive blood and CSF rapid plasma reagin circle card test (RPR) and Treponema pallidum particle agglutination test (TPPA); (2) met the criteria for dementia in the Diagnostic and Statistical Manual of Mental Disorders, 4^th edition (DSM-IV); and (3) clinical stage and diagnosis confirmed by trained physicians from our hospital. HCs were included if they: (1) exhibited normal global cognitive function as evidenced by Mini-Mental State Examination (MMSE) scores higher than the adjusted scores (illiterate ≥17 points, primary school ≥20 points, and above middle school ≥24); and (2) had no history of syphilis.

The exclusion criteria for all subjects were: (1) history of cognitive disorder; (2) family history of dementia; (3) neurological disease, such as a brain tumor or stroke; and (4) other conditions that significantly influence olfactory function, including active upper respiratory/sinus infection or respiratory distress at the time of testing, congenital or traumatic anosmia, known nasal polyps or tumors, current or recent (past 6 months) smoking, and alcohol or substance dependence. The diagnoses were made by at least 2 neurologists with dementia expertise.

NEUROPSYCHOLOGICAL ASSESSMENTS:

The Chinese version of the MMSE (score range 0–30) was utilized to assess global cognitive function, including orientation, memory, attention, calculation, and language skills [22]. The Neuropsychiatric Inventory (NPI) was employed to measure neuropsychiatric symptoms [23]. The NPI is a structured caregiver interview comprising 12 subscales: delusions, hallucinations, agitation, depression, anxiety, euphoria, apathy, disinhibition, irritability, aberrant motor behavior, nighttime behavioral disturbances, and appetite. Based on scripted questions posed to caregivers, the severity (rated 1–3) and frequency (rated 1–4) of each neuropsychiatric symptom over the past month were graded. The composite score of each subscale was obtained by multiplying the frequency and severity scores, and then the sum of the scores from each subscale was added to calculate the total NPI score (score range, 0–144).

ASSESSMENT OF OLFACTORY FUNCTION:

The olfactory function of all participants was measured using the “Sniffin Sticks” test battery, which consists of 3 subtests: OT, OD, and OI [24]. Odors were presented using pen-shaped dispensers. In each subtest, the operator removed the pen cap and positioned the felt-tip of the pen approximately 2 cm in front of the participant’s nostrils on both sides, for approximately 3 seconds.

OT was evaluated using a forced-choice paradigm, where participants were required to differentiate between 1 odor (phenethyl alcohol, PEA) and 2 blanks (filled with the solvent propylene glycol). Starting from the lowest concentration of PEA, the concentration was gradually increased until the participant could correctly identify it twice in a row, which defined the first reversal point. Subsequently, testing was performed with a concentration level lower by 1 step (a higher numerical value indicates a lower concentration level), continuing until the participant made an incorrect judgement, marking the second reversal point. This staircase test was repeated 7 times, with 2 correct or 1 incorrect answer leading to a decrease or increase in concentration, respectively. The average of the last 4 reversal points was designated as the odor threshold score.

For OD, participants were presented with 3 odorized pens where 2 pens had the same odor and the third had a different scent. The participants were asked to identify the different pen. The last subtest performed was the OI subtest, where 16 pens with different odors were presented. Participants were instructed to choose the item that most accurately described the odor from 16 sets of 4 option cards. The results of the 3 subtests are summed up for a composite TDI score. According to the normative data of the Sniffin’ Sticks, TDI scores less than 16.5 were defined as anosmia, and TDI scores less than 30.3 (in people aged 16–35 years), 27.3 (in people aged 36–55 years), or 19.6 (in people over age 55 years) were defined as hyposmia [25].

BLOOD AND CSF BIOMARKER COLLECTION:

The rapid plasma reagin circle card test (RPR) and Treponema pallidum particle agglutination test (TPPA) results were obtained from blood and cerebral spinal fluid (CSF) samples collected from all patients with GPI. We also summarized the white blood cell (WBC) count and protein level in the CSF.

NEUROIMAGING ASSESSMENT:

A Philips 3.0 T MR system at the Affiliated Brain Hospital of Guangzhou Medical University was used to acquire the imaging data. The MTA visual rating scale [26] was used to evaluate the severity of MTA, which was divided into 5 grades, from 0 to 4. Higher grades indicated more severe MTA. The severity of WMH was evaluated by the Fazekas scale [27] and the score ranged from 0 to 3, with a higher score representing a larger range of WMH.

STATISTICS:

SPSS version 25.0 (IBM) was used to perform the statistical analyses. The differences in the demographics, clinical characteristics, and olfactory function between the groups were evaluated by performing χ² analysis, independent sample t tests, and Wilcoxon Mann-Whitney tests as needed. Partial correlation analyses were performed to explore the associations between olfactory function and cognitive function, neuropsychiatric symptoms, blood and CSF biomarkers, and brain structural abnormalities in patients with GPI, and the control variables included age and sex. Furthermore, partial correlation analyses were conducted for all subjects to explore the relationships between olfactory function and cognitive function in the overall sample. Benjamini-Hochberg correction was applied to correct P values in the correlation analysis.

We conducted mediation analyses to investigate whether sensory or cognitive impairments collectively contributed to OI dysfunction in patients with GPI. When the following conditions were met, a mediation model was established: (1) the independent variable (IV) had a significant effect on the dependent variable (DV); (2) the IV had a significant predictive effect on the mediator; (3) the mediator had a significant influence on the DV; and (4) when the mediator was excluded from the model, the effect of the IV on the DV weakened. In our analysis, group was regarded as the IV, OI scores were regarded as DVs, and MMSE and OT scores were regarded as mediators. PROCESS 3.2 was used to investigate the mediation model among variables. Indirect effects were estimated with 5000 bootstrapped samples. Moreover, the Sobel test was performed to verify whether the mediating effect was significant.

Results

DEMOGRAPHIC DATA, CLINICAL CHARACTERISTICS, AND OLFACTORY FUNCTION OF THE DIFFERENT GROUPS:

The demographic, clinical, and olfactory information of all the subjects is shown in Table 1. Compared with the HC group, the patients with GPI had lower MMSE scores and higher total NPI scores (P<0.001). In terms of olfactory function, the patients with GPI exhibited significantly lower OT, OD, OI, and TDI scores compared with the HC group (all P<0.01) (Figure 1). In addition, we selected some clinical MR images to demonstrate neuroimaging abnormalities, including medial temporal lobe atrophy and white matter hyperintensity, in patients with GPI (Figure 2).

NEUROPSYCHIATRIC SYMPTOMS IN PATIENTS WITH GPI:

The percentage of each neuropsychiatric symptoms in patients with GPI is presented in Table 2 and the scores for each symptom are shown in Figure 3. As the results show, all 12 symptoms in the NPI were found in patients with GPI, with nighttime behavioral disturbances (78.4%), irritability (51.4%), and apathy (51.4%) being the 3 symptoms with the highest incidence and scores.

CORRELATIONS BETWEEN OLFACTORY FUNCTION AND COGNITIVE FUNCTION, NEUROPSYCHIATRIC SYMPTOMS, BLOOD AND CSF BIOMARKERS, AND BRAIN STRUCTURAL ABNORMALITIES:

For the patients with GPI, there was a positive correlation between the OI and MMSE scores adjusted for age and sex (r=0.570, P<0.001) (Figure 4C). However, there was no significant correlation between olfactory function and neuropsychiatric symptoms, blood or CSF TPPA, blood or CSF RPR, CSF WBC, CSF protein, MTA scale scores, or Fazekas scale scores (P>0.05).

Across all subjects, partial correlation analyses showed that the MMSE score was positively associated with OT (r=0.419, P<0.001), OD (r=0.597, P<0.001), OI (r=0.680, P<0.001), and TDI scores (r=0.687, P<0.001) after adjustment for sex and age (Figure 4A–4D).

THE REDUCED OI OF PATIENTS WITH GPI WAS MEDIATED BY DECREASED MMSE AND OT SCORES:

The total effect of group (GPI vs HCs) on OI was β=2.965 (t=5.853, P<0.001), indicating a significant positive effect of group on OI. The indirect effect of group on OI through the MMSE was 2.066 (z=4.378, P<0.001), which suggested that the MMSE mediated the effect of group on OI. After controlling for MMSE (path c’ in Figure 5A), the remaining direct effect of group on OI was not significant at β=0.899 (t=1.551, P=0.125), with the effect of group on MMSE being β=9.258 (t=7.510, P<0.001) and the effect of MMSE on OI being β=0.223 (t=5.363, P<0.001).

The total effect of group (GPI vs HCs) on OI was β=2.865 (t=5.705, P<0.001), which manifested a significant positive effect of group on OI. The indirect effect of group on OI through OT was 0.588 (z=2.233, P=0.025), which suggested that OT as a mediator annulated the effect of group on OI. After controlling for OT (path c’ on Figure 5B), the remaining direct effect of group on OI was still significant at β=2.277 (t=4.412, P<0.001), with the effect of group on OT being β=1.818 (t=3.544, P<0.001) and the effect of OT on OI being β=0.323 (t=2.988, P<0.001).

No significantly mediated effect was found in other indicators on the association between group and OI (P>0.05). In conclusion, the above results showed that the MMSE and OT mediated the association between group and OI.

Discussion

The present study aimed to investigate the presence and pattern of olfactory dysfunction in patients with GPI and its association with clinical manifestations, including cognitive impairment, neuropsychiatric symptoms, blood and CSF biomarkers, and brain structural abnormalities. The primary findings were as follows: (1) Compared with HCs, the patients with GPI exhibited significant overall olfactory dysfunction including deficits in OT, OD, and OI; (2) Among patients with GPI, there was a positive correlation between the OI and MMSE scores, whereas no significant correlations were found between olfactory function and other clinical manifestations including neuropsychiatric symptoms, blood or CSF TPPA, blood or CSF RPR, CSF WBC, CSF protein, MTA scale scores, or Fazekas scale scores; and (3) Mediation analysis indicated that impaired OI in patients with GPI was mediated by global cognitive function (MMSE) and OT.

Odor perception requires a functional peripheral sensory organ (olfactory epithelium [OE] and OB) and central pathways [13]. The impaired OT, OD, and OI in patients with GPI may indicate that both the peripheral sensory organ and the olfactory cortex are affected by GPI. Specifically, OT has been shown to correlate with the volume of the OB, a structure related to peripheral sensory input [28]. OT dysfunction is believed to result from changes in the detection threshold caused by damage to the peripheral structure (eg, the OE and OB) of the olfactory system [20]. Previous studies have demonstrated that individuals with OB atrophy [8] or OE damage [7] tend to exhibit decreased OT. As a result, OT dysfunction in patients with GPI suggests that the peripheral structure of the olfactory system has been damaged by GPI, which is supported by the findings of 3 autopsy reports demonstrating OB atrophy in patients with GPI [14–16].

OD and OI deficits might be a consequence of a combination of sensory and cognitive impairments. During OD testing, the subject is required to detect similarities and differences between odorants. Thus, discrimination tasks put demands on sensory function and executive functions, as subjects require the ability to form a transient representation of a target odor and to use this representation for comparison and subsequent selection of the target odor from the distractors [20]. In the assessment of OI, higher cognitive processes such as olfactory memory, semantic understanding, and object naming are recruited [29,30]. Studies indicate that the perceived intensity of the odor is a predictor of higher order olfactory performance [31]. We found that both the OD and OI were significantly correlated with the MMSE score in the overall sample, whereas in patients with GPI, a correlation between the OD score and MMSE score was not detected. This finding suggests that cognitive deficits in patients with GPI may specifically affect the ability to identify and label odors rather than the ability to discriminate between different odors. In patients with GPI, we also observed a significant correlation between OT and OD scores, whereas no correlation was observed between OT and OI scores. This finding indicates that basic sensory abilities are foundational to the ability to discriminate between different odors. If a patient has trouble detecting an odor, it naturally follows that discriminating between odors would also be compromised. Thus, the impairment in OD might be partly driven by the deficits in OT.

Notably, significant correlations between OI and cognitive function were observed in both the overall sample and patients with GPI. Furthermore, the results of mediation analysis indicated that cognitive impairment and impaired OT in patients with GPI contribute to the impaired OI. A relation between a decrease in general cognitive abilities and decreased OI has been observed [32,33]. Numerous studies have confirmed that OI dysfunction is associated with cortical thinning and reduced grey matter volume in brain regions involved in olfaction and cognition processing, such as the entorhinal cortex, hippocampus, orbitofrontal cortex, and prefrontal cortex [34,35]. Thus, our findings suggest that OI may serve as a marker reflecting the cognitive function in GPI.

Previous studies have demonstrated that affective symptoms and psychosis symptoms are associated with olfactory dysfunction in patients with AD [36], major depressive disorder [7], schizophrenia [12], and bipolar disorder [9]. There was no correlation between olfactory function and neuropsychiatric symptoms in patients with GPI. Regions, including the amygdala, hippocampus, insula, anterior cingulate cortex, and anterior orbital cortex, have been described as common neural substrates for emotional symptoms, psychotic symptoms, and olfactory processing. Neuroradiological examinations of patients with neurosyphilis have demonstrated that the hyperintensities, hypoperfusion, or atrophy detected by MRI extend across the whole brain, suggesting that the distribution of neuropathology in neurosyphilis has no spatiotemporal characteristics [4,37]. The abundant neuropsychiatric symptoms reported in GPI may have been due to the wide range of brain damage. Thus, the different relationships between olfactory dysfunction and neuropsychiatric symptoms in GPI and other neuropsychiatric diseases may reflect their different patterns of brain damage.

In addition, the current study did not find any correlation between olfactory function and blood or CSF biomarkers and neuroimaging abnormalities. MTA before treatment was correlated with severe cognitive impairment and poor cognitive outcomes in patients with GPI [18]. Previous studies have reported that olfactory dysfunction is associated with changes in the volume and metabolism of the medial temporal cortex [38] and brain white matter lesions [39], but these associations were not observed in the present study. Since all patients with GPI in our study were hospitalized patients who received standardized treatment, abnormalities in CSF and blood and neuroimaging may be influenced by antibiotic treatment. To clarify the role of olfactory function in GPI, future studies need to examine the clinical manifestations and olfactory function differences between patients with GPI before and after treatment.

The present study provides evidence about olfactory dysfunction associated with GPI and proves that OI can serve as a marker for reflecting cognitive function in patients with GPI. In clinical practice, the OI test (which takes about 5 minutes) provides a noninvasive, easy-to-use, and cost-effective screening tool to evaluate the cognitive function in patients with GPI, thereby enhancing evaluation efficiency and assisting in early diagnosis. Some limitations of this work must be considered. First, our observations were based on cross-sectional data, and further longitudinal studies are needed to confirm these findings. In addition, future research should employ comprehensive neuropsychological assessments and multimodal MRI analyses to further explore the relationships between olfactory function and specific cognitive domains and between olfactory function and brain structural abnormalities in GPI.

Conclusions

The present study confirmed the presence of olfactory dysfunction in patients with GPI and revealed the overall pattern of olfactory dysfunction including deficits in OT, OD, and OI. Additionally, OI may serve as a marker for cognitive function in patients with GPI. Future longitudinal studies with larger sample sizes are needed to confirm these results and investigate the intricate interplay between olfactory function and cognition, as well as the underlying mechanisms of GPI.