Minor Hallucinations in Alzheimer’s Disease

INTRODUCTION

Behavioral and neuropsychiatric symptoms develop during the middle-stages of Alzheimer’s disease (AD) in a high percentage of patients. They include mood disturbances, psychosis (delusions and hallucinations), personality changes, and behavioral disorders.

The prevalence of delusions and hallucinations in patients with AD are approximately 23% and 8.8%, respectively. About 18.9% of AD patients develop both symptoms. Delusions and hallucinations are both associated with higher dementia severity, poorer cognition and function, greater severity in other neuropsychiatric symptoms, and higher caregiver burden. It has been suggested that the presence of delusions, alone or in combination with hallucinations, may help to predict institutionalization [1].

The hallucinatory phenomenon, especially well-structured visual experiences, also appears in other neurodegenerative disorders, such as dementia with Lewy Bodies [2] or Parkinson’s disease (PD) [3]. However, hallucinations may have a broader spectrum and include so-called minor hallucinations (MHs). MHs include passage hallucinations (PHs), visual illusions, and presence hallucinations (PrHs) [4, 5].

PHs consisted of brief visions of a person or an animal passing from the periphery of the visual field and they are generally identified as a person or an animal of a particular species (commonly a cat or a dog). These characteristics suggest that disinhibition of an early part of the visual cognitive process, such as perceptual representation system can lead to the misinterpretation of a very flimsy perception and the emergence of PHs [6]. Visual illusions are transitory misperceptions of living beings or objects that are different from objective reality. A PrH, or sense of presence, is an experience in which the person has a strong feeling of the presence of someone (person or animal), either somewhere in the room or, less often, behind him or her, without any apparent justifying sensory stimulus. Frequently, the presence is that of a relative. PH and illusion are visual phenomena, but PrH cannot be related to a specific sensory modality. PrHs might arise from a conflict between a preconscious visual perception (hallucinated) and actual normal vision [6].

Recent studies have shown that MHs seem to be the most frequent type of hallucinatory symptoms in PD; they can occur in early stages of the disease, even before the recognition of motor symptoms, and may herald the progression to more severe psychotic experiences [7, 8]. They have also been described in Lewy Body dementia [9] and in healthy older people [10]. However, their prevalence in AD patients has never been evaluated.

The first goal of the present study was to determine the prevalence of minor hallucinatory phenomena in AD, as there is no data available in this regard. Second, to describe the characteristics of this minor hallucinatory phenomenon in patients with AD and amnestic mild cognitive impairment (aMCI). Third, to describe the potential relationship of this phenomenon with cognition, the presence of other psychological and behavioral symptoms and the use of psychoactive drugs in patients with mild and moderate AD.

METHODS

We have recruited prospectively and consecutively 268 subjects from a sample of outpatients assisted to the Cognitive Disorders Unit at Hospital Universitari de Santa Maria in Lleida between April 2016 and August 2017. Ninety acetylcholinesterase inhibitors drug-naïve patients recently diagnosed as AD in mild to moderate stages (Mini-Mental State Examination (MMSE)>20), seventy-eight patients with aMCI and one hundred controls without cognitive impairment. Volunteers, relatives, or friends of the patients were enrolled as controls. The diagnosis of AD and MCI was made according to the clinical criteria of National Institute of Aging-Alzheimer's Association (NIA-AA) [11, 12]. Based on these criteria, MCI refers to an state in which there should be evidence of concern about a change in cognition in comparison with the person’s previous level, evidence of lower performance in one or more cognitive domains that is greater than would be expected for the patient’s age and educational background and finally the patient must maintain the functionality preserved for activities of daily living. Only aMCI patients were included in the study. Since visual and auditory hallucinations with no clinical significance have also been reported in the healthy population, the AD patients were matched with cognitively healthy subjects by age, sex, and education. All healthy controls and patients had normal or corrected-to-normal vision and hearing. All participants provided written informed consent.

Excluding criteria were as follows: 1) patients with cognitive impairment caused by other entities, such as the presence of lesions identified on CT or brain MRI or analytical alterations (ionic alterations, hypothyroidism, deficit of cobalamin or folic acid, positive syphilitic serology) that could justify the presence of cognitive impairment; 2) Patients with AD and MMSE<20 because we considered that patients with moderate-severe or severe cognitive impairment could not adequately respond to the semi-structured interview; 3) Control subjects who showed deficits in cognitive screening tests during the interview (MMSE≤26).

All participants underwent cognitive and behavioral assessment through MMSE [13] and Neuropsychiatric Inventory (NPI) [14] respectively. The NPI appraises the severity and frequency of the following 12 domains of behavioral functioning: delusions, hallucinations, agitation/aggression, dysphoria, anxiety, euphoria, apathy, disinhibition, irritability/lability, aberrant motor activity, night-time behavioral disturbances, and appetite and eating abnormalities. The total value is obtained by multiplying the frequency (0 = absent, 1 = occasionally, less than 1 per week, 2 = often, about 1 time per week, 3 = frequently, several times per week but not daily, 4 = very frequently, daily or continuously) by severity (1 = mild, 2 = moderate, 3 = severe) of each of these items based on screening questions that usually are asked from the caregiver. Total score is the sum of the subscale scores.

Both the subjects and their relatives responded to a semi-structured questionnaire in order to rate psychotic phenomena. For this purpose, we made use of the sample interview in Pagonabarraga’s et al. study about Minor hallucinations in Parkinson’s Disease drug-naive patients [7]. The questionnaire includes questions on hallucinations in different modalities (visual, olfactory, tactile, auditory, PrHs, illusions, and PHs) that had occurred in the previous month. It also obtains some other details, such as when they started, how they developed, their frequency, intensity, duration, and time of the last episode.

General medical aspects such as age, gender, schooling, smoking, drinking, family history of AD, or other neurological pathology, personal pathological antecedents such as hypertension, diabetes mellitus, hypercholesterolemia, stroke, depression, psychiatric history, and psychoactive drugs consumption were also evaluated in the whole sample.

Clinical, cognitive, and demographic data from AD patients were benchmarked against results obtained from healthy controls. The statistical study was performed by the SPSS 19.0 program (SPSS, Chicago, IL). The comparison of means was made by the Student t and ANOVA tests. Bonferroni post hoc analysis was performed to evaluate positive ANOVA test. The comparison between categorical variables was performed using the chi test square Pearson and Fisher’s exact test. Spearman’s correlation was used for quantitative variables. Significance was set at p < 0.05 for all analyses.

RESULTS

We collected a consecutive series of 268 subjects, 90 with recent diagnosis of mild-moderate AD, 78 with aMCI and 100 controls. These subjects were aged between 56 and 86 years with 53.0% of them being women. Patients with aMCI were younger than the controls and AD (p < 0.01). The groups did not differ in terms of smoking, excessive alcohol consumption, family history of AD, or personal background. Obviously, MMSE score was lower in AD patients than aMCI patients than controls (p < 0.01 for comparisons between all groups). Patients with AD showed greater presence of neuropsychiatric symptoms according to the NPI scale than controls (p < 0.01) but not compared to MCI patients (p = 0.06). Controls had a greater NPI total score than MCI patients (p < 0.01). After Bonferroni analysis the differences remained significant for depression, anxiety, apathy, and irritability (p < 0.05).

Sociodemographic characteristics and the prevalence of antidepressive, hypnotic/anxyolitic, and antipsychotic drugs consumption are shown in Table 1. No patient was admitted to a nursing home.

Prevalence of MHs was 21.1% (19) in AD, 12.8% (10) in aMCI, and 3% (3) in controls (p < 0.01). The most frequent MH was PrH (9.3%), followed by PH (4.9%), and finally illusion (0.7%). For AD subgroup, the frequencies of PrHs, PHs and illusions were 16.7%, 10.0%, and 0% respectively; and for aMCI subgroup were 9%, 5.1%, and 2.6% respectively.

In the group of patients with MHs, eight (27.8%) patients had more than one MH and the most frequent association was observed between PHs and PrHs. In 13 (44.8%) patients MHs (one or more than one) were associated with other hallucinatory modalities (one or more than one). The association was observed with tactile (6 patients), olfactory (6 patients), auditory (3 patients), and finally visual and sensory hallucinations (2 patients). No gustatory hallucinations were registered.

The description of MHs was very variable. The most frequent PrHs were described as follows: the patient noticed someone passing behind him in the street, felt the presence at home or at the side of the bed, or even a patient had the feeling that there was someone in the street every night and, although he knew it was a feeling that did not correspond to reality, before going to bed he would look down at the sidewalk (no delusional idea of control or persecution was present).

The PHs included unstructured visions such as sideways passage of an indefinable object or well-defined objects such as a black ball, a leaf of a tree, flies, or a bird. The frequency of these episodes was very erratic, some patients had 4-5 MHs a week and some once a month or less. We did not find relationship between the frequency or the type of MHs and the presence or type of main hallucinations.

In healthy controls, 3 subjects presented MH (all PrH). All cases experienced the presence of someone at home; in two of them this feeling was related to the loss of one of the family members.

We emphasize that among these three cases only one of the patients expressed the experience of MH spontaneously when he was asked in general for hallucinations, while the rest of controls and patients only mentioned them after completing the questionnaire. This highlights the importance of asking specifically for this type of hallucinations.

After adjusting age and gender, there was a negative correlation between MMSE score and the presence of MHs (r = –0.261; p < 0.01) while there was a positive correlation between NPI score and the presence of MHs (r = 0.237; p < 0.01).

In AD patients, when we compared those with MH and those without, we found that they did not show significant differences at age and gender, clinical features or cognitive function according to the scales used in our study. Globally, the total score for MMSE and total NPI was similar in both groups (p = 0.5 and p = 0.3, respectively). However, they showed significant differences in the hallucinations, depression/dysphoria, and sleep disorders items of NPI (see Table 2).

In some studies, the presence of MHs has been related to the frequency of psychoactive drugs consumption. In our study, no significant correlation was observed between the use of antidepressant (30% versus 50%; p = 0.19), hypnotics (19.0% versus 22.2%; p = 0.76), or neuroleptics drugs (4.8% versus 0%; p = 0.34) and the presence or absence of MHs in patients with AD.

In aMCI patients, we found that those patients with MH had lower MMSE score (24.5±0.8 versus 26.1±0.2; p = 0.02) and higher prevalence of depression (60.0% versus 20.6%; p < 0.01) compared with subjects with aMCI without MH. We did not find any other difference at the demographic or clinical characteristics except some items of NPI scale, such as the presence of delirium, night-time behavioral disturbances, and appetite/eating disorders (Table 2). In this subgroup, we did not either observed any significant correlation between the use of antidepressant (25.4% versus 40.0%; p = 0.33), hypnotic (29.9% versus 30.0%; p = 0.99), or neuroleptic (6.0% versus 10.0%; p = 0.63) drugs and the presence or absence of MHs.

DISCUSSION

We found that 21.1% of patients with AD and 12.8% of patients with MCI have experienced MH. To our knowledge, this is the first report describing MHs in newly diagnosed drug-naïve AD and aMCI patients, therefore these findings cannot be compared with the previous data.

There are studies that have evaluated the prevalence of psychotic symptoms and MHs in elderly healthy population or in other degenerative pathologies such as PD or Lewy body dementia. In a study of elderly population, the prevalence of psychotic symptoms was between 4–24% [15, 16], and this prevalence increased especially in institutionalized subjects or those with sensory deficits [17]. In our sample population, both groups of patients were excluded.

Referring specifically to MHs, Soulas et al. [10] observed that up to 13% of elderly subjects had MHs and their presence increased with age and use of psychoactive drugs. Furthermore, the presence of major and minor hallucinations together was reported very frequent. As in our study, they observed no case of gustatory hallucinations, demonstrating its low frequency in this type of population.

In a similar study designf on consecutive non-demented, newly presented and un-medicated patients, Pagonabarraga et al. reported a 42% prevalence of MHs in 50 patients with PD [7]. Similarly, in the study conducted by Nishio et al. the prevalence of MHs was reported 44.4% in patients with PD [8]. However, a more previous and larger study evaluated the prevalence of MHs in 216 consecutive patients, most taking PD medications. They reported that MHs affected approximately 26% of the study population [6]. In conclusion, it seems that the prevalence of MHs in PD is greater than AD and aMCI patients.

In both AD and MCI patients, MHs were associated with an increase in the frequency of depression. This data has been also observed in PD patients [7, 8] and in psychiatric population. The association between psychotic symptoms and depression is well known now [18]. According to Nishio et al., it is possible that depression and other types of negative moods play an important role in the process by which MHs/illusions evolve into frank visual hallucinations through non-neural and psychological factors. The other possibility is that depression is a psychological reaction to, or consequence of, visual hallucinations and delusions [8].

It is important to highlight that the prevalence of MHs in the control group in the Pagonabarraga study [7] and our study was very similar (4.5% compared to 3%, respectively), suggesting that this prevalence might be consistent in Spanish people without cognitive impairment. The most frequent type of hallucination observed in the controls was PrH. Hallucinations involving the deceased spouse have been reported in up to half of widowed persons, with a higher frequency in the elderly [19, 20]. The incidence of hallucination increases with the duration of marriage and is particularly associated with a happy marriage and parenthood [21].

The physiopathological process of MHs is not known yet. Several studies in patients with PD found a bilateral reduction of gray matter volume in different areas of the dorsal visual stream and in functionally related midbrain/thalamus posterior neocortex and cerebellar structures [7, 22]. However, no similar study in AD patients exists.

Given that our study is a cross-sectional study, we do not know what consequences could MHs have on the progression and prognosis of AD. Furthermore, considering that the major hallucinations have shown to progress in frequency and severity [23] and are related to higher mortality [24], we ignore the impact of MHs on subsequent appearance of major hallucinations or mortality. Our patients did not receive acetylcholinesterase inhibitors and we ignore the effect of these inhibitors on the prevalence of MHs.

None of the patients had extrapyramidal symptoms suggesting that patients may not have other types of dementia. However, we cannot rule out the development of other types of dementia, such as Lewy body dementia, during the progression of the disease. However, we believe that the high prevalence of MHs found in AD emphasizes that this prevalence cannot be explained by the presence of patients who may end up developing extrapyramidal symptoms. Similarly, among the control subjects with MH, we cannot rule out the presence of asymptomatic α-synuclein deposits.

One of the main limitations of our study is that we have only been able to evaluate the presence of MHs in patients in the mild-moderate phase of the disease. We believe that in more advanced stages of the disease the cognitive deterioration of the patient would limit the validity of the results. Another aspect to consider is that the evaluations were not blinded, so the effect of suggested answers cannot be excluded. Patients and caregivers may be more likely to give more importance to perceptions or unusual events than healthy subjects, and researchers may be more likely to find hallucinations in the AD group. In addition, given that the recruitment of the sample was carried out in a unit of cognitive disorders through a structured non-validated interview, the generalization of the high prevalence of MHs in patients with AD free of treatment requires replication in community patients.

Finally, in patients with MCI, follow-up is necessary to evaluate the risk of progression of these patients to AD or other dementias. It would also be interesting to assess the possible association between the presence of MHs and the presence of AD biomarkers using cerebrospinal fluid and/or PET, in order to assess whether MHs are really a clinical finding that can be considered characteristic of preclinical stages of the disease such as anosmia or constipation in PD.

In conclusion, our study has shown the presence of MHs in patients with newly diagnosed AD, and aMCI, with a frequency much higher than healthy controls. This prevalence is correlated with the presence of other cognitive behavioral symptoms and with a worse cognitive performance. Therefore, we suggest the interrogation about the presence of this type of hallucinations as a clinical feature of this population.