Correlation of Hearing Loss and Chronic Falling Among Patients With Dementia in 3 Memory-Care Communities

Dementia, caused by physical changes in the brain, is a term used to imply a loss of memory or other mental abilities which interfere with everyday life.¹ Among older adults with dementia, there is an increased risk to fall; patients with dementia fall twice as often as older adults without cognitive impairment and are at greater risk of injurious falls.² Falls can lead to severe complications such as fractures, concussions, lacerations, and an increased number of hospitalizations; falls can also contribute to comorbidities and aggravate dementia symptoms.

Persons with dementia are at greater risk of experiencing a fall as a result of the loss of motor function, cognition, and difficulty in judging aspects of their environment. These falls can cause injuries, lower patients’ self-esteem, and contribute to a further loss of cognition as a result of striking the head. Coupled with the increased number of hospital visits and the potential for removal of patients from a group-living environment, the consequences of falling among residents of assisted-living memory care facilities with dementia are multifaceted and improved methods of fall-prevention are needed. 

Lin and colleagues have shown that hearing loss leads to greater brain degeneration.³ Lin and Ferrucci, in another study, showed that hearing loss directly influences fall rates.⁴ The present observation study sought to investigate the relationship between hearing loss and falls in older adults with dementia, specifically nursing facility residents. By identifying those residents at a greater risk, measures may be implemented to curtail falling and therefore prevent associated complications. 

 Participants and Study Design

The study took place in 3 freestanding assisted-living memory-care facilities in Illinois. A total of 153 residents with dementia were studied, 51 of whom were men and 102 of whom were women. Participants’ ages ranged from 65 to 99 years. This was an institutional review board (IRB)-approved study, and, given the loss of decision-making capacity among the participants, consent was obtained from the power of attorney (POA) agent. 

Data was collected from March 1, 2016 through July 31, 2016. Each participant was studied for a total of 90 continuous days within this 4-month study period; the Mini-Mental State Examination (MMSE) was administered at the end of an individual’s 90-day period.⁵ The MMSE was administered in a private room; participants and their POA agent had the option to stop, postpone, or split the MMSE into multiple sessions. The MMSE was administered to participants at all facilities during the study period between 9am and 12 noon to avoid changes in behavior related to sundowning, which refers to the change in cognition that can occur later in the day in some persons with dementia.⁶ 

Of the 153 participants enrolled, 51 residents were from site A, 68 from site B, and 34 from site C. The number of each participant’s falls in the 3 months prior to the administration of the MMSE was recorded. Age, gender, and medical history (eg, Parkinson disease [PD], osteoporosis, hyperlipidemia, hypothyroidism, chronic obstructive pulmonary disease [COPD], hypertension, other chronic diseases) were gathered from each participant’s medical record. Additionally, the number of medications taken by each participant and their use of antibiotics also were analyzed.

Because of participants’ dementia and associated decreased cognition, the social history was obtained through a questionnaire and conversation with each participant’s POA agent. The social history characteristics considered were tobacco use, military service, prior participation in athletics, and current use of glasses and/or hearing aids.

Data Analysis

De-identified, encrypted data were compiled in a spreadsheet, and strict IRB guidelines were followed. Pearson correlation coefficients were calculated between the number of falls for each individual and an array of variables, including PD, osteoporosis, hyperlipidemia, hypothyroidism, COPD, hypertension, tobacco use, military service, prior participation in athletics, current use of glasses, and current use of hearing aids. Correlations between falling and antibiotics use (Pearson r = 0.33) elevated hemoglobin A1c level (Pearson r = 0.22) and gait disorder (Pearson r = 0.20) were found; however, this study focuses on hearing impairment, since it proved to be the most statistically significant variable with respect to the occurrence of falls.

The raw data were split into 8 groups. Group 1 contained all data of participants who had not fallen in the 3-month period, group 2 contained data of participants who had fallen once, group 3 contained data of participants who had fallen twice, and so on, ending with group 8, which contained the data of only those participants who had fallen 7 or more times. For each group, correlation coefficients were calculated between the number of falls and the variables listed above. The raw data for each participant’s hearing loss and falls is shown in Table 1 (refer to page 2). We averaged number of falls between participants having and not having a certain trait and subtracted these values to investigate the effects of a variable on falls. The data were organized in order of having or not having a trait. For instance, in the case of hearing, all participants who had an indication of hearing loss, denoted by a 1 on the spreadsheet, were placed at the top of the spreadsheet, with those who did not have an indication of hearing loss, denoted by a 0 on the spreadsheet, coming at the bottom. The average number of falls for those labeled 1 (ie, having hearing loss) was calculated. This was repeated for all participants with a 0 for hearing loss (the lack of that trait). The difference in these averages became the statistic attributed to the condition’s effect on the number of falls for that specific participant. This process was repeated for the entire array of variables.

 The total number of participants who had an indication of hearing loss was 47, while the number of participants without an indication of hearing loss was 106, for a total of 153 participants.

Participants having an indication of hearing loss fell chronically (5 or more times) in greater numbers than did those without an indication of hearing loss. Figure 1 shows that participants without an indication of hearing loss (red line) tended to fall relatively fewer times, while participants with hearing loss tended to experience a much greater relative number of falls in the 3-month period. The red line represents participants with no indication of hearing loss, while the blue line represents participants with an indication of hearing loss. For example, 18 total participants fell 2 times, 12 of whom did not have an indication of hearing loss and 6 of whom had an indication of hearing loss.

Assuming that hearing loss has no effect on the number of falls, the group without an indication of hearing loss (red line) should always be graphically above the group with an indication of hearing loss (blue line), because the hearing-loss group had 59 more participants. However, the data show that the hearing-loss group surpassed the other group at 3 falls and again at approximately 8 falls, therefore showing that hearing loss seems to affect number of falls.

Of participants without an indication of hearing loss, 89.6% (95/106) fell 3 or fewer times, and 10.4% (11/106) fell more than 3 times. Of the participants with an indication of hearing loss, 76.7% (36/47) fell 3 or fewer times, and 23.3% (11/47) fell more than 3 times. Furthermore, only 4.7% (5/106) of participants without an indication of hearing loss fell 5 or more times. However, 14.9% (7/47) of participants with an indication of hearing loss fell 5 or more times. None (0/106) of the participants without an indication of hearing loss fell 8 or more times, while 10.6% (5/47) of participants with an indication of hearing loss fell 8 or more times.

Two noteworthy results emerged at the completion of data analysis. First, as the number of falls increased, the correlation between hearing loss and the number of falls grew stronger (ie, the r value increased) (Table 2). Second, the relation between the number of falls and hearing loss is determined by calculating the difference between the average number of falls for participants with a hearing loss indication and for those without (Table 3).

The average number of falls for a participant with a hearing loss indication was 2.575. The average number of falls for a participant without a hearing loss indication was 1.377. It is postulated that the difference of 1.198 falls is due to hearing loss.

Hearing loss puts a resident with dementia at a greater risk of falls. This becomes especially important when dealing with chronic fallers, defined as those falling 5 or more times in a 3-month period. The results presented add support to the position that hearing loss is correlated with increased numbers of falls in older adults with dementia. 

It is important to note that hearing loss was assigned by an interview with the patients’ POA agent in this study. If the resident used a hearing aid, according to the POA agent, the resident was considered to have an indication of hearing loss. Therefore, every participant who was marked 1, with an indication of hearing loss, was already being treated for hearing loss given that they had a hearing aid. However, persons with dementia often forget to use their hearing devices. Because hearing treatment often is neglected, participants with hearing loss may be at a greater risk of falling. In the study, hearing loss was only assigned as a 1 or a 0 by the POA agent; it is unlikely that a majority of participants from an older adult population did not have any hearing loss. The study was limited by being unable to show hearing loss for any participants not currently using a hearing aid.^7-9

According to the Hearing Loss Association of America, hearing loss may increase the risk of cognitive problems and even dementia.¹⁰ Lin and colleagues published a study in 2014 suggesting that persons with hearing loss prior to brain atrophy exhibit a loss of brain matter at a much higher rate compared with persons who exhibit brain atrophy without hearing loss.³ This study provides evidence that a correlation exists between the decay of a resident’s hearing and the number of falls the resident experiences. Lin’s research confirms that hearing loss in patients with dementia is furthering the rate of brain atrophy, creating more symptoms of noncognition and confusion.^3,4,11

It is known that brain tissue loss is more severe among persons with hearing loss.¹¹ This particularly severe form of dementia provides a unique perspective on the link between hearing loss and chronic falls. The results of the current study support a link between hearing loss and chronic falls in a population with dementia. When only including participants who fell less than 5 times in the studied period (nonchronic fallers), no correlation existed with hearing loss. However, the correlation rises dramatically when considering participants who experience chronic falls (ie, individuals who fell 5 or more times in the 3-month study period).

The correlation between hearing loss indication and falls for participants who fell fewer than 5 times could be related to other causes such as medication use, poor vision, confusion, agitation, paranoia, and other dementia symptoms. However, it is postulated that hearing loss strongly contributes to chronic falling; in particular, the Pearson value of 0.661 provides statistically significant support. This finding is further corroborated by the data showing the difference between the average falls of the participants with hearing loss and the average falls of the participants without hearing loss to be 1.198 falls.

It is important to determine the types and causes of hearing loss that may contribute to falls. Persons with hearing loss related to the following causes should be studied: GJB2 mutation, illness, ototoxic drugs, excessive noise, tumor, and head injury. By studying a larger number of people in different clinical environments, future studies may categorize persons with hearing loss into multiple groups. By studying persons in these hearing-loss categories, future research may further identify with greater specificity those individuals who are at an increased risk of chronic falls. After identifying individuals with dementia and hearing loss who are at risk of chronic falls, methods such as hearing aids, cochlear implants, increased assistance, amplifiers, visual aids, periodic ear examinations, and decreased use of ototoxic drugs may be implemented to prevent falls.

This study has certain limitations. First, the number of participants (N = 153) is large enough to provide statistical evidence supporting this correlation but not sufficiently large to affirm a specific theory. In future studies, a larger number of participants would provide greater variability of patient medical history. Moreover, in future studies it will be important to quantify the severity of hearing loss through a formal audiology examination; a χ2 test of association could be performed with both variables (falls and hearing loss). Furthermore, it would be interesting to study the correlation between GJB2-related hearing loss and dementia, as both have a genetic disposition. However, it is also conceivable that the connection between hearing loss and dementia/falls is more complex. For example, when patients with dementia have hearing loss, they often feel alone, alienated, and agitated.¹² Often, hearing loss leads a person with dementia to think that he or she is being misunderstood.¹²

In summary, the present study lends support to prior studies that found a correlation between hearing loss and falls in older adults with dementia. By identifying older adults with dementia with a history of chronic falls who also have hearing loss, interventions may be found to improve hearing quality so as to reduce the number of falls.

Authors: Ethan Schonfeld¹; Liz Tusler Meyer, AuD²; Andrew Becker, MS³; Katrina Tate, MS³; Meghana Moodabagil, MS³; Charlotte McSharry, MS³; Mahta Amidi, MS³; Andrew Bestler^3; Ricardo G Senno, MD, MS, FAAPMR³

Affiliations:
¹ Northwestern University, Evanston, IL
² Audiology Associates of Deerfield, Deerfield, IL
³ Chicago Medical School, Rosalind Franklin University of Medicine and Science, Chicago, IL

Disclosures:
The authors report no relevant financial relationships.

Address correspondence to:
Ethan Schonfeld
2720 Canterbury Drive
Phone: (224) 735-1990 Fax: (847) 272-8221
Email: ethanschonfeld@gmail.com