Short Physical Performance Battery and Mediation of the Effect of Mild Cognitive Impairment on Falls by Community-Dwelling Older Adults
Publication: The Journal of Neuropsychiatry and Clinical Neurosciences
Abstract
Objective:
The authors examined the association among cognitive function, falling, and physical performance among community-dwelling older adults (ages ≥65 years).
Methods:
Eight waves of the National Health and Aging Trends Study (NHATS; 2011–2018) were assessed, with 1,225 respondents who participated in all waves. The outcomes were self-reported number of falls and NHATS Short Physical Performance Battery (SPPB) score. The Clock Drawing Test measured participants’ executive function, and immediate and delayed word recall tests assessed memory.
Results:
The analyses indicated no direct correlation between executive function and fall risk when controlled for contributing factors. However, executive function and memory significantly predicted the risk for poor physical performance, defined by the NHATS SPPB score. The interaction between pain medication and memory worsened poor physical performance among participants with mild and severe memory impairment, as well as among those with mild to moderate impairment in executive function.
Conclusions:
Screening older adults living in the community for executive function, memory impairment, and physical performance can predict the risk for falls and the subsequent consequences of falling.
Many older adults experience a fall annually, mostly due to, but not limited to, poor balance, physical performance problems, cognitive impairment, sleep disturbance, and vision impairment (1). The literature has shown that executive function and episodic memory reserve are significantly correlated with physical performance, in particular slow gait speed, which therefore predicts an increased risk for falling (2, 3). Impaired executive function may be associated with a lower level of complex motor tasks, which may also increase fall risk by decreasing the ability to maintain balance and to integrate cognitive skills related to attention and perception in motor planning (4). Some researchers have reported that memory reserve does not predict falling (5). In contrast, other investigators have reported that immediate memory impairment could be associated with increased fall risk, especially recurrent falls among older adults, mediated by the effects of activity, mobility, and grip strength (6). Conversely, older adults with a history of recurrent falls tend to score significantly lower on all cognitive tests (6). This counterintuitive relationship between cognitive reserve, physical function, and falling has been shown to be equivocal in the literature and requires further investigation (2). As an indicator of physical performance, the Short Physical Performance Battery (SPPB) tests gait speed, timed chair stand, and balance (7). To our knowledge, the associations between executive function and memory, SPPB, and falls have not been investigated in longitudinal studies with large sample sizes.
Mild cognitive impairment can be considered a form of cognitive decline in one or more domains of executive function and memory that is atypical of normal aging and may be a precursor to dementia (8). To our knowledge, whether the effect of mild cognitive impairment on physical function can increase the risk for falling has not been investigated in longitudinal studies.
Sleep and pain medications should be considered when examining falls among older adults because of their effects on balance and their association with potential risk for falling. Sleep medications can cause confusion, delirium, and falls by increasing involuntary movements, which can then lead to a loss of balance (9). In addition, nonsteroidal anti-inflammatory drugs (NSAIDs), the first-line treatment to control chronic pain among older adults, can significantly increase the risk for falling in inpatient settings (10, 11). However, information about the role of these medications in falls among community-dwelling older adults is limited.
In this study, we examined the relationship between executive function, memory, physical function, and falling in a longitudinal model, while controlling for the use of sleep and pain medications, fear of falling, anxiety, visual impairment, and depression among community-dwelling older adults with and without mild cognitive impairment. We hypothesized that decreased cognitive function would correlate with increased frequency of falls mediated by physical function.
Methods
Study Data
We used 8 years (2011–2018) of data from the National Health and Aging Trends Study (NHATS), which comprises a publicly available data set funded by the National Institute on Aging. In NHATS, which serves as a platform to study the trend of functional limitations among older adults, annual in-person interviews were conducted with a nationally representative sample of older adults (≥65 years) who were Medicare beneficiaries living in the community and in residential care centers (12). To develop nationally representative data sets, selection of a group of counties was conducted in the first stage, followed by selection of zip codes in the second stage. In 2011 (the first wave), participants were selected on the basis of demographic variables (i.e., age and ethnicity) within each zip code (13). NHATS has been reviewed by and received approval from the Johns Hopkins Bloomberg School of Public Health’s Institutional Review Board. The results of this research can be generalized to the older adult population, with the possibility of underrepresentation of some groups (e.g., immigrants who are not eligible for Medicare), for the following reasons: the population of older adults in the United States was 41.4 million in 2011; 38.9 million people enrolled in Medicare in 2011 (14, 15); and 8,245 older adults participated in NHATS in 2011, with a 71% weighted response rate (16). The trained interviewers conducted the interviews and assessments under the supervision of a field supervisor in the location where the participants lived. Further details about data collection tools and methods are described elsewhere (17).
Data Sets
In this study, we used the deidentified data provided by NHATS. We requested access to sensitive data (age, gender, income, and ethnicity) regarding the study participants. After the request was reviewed, we received the deidentified sensitive data. By using the sample person ID (SPID) numbers, which are not related to the identity of the participants and are unique for each participant across 8 years, we merged the two data sets for each wave. After selection of the dependent, independent, and control variables, on the basis of the hypothesis described below, new variables were created in all eight waves of data. For each SPID and year, we created a global variable for each of the dependent, independent, and control variables. Participants were excluded from the study if they were diagnosed with dementia or moderate to severe or severe impairment in executive function. Only data from those living in the community remained in the data set. Year of participation (2011–2018) was considered as the time variable in the data analyses. By using SPID and year, the eight files were appended first, then balanced (the same respondents participated in eight waves [N=9,800 or N=1,225 per wave], representing a balanced data set). Participants ages ≥95 years were excluded from the regression models.
Variables
In two separate questions, participants were asked whether they had any fall events, defined as losing balance and landing on the floor or ground in the past month and in the past 12 months (17). The NHATS SPPB score was used as an indicator of physical performance to evaluate balance stance (side-by-side, semitandem, and tandem), gait speed, and chair stand (single and repeated). Participants who did not attempt the exercises or were ineligible (because of safety concerns) received a score of zero. The NHATS SPPB score range is 0–12, after summing the scores of balance stand, gait speed, and chair stand time (i.e., score range for each measure was 0–4) (18).
Mild cognitive impairment is defined as a defect in one or more of the main components of cognitive function (i.e., memory, language, attention, visuospatial function, and executive function). Executive function was measured with the Clock Drawing Test (CDT), with scores categorized in six levels: 0–5, with a score of 0 indicating no executive function impairment and a score of 5 indicating very severe impairment (19). Participants drew a given time (e.g., 11:10) on a piece of paper.
Immediate memory and delayed memory were tested with the Immediate Word Recall Test (IWRT) and the Delayed Word Recall Test (DWRT), respectively, which have 100% sensitivity and specificity for screening for memory impairment among older adults (13, 20). For the IWRT, which was given before the CDT, participants listened to a list of 10 words and were asked to repeat each word immediately. For the DWRT, which was given after completion of the CDT, participants were asked to repeat the same words. On the basis of the number of words remembered, memory was categorized into three ordinal levels: normal (seven or more words), mild impairment (five to six words), and severe (less than five words) (21).
The function score was developed with the model by Gill and Williams (22), which surveys a person’s ability to go outside, get around inside, get out of bed, eat, bathe, use the toilet, and dress. Performance of each activity was given a score of 1, 2, or 4 on the basis of whether the participant was independent, was vulnerable, or needed assistance, respectively. Hence, the maximum total score for each function was 28, indicating “most dependent,” and the minimum score was 7, indicating “most independent.”
The Patient Health Questionnaire–2 and the Generalized Anxiety Disorder 2-item questionnaire were used to screen for depression and anxiety, respectively (23, 24). Participants were asked whether they used pain or sleep medication; their responses pertaining to the use of each type of medication were collapsed into two categories: rarely and 2–7 days per week. The history of falling in the past month and past year, at the time of the study, and fear of falls were self-reported.
Analysis
We used longitudinal regression analyses for different dependent variables (i.e., falling events and NHATS SPPB score), using STATA, version 16.0 (StataCorp, College Station, Tex.). Considering the research hypothesis, which was to examine any direct correlation between the risk for falling and executive function and memory, we developed a longitudinal logistic regression model in which executive function, memory, and NHATS SPPB score (as the indicator of physical function) were predictors of falling (yes or no). On the basis of the results of the first regression model (i.e., no significant correlation between cognitive domains and falling), we developed the second, third, and fourth regression models to examine the correlation between NHATS SPPB score and its cognitive function predictors (i.e., executive function, delayed memory, and immediate memory, respectively). The NHATS SPPB score was treated as a continuous variable (0–12), and longitudinal linear regressions (random effect) were used to examine its correlation with the predictors.
Results
The mean age of participants was 74.2 years (SD=6.0) in 2011, and 57.7% were women. The frequency of falls within the past month and past 12 months significantly increased within the 8 years of the study, from 7.0% and 19.5% in 2011 to 10.8% and 24.2% in 2018, respectively. For multiple fall events, the rate increased from 36.8% in 2011 to 41.8% in 2018. Fear of falling increased >15.0 percentage points, from 22.4% to 37.5%. The average NHATS SPPB score decreased from 7.9 in 2011 to 6.8 in 2018. The average function score increased from 7.7 to 8.6 within the 8 years (Table 1).
| Variable | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | % | N | % | N | % | N | % | N | % | N | % | N | % | N | % | |
| Fall | ||||||||||||||||
| Past month | 86 | 7.0 | 104 | 8.5 | 104 | 8.5 | 119 | 9.8 | 132 | 10.8 | 111 | 9.1 | 129 | 10.6 | 131 | 10.8 |
| Past 12 months | 222 | 19.5 | 249 | 22.3 | 266 | 23.8 | 256 | 23.3 | 260 | 24.0 | 249 | 22.6 | 279 | 25.8 | 259 | 24.2 |
| Multiple falls | 113 | 36.8 | 125 | 35.4 | 132 | 35.9 | 149 | 39.7 | 142 | 36.5 | 141 | 39.2 | 163 | 40.0 | 161 | 41.8 |
| Fear of falling | 274 | 22.4 | 280 | 22.9 | 294 | 24.0 | 323 | 26.5 | 340 | 27.9 | 380 | 31.3 | 413 | 34.1 | 453 | 37.5 |
| Executive function (Clock Drawing Test)a | ||||||||||||||||
| Normal | 349 | 28.5 | 412 | 33.7 | 379 | 31.0 | 553 | 45.3 | 510 | 41.9 | 518 | 42.6 | 465 | 38.4 | 566 | 46.9 |
| Mild | 628 | 51.3 | 625 | 51.1 | 672 | 55.0 | 481 | 39.4 | 556 | 45.7 | 561 | 46.2 | 580 | 47.9 | 473 | 39.2 |
| Mild to moderate | 247 | 20.2 | 186 | 15.2 | 172 | 14.1 | 187 | 15.3 | 152 | 12.5 | 136 | 11.2 | 167 | 13.8 | 169 | 14.0 |
| Immediate memory (Immediate Word Recall Test) | ||||||||||||||||
| Normal | 276 | 22.6 | 342 | 28.0 | 318 | 26.0 | 315 | 25.8 | 297 | 24.4 | 278 | 22.9 | 241 | 19.9 | 269 | 22.3 |
| Mild | 666 | 54.4 | 607 | 49.6 | 636 | 52.0 | 623 | 51.0 | 616 | 50.6 | 612 | 50.4 | 606 | 50.0 | 557 | 46.1 |
| Severe | 282 | 23.0 | 274 | 22.4 | 269 | 22.0 | 283 | 23.2 | 305 | 25.0 | 325 | 26.8 | 365 | 30.1 | 382 | 31.6 |
| Delayed memory (Delayed Word Recall Test) | ||||||||||||||||
| Normal | 107 | 8.7 | 138 | 11.3 | 138 | 11.3 | 140 | 11.5 | 146 | 12.0 | 127 | 10.5 | 125 | 10.3 | 109 | 9.0 |
| Mild | 462 | 37.8 | 471 | 38.5 | 461 | 37.7 | 462 | 37.8 | 433 | 35.6 | 456 | 37.5 | 411 | 33.9 | 395 | 32.7 |
| Severe | 655 | 53.5 | 614 | 50.2 | 624 | 51.0 | 619 | 50.7 | 639 | 52.5 | 632 | 52.0 | 676 | 55.8 | 704 | 58.3 |
| Visual impairment | 47 | 3.8 | 48 | 3.9 | 45 | 3.7 | 49 | 4.0 | 56 | 4.6 | 50 | 4.1 | 58 | 4.8 | 67 | 5.6 |
| Sleep medication use | ||||||||||||||||
| Never/rarely | 983 | 86.1 | 978 | 86.3 | 968 | 84.5 | 961 | 83.9 | 972 | 84.7 | 972 | 84.8 | 956 | 85.1 | 948 | 84.1 |
| 2–7 days/week | 159 | 13.9 | 155 | 13.7 | 177 | 15.5 | 185 | 16.1 | 175 | 15.3 | 174 | 15.2 | 168 | 15.0 | 179 | 15.9 |
| Pain medication use | ||||||||||||||||
| Never/rarely | 774 | 63.2 | 749 | 61.3 | 734 | 60.1 | 731 | 59.9 | 738 | 60.6 | 742 | 61.2 | 747 | 61.6 | 713 | 59.3 |
| 2–7 days/week | 450 | 36.8 | 473 | 38.7 | 487 | 39.9 | 489 | 40.1 | 479 | 39.4 | 471 | 38.8 | 465 | 38.4 | 490 | 40.7 |
| Depression | 430 | 35.2 | 436 | 35.8 | 437 | 35.9 | 443 | 36.4 | 452 | 37.2 | 464 | 38.4 | 458 | 37.9 | 484 | 40.2 |
| Anxiety | 469 | 38.4 | 462 | 37.9 | 461 | 37.8 | 460 | 37.8 | 436 | 35.9 | 480 | 39.7 | 486 | 40.3 | 448 | 37.1 |
| Race/ethnicity | ||||||||||||||||
| White | 1,004 | 82.0 | ||||||||||||||
| Black | 162 | 13.2 | ||||||||||||||
| Hispanic | 40 | 3.3 | ||||||||||||||
| Other | 19 | 1.6 | ||||||||||||||
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| NHATS SPPB score | 7.92 | 2.74 | 7.93 | 2.74 | 7.92 | 2.87 | 7.73 | 3.01 | 7.48 | 2.95 | 7.43 | 2.98 | 7.23 | 3.03 | 6.82 | 3.21 |
| Function score | 7.73 | 1.93 | 7.75 | 1.96 | 7.81 | 2.07 | 7.96 | 2.42 | 7.93 | 2.16 | 8.09 | 2.43 | 8.32 | 2.82 | 8.62 | 3.28 |
| Age (years) | 74.17 | 5.93 | 75.16 | 5.90 | 76.16 | 5.90 | 77.13 | 5.85 | 78.08 | 5.79 | 79.04 | 5.74 | 80.00 | 5.70 | 80.96 | 5.63 |
a
Some percentages are based on denominators smaller than the total (N=1,225) because of missing values.
CDT results revealed that in 2011, 28.5% and 51.3% of participants had normal and mild executive function impairment, respectively. CDT results also revealed that approximately 57.0% of participants maintained their executive function status throughout the 8 years of the study, while 36.2% developed mild impairment in executive function, and 6.9% developed mild to moderate impairment. In contrast, 34.3% of participants with mild to moderate impairment and 32.3% with mild impairment in 2011 regained normal executive function during the 8 years of the study. DWRT results revealed that among participants with normal delayed recall, approximately 40.0% remained in the same category, and 45.0% and 15.4% developed mild and severe impairment, respectively, within 8 years. Approximately 72.0% of participants with severe impairment remained in the same category. IWRT results revealed that among participants within the normal range of immediate recall, 50.5% remained in the same category, but 43.6% and 6.0% developed mild and severe impairment in immediate recall, respectively, in 8 years. The IWRT results also revealed that about half of participants with severe impairment in immediate recall did not improve within the 8 years. The proportion of depression increased from 12.0% to 13.4% within the 8 years, when approximately 20.0% of participants received a new diagnosis of depression and one-third of depressed participants returned to a nondepressed state.
The results of longitudinal logistic regression analyses revealed that a one-unit increase in the NHATS SPPB score reduced the risk for falling within the past month by approximately 12.0%, when fear of falling increased this risk by 41.0%. Cognitive impairment (i.e., as measured by executive function and the IWRT and DWRT) did not significantly predict falling risk, although better memory lowered the risk for falling. Depression increased the falling risk by 33.0%. Women were less likely to experience falling (by 21.0%) compared with men. Overall, the risk for falling increased by 6.0% in each year throughout the 8 years of the study (Table 2).
| Predictor | Odds ratio | SE | 95% CI |
|---|---|---|---|
| CDT score | 0.99 | 0.06 | 0.88–1.12 |
| DWRT score | 1.02 | 0.08 | 0.87–1.19 |
| IWRT score | 1.06 | 0.08 | 0.91–1.23 |
| NHATS SPPB score | 0.88*** | 0.02 | 0.86–0.91 |
| Fear of falling (reference group: no fear of falling) | 1.41*** | 0.14 | 1.17–1.70 |
| Depression (reference group no depression) | 1.33** | 0.13 | 1.10–1.61 |
| Anxiety (reference group: no anxiety) | 1.13 | 0.11 | 0.94–1.36 |
| Age (years) | 0.98* | 0.01 | 0.96–0.99 |
| Race/ethnicity (reference group: White) | |||
| Hispanic | 0.75 | 0.21 | 0.43–1.31 |
| Black | 0.77 | 0.12 | 0.57–1.05 |
| Other | 0.95 | 0.37 | 0.45–2.03 |
| Women (reference group: men) | 0.79* | 0.08 | 0.64–0.97 |
| Year | 1.06** | 0.02 | 1.02–1.10 |
a
Longitudinal logistic regression was used. Summary statistics were as follows: log likelihood=−2,613.81, Wald χ2=140.77, df=13, p<0.001; N=9,036. CDT=Clock Drawing Test; DWRT=Delayed Word Recall Test; IWRT=Immediate Word Recall Test; NHATS=National Health and Aging Trends Study; SPPB=Short Physical Performance Battery.
*p<0.05, **p<0.01, ***p<0.001.
Mild to moderate impairment in executive function reduced the NHATS SPPB score by 0.15, compared with normal executive function. Pain medication, sleep medication, fear of falling, functional decline, and depression decreased this score by 0.21, 0.16, 0.34, 0.24, and 0.26, respectively. A 1-year increase in age reduced the chance of a higher NHATS SPPB score by 0.16. Hispanics and African Americans had lower probability of higher NHATS SPPB scores, by 1.61 and 1.32, respectively, compared with non-Hispanic Whites. Women had reduced NHATS SPPB scores (by 0.65) compared with men (Table 3; model 1).
| Model 1b | Model 2c | Model 3d | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Predictor | Coefficient | SE | 95% CI | Coefficient | SE | 95% CI | Coefficient | SE | 95% CI |
| CDT score | |||||||||
| Mild impairment | 0.04 | 0.04 | −0.05 to 0.12 | – | – | 0.04 | 0.04 | −0.05 to 0.12 | |
| Mild to moderate impairment | −0.15* | 0.06 | −0.28 to −0.03 | −0.15* | 0.06 | −0.28 to −0.03 | – | – | |
| Pain medication | −0.21*** | 0.05 | −0.30 to −0.12 | – | – | – | – | ||
| Mild executive function impairment-by-pain medication use | |||||||||
| No mild impairment plus 2–7 days/week | – | – | −0.20** | 0.06 | −0.32 to −0.08 | – | – | ||
| Mild impairment plus rarely | – | – | 0.04 | 0.05 | −0.06 to 0.14 | – | – | ||
| Mild impairment plus 2–7 days/week | – | – | −0.18** | 0.07 | −0.30 to −0.05 | – | – | ||
| Mild to moderate executive function impairment-by-pain medication use | |||||||||
| No mild to moderate impairment plus 2–7 days/week | – | – | – | – | −0.23*** | 0.05 | −0.33 to −0.13 | ||
| Mild to moderate impairment plus rarely | – | – | – | – | −0.21** | 0.08 | −0.36 to −0.06 | ||
| Mild to moderate impairment plus 2–7 days/week | – | – | – | – | −0.29** | 0.09 | −0.48 to −0.11 | ||
| Sleep medication | −0.16* | 0.08 | −0.31 to −0.01 | −0.16* | 0.08 | −0.31 to −0.01 | −0.16* | 0.08 | −0.31 to −0.01 |
| Fear of falling | −0.34*** | 0.05 | −0.45 to −0.24 | −0.34*** | 0.05 | −0.45 to −0.24 | −0.35*** | 0.05 | −0.45 to −0.24 |
| Function | −0.24*** | 0.01 | −0.26 to −0.22 | −0.24*** | 0.01 | −0.26 to −0.22 | −0.24*** | 0.01 | −0.26 to −0.22 |
| No visual impairment | 0.32** | 0.11 | 0.11 to 0.53 | 0.32** | 0.11 | 0.11 to 0.53 | 0.32** | 0.11 | 0.11 to 0.53 |
| Depression | −0.26*** | 0.05 | −0.36 to −0.16 | −0.26*** | 0.05 | −0.36 to −0.16 | −0.26*** | 0.05 | −0.36 to −0.17 |
| Anxiety | 0.01 | 0.05 | −0.09 to 0.10 | 0.01 | 0.05 | −0.09 to 0.10 | 0.01 | 0.05 | −0.09 to 0.10 |
| Age (years) | −0.16*** | 0.01 | −0.17 to −0.14 | −0.16*** | 0.01 | −0.17 to −0.14 | −0.16*** | 0.01 | −0.17 to −0.14 |
| Race/ethnicity | |||||||||
| Hispanic | −1.16*** | 0.01 | −1.76 to −0.56 | −1.16*** | 0.31 | −1.76 to −0.56 | −1.17*** | 0.31 | −1.77 to −0.57 |
| Black | −1.32*** | 0.16 | −1.64 to −1.00 | −1.32*** | 0.16 | −1.64 to −1.00 | −1.32*** | 0.16 | −1.64 to −1.01 |
| Other | −1.18** | 0.44 | −2.03 to −0.32 | −1.18** | 0.44 | −2.03 to −0.32 | −1.18** | 0.43 | −2.03 to −0.33 |
| Female | −0.65*** | 0.11 | −0.87 to −0.44 | −0.65*** | 0.11 | −0.87 to −0.44 | −0.65*** | 0.11 | −0.87 to −0.44 |
| Year | 0.04** | 0.01 | 0.02 to 0.06 | 0.04** | 0.11 | 0.02 to 0.06 | 0.04** | 0.01 | 0.02 to 0.06 |
| Constant | −55.19* | 23.27 | −100.80 to −9.58 | −55.38* | 23.27 | −100.98 to −9.77 | −55.25* | 23.23 | −100.79 to −9.72 |
a
Longitudinal linear regression was used. CDT=Clock Drawing Test.
b
Model 1, predictors with no interaction terms: R2=0.33, Wald χ2=1,795.90, df=15, p<0.001; N=8,394.
c
Model 2, controlling for pain medicine × mild impairment: R2=0.33, Wald χ2=1,796.48, df=16, p<0.001; N=8,394.
d
Model 3, controlling for pain medicine × mild to moderate impairment: R2=0.33, Wald χ2=1,803.22, df=16, p<0.001; N=8,394.
*p<0.05, **p<0.01, ***p<0.001.
The interactions between executive function and pain medication revealed that participants with mild impairment in executive function who took pain medication 2–7 days a week could have a lower NHATS SPPB score (by 0.18) compared with those with normal executive function who did not take pain medication (Table 3; model 2). In the mild to moderate executive function impairment group, use of pain medications increased the chance of lower NHATS SPPB scores (coefficient=−0.21 for rare medication use and coefficient=−0.29 for use on 2–7 days per week, compared with the normal executive function group without pain medication) compared with those with mild to moderate executive function impairment who did not use pain medications (Table 3; model 3).
Severe impairment in delayed recall (DWRT) lowered the NHATS SPPB score by 0.31 compared with normal DWRT (Table 4; model 1). The interaction between DWRT score and pain medication revealed that use of pain medication 2–7 days per week combined with mild impairment in delayed recall reduced the NHATS SPPB score by 0.36, which was significantly greater than the effects of mild impairment in delayed recall and pain medication use independently (Table 4; model 2). The score reduction increased to 0.51 for those with severe impairment in delayed recall who took pain medication 2–7 days per week (Table 4; model 3).
| Model 1b | Model 2c | Model 3d | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Predictor | Coefficient | SE | 95% CI | Coefficient | SE | 95% CI | Coefficient | SE | 95% CI |
| DWRT score | |||||||||
| Mild impairment | −0.13 | 0.07 | −0.27 to 0.01 | – | – | −0.13 | 0.07 | −0.26 to 0.01 | |
| Severe impairment | −0.31*** | 0.07 | −0.45 to −0.16 | −0.31*** | 0.07 | −0.45 to −0.16 | – | – | |
| Pain medication | −0.21*** | 0.05 | −0.31 to −0.12 | – | – | – | – | ||
| Mild DWRT impairment-by-pain medication use | |||||||||
| No mild impairment plus 2–7 days/week | – | – | −0.18** | 0.06 | −0.29 to −0.07 | – | – | ||
| Mild impairment plus rarely | – | – | −0.10 | 0.07 | −0.25 to 0.04 | – | – | ||
| Mild impairment plus 2–7 days/week | – | – | −0.36*** | 0.09 | −0.54 to −0.19 | – | – | ||
| Severe DWRT impairment plus pain medication use | |||||||||
| No severe impairment plus 2–7 days/week | – | – | – | – | −0.25*** | 0.06 | −0.38 to −0.13 | ||
| Severe impairment plus rarely | – | – | – | – | −0.34*** | 0.08 | −0.49 to −0.18 | ||
| Severe impairment plus 2–7 days/week | – | – | – | – | −0.51*** | 0.09 | −0.69 to −0.34 | ||
| Sleep medication | −0.16* | 0.08 | −0.31 to −0.01 | −0.16* | 0.08 | −0.31 to −0.01 | −0.16* | 0.08 | −0.31 to −0.01 |
| Fear of falling | −0.34*** | 0.05 | −0.45 to −0.24 | −0.34*** | 0.05 | −0.45 to −0.24 | −0.34*** | 0.05 | −0.45 to −0.24 |
| Function | −0.25*** | 0.01 | −0.27 to −0.23 | −0.25*** | 0.01 | −0.27 to −0.23 | −0.25*** | 0.01 | −0.27 to −0.23 |
| No visual impairment | 0.31** | 0.11 | 0.10 to 0.52 | 0.31** | 0.11 | 0.10 to 0.52 | 0.31*** | 0.11 | 0.10 to 0.52 |
| Depression | −0.26*** | 0.05 | −0.36 to −0.17 | −0.26*** | 0.05 | −0.36 to −0.17 | −0.26*** | 0.05 | −0.36 to −0.17 |
| Anxiety | 0.01 | 0.05 | −0.08 to 0.10 | 0.01 | 0.05 | −0.08 to 1.00 | 0.01 | 0.05 | −0.09 to 1.00 |
| Age (years) | −0.15*** | 0.01 | −0.17 to −0.13 | −0.15*** | 0.01 | −0.17 to −0.13 | −0.15*** | 0.01 | −0.17 to −0.13 |
| Race/ethnicity | |||||||||
| Hispanic | −1.10*** | 0.30 | −1.69 to −0.50 | −1.10*** | 0.30 | −1.70 to −0.50 | −1.10*** | 0.30 | −1.70 to −0.50 |
| Black | −1.29*** | 0.16 | −1.60 to −0.97 | −1.29*** | 0.16 | −1.60 to −0.97 | −1.29*** | 0.16 | −1.60 to −0.97 |
| Other | −1.16** | 0.43 | −2.01 to −0.31 | −1.16** | 0.43 | −2.01 to −0.31 | −1.16** | 0.43 | −2.01 to −0.32 |
| Women | −0.68*** | 0.11 | −0.90 to −0.47 | −0.68*** | 0.11 | −0.90 to −0.47 | −0.68*** | 0.11 | −0.90 to −0.47 |
| Year | 0.04** | 0.01 | 0.01 to 0.06 | 0.04** | 0.01 | 0.01 to 0.06 | 0.04** | 0.01 | 0.01 to 0.06 |
| Constant | −51.61* | 23.02 | −96.72 to −6.50 | −51.67* | 23.02 | −96.79 to −6.56 | −51.88* | 23.02 | −96.99 to −6.77 |
a
Longitudinal linear regression was used.
b
Model 1, predictors with no interaction terms: R2=0.34, Wald χ2=1,826.44, df=15, p<0.001; N=8,394.
c
Model 2, controlling for pain medicine × mild impairment: R2=0.34, Wald χ2=1,826.82, df=16, p<0.001; N=8,394.
d
Model 3, controlling for pain medicine × severe impairment: R2=0.34, Wald χ2=1,827.01, df=16, p<0.001; N=8,394.
*p<0.05, **p<0.01, ***p<0.001.
Both mild and severe impairment in immediate recall (IWRT) lowered NHATS SPPB scores by 0.20 and 0.35, respectively (Table 5; model 1). Use of pain medication 2–7 days per week combined with mild impairment in immediate recall reduced the NHATS SPPB score by 0.43 (approximately two times greater than without pain medication) (Table 5; model 2). This score reduction increased by 0.52 for those with severe impairment in immediate recall who took pain medication 2–7 days per week (Table 5; model 3).
| Model 1b | Model 2c | Model 3d | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Predictor | Coefficient | SE | 95% CI | Coefficient | SE | 95% CI | Coefficient | SE | 95% CI |
| IWRT score | |||||||||
| Mild impairment | −0.20*** | 0.05 | −0.29 to −0.10 | – | – | −0.19*** | 0.05 | −0.29 to −0.10 | |
| Severe impairment | −0.35*** | 0.07 | −0.48 to −0.22 | −0.35*** | 0.07 | −0.48 to −0.22 | – | – | |
| Pain medication | −0.21*** | 0.05 | −0.30 to −0.12 | – | – | – | – | ||
| Mild IWRT impairment-by-pain medication use | |||||||||
| No mild impairment plus 2–7 days/week | – | – | −0.15* | 0.06 | −0.27 to −0.02 | – | – | ||
| Mild impairment plus rarely | – | – | −0.15* | 0.06 | −0.26 to −0.04 | – | – | ||
| Mild impairment plus 2–7 days/week | – | – | −0.43*** | 0.07 | −0.57 to −0.29 | – | – | ||
| Severe IWRT impairment-by-pain medication use | |||||||||
| No severe impairment plus 2–7 days/week | – | – | – | – | −0.24*** | 0.05 | −0.35 to −0.14 | ||
| Severe impairment plus rarely | – | – | – | – | −0.40 *** | 0.07 | −0.54 to −0.25 | ||
| Severe impairment plus 2–7 days/week | – | – | – | – | 0.52*** | 0.09 | −0.69 to −0.35 | ||
| Sleep medication | −0.16* | 0.08 | −0.31 to −0.01 | −0.16* | 0.08 | −0.31 to −0.01 | −0.16* | 0.08 | −0.32 to −0.01 |
| Fear of falling | −0.35*** | 0.05 | −0.45 to −0.25 | −0.35*** | 0.05 | −0.45 to −0.25 | −0.35*** | 0.05 | −0.45 to −0.25 |
| Function | −0.25*** | 0.01 | −0.27 to −0.23 | −0.25*** | 0.01 | −0.27 to −0.23 | −0.25*** | 0.01 | −0.27 to −0.23 |
| No visual impairment | 0.32** | 0.11 | 0.11–0.53 | 0.31** | 0.10 | 0.11–0.53 | 0.32** | 0.11 | 0.11–0.53 |
| Depression | −0.26*** | 0.05 | −0.36 to −0.17 | −0.26*** | 0.05 | −0.36 to −0.17 | −0.26*** | 0.05 | −0.36 to −0.17 |
| Anxiety | 0.01 | 0.05 | −0.08 to 0.11 | 0.01 | 0.05 | −0.08 to 0.11 | 0.01 | 0.05 | −0.08 to 0.11 |
| Age (years) | −0.15*** | 0.01 | −0.17 to −0.13 | −0.15*** | 0.01 | −0.17 to −0.13 | −0.15*** | 0.01 | −0.17 to −0.13 |
| Race/ethnicity | |||||||||
| Hispanic | −1.08*** | 0.30 | −1.67 to −0.49 | −1.08*** | 0.30 | −1.67 to −0.49 | −1.08*** | 0.30 | −1.67 to −0.49 |
| Black | −1.29*** | 0.16 | −1.60 to −0.98 | −1.29*** | 0.16 | −1.60 to −0.98 | −1.30*** | 0.16 | −1.61 to −0.99 |
| Other | −1.16 ** | 0.43 | −1.99 to −0.33 | −1.16** | 0.43 | −1.99 to −0.33 | −1.17** | 0.43 | −2.00 to −0.33 |
| Women | −0.68*** | 0.11 | −0.89 to −0.47 | −0.68*** | 0.11 | −0.89 to −0.47 | −0.68*** | 0.11 | −0.89 to −0.47 |
| Year | 0.04** | 0.01 | 0.02–0.06 | 0.04** | 0.01 | 0.02–0.06 | 0.04** | 0.01 | 0.02–0.06 |
| Constant | −54.34* | 22.86 | −99.14 to −9.54 | −54.76* | 22.85 | −99.14 to −9.54 | −54.82* | 22.84 | −99.60 to −10.05 |
a
Longitudinal linear regression was used.
b
Model 1, predictors with no interaction terms: R2=0.34, Wald χ2=1,849.38, df=15, p<0.001; N=8,394.
c
Model 2, controlling for pain medicine × mild impairment: R2=0.34, Wald χ2=1,853.52, df=16, p<0.001; N=8,394.
d
Model 3, controlling for pain medicine × severe impairment: R2=0.34, Wald χ2=1,853.31, df=16, p<0.001; N=8,394.
*p<0.05, **p<0.01, ***p<0.001.
Discussion
In this longitudinal study, we examined the relationship between executive function, memory, physical function, and falls among community-dwelling older adults with and without mild cognitive impairment. Our findings revealed that cognitive impairment was not directly correlated with falling, which contradicts findings from previous studies indicating that impairment in cognitive function is positively associated with falling (5, 25). Our results suggest that cognitive function can predict physical performance, which is strongly correlated with increased fall risk among community-dwelling older adults. Hence, physical performance can be a mediating factor between cognitive function and falling, because a higher cognitive reserve can protect gait speed and subsequently reduce the risk for falling (2).
We found that mild to moderate impairment in executive function predicted impairment in physical performance; the combination of pain medication use and mild to moderate impairment in executive function increased the risk for poor physical performance and, subsequently and indirectly, the risk for falling. Memory impairment negatively affected physical performance among the community-dwelling older adults in this study. Pain medication combined with impairment in immediate and delayed recall synergized this negative effect on physical performance and, subsequently, the risk for falling.
Physical instability and its predictors, such as balance deficits, reduced cognitive function, postural deformity, musculoskeletal dysfunction, visual impairment, medication, and altered gait speed, are critical factors in falling (25, 26). Most data on the circumstances of falls have been limited to older adults without cognitive impairment or with only gait dysfunction associated with cognitive impairment in falling; therefore, the present study contributes to our understanding of falling associated with impaired cognitive function. Previous studies have indicated that most falls occur during walking or changing position (i.e., getting up or sitting down). According to Berg et al. (27), standing still and high-level activities, such as climbing and running, reduce the risk for falls. Our findings contradict those of Delbaere et al. (25), which—based on a longitudinal study with a 1-year follow-up—suggested that mild cognitive impairment, particularly executive function impairment, predicts falling independently of medication use and other physiological factors that can affect balance. However, our findings are consistent with those of Liu-Ambrose et al. (1), who reported that older women with mild cognitive impairment are at greater risk for falling because of their high risk of sway in posture; the investigators suggested testing for balance and controlling for actual fall events in this population. We examined the correlation between cognitive function (executive function and memory) and physical performance, controlling for medication use and other contributing factors in an 8-year longitudinal study.
Herman et al. (5) reported executive function as a fall predictor and balance as a mediating factor between memory impairment and falls among community-dwelling older adults. Holtzer and colleagues (2) found no direct correlation between gait speed and episodic memory, whereas van Schoor et al. (6) reported impairment in immediate memory as the strongest predictor of recurrent falls. Our findings suggest that immediate recall (IWRT) predicts physical performance and, hence, fall risk. Mild and severe impairment in immediate recall significantly reduced the physical performance score by 0.20 and 0.35, respectively. Controlling for the interactions between pain medication use and immediate recall showed that taking pain medication 2–7 days a week increased the risk for poor physical performance among older adults with mild and severe immediate recall impairment when the SPPB score was reduced by 0.43 and 0.52, respectively. Although mild to moderate impairment in executive function also predicted the risk for poor physical performance, the relationship between immediate recall and SPPB score was stronger than impairment in executive function among community-dwelling older adults. This relationship may be explained by the correlation between immediate memory and attention, as well as the neurodegenerative diseases that can affect physical performance and increase the risk for falls.
Chronic musculoskeletal pain, particularly in the lower extremities, is one of the known risk factors for falling among older adults, because reducing pain is suggested to decrease the risk for falling (28). Conversely, use of NSAIDs increases fall risk (11). Our findings suggest that use of pain medication significantly reduces the SPPB score, by 0.21. Moreover, the effect of pain medication among those with mild to moderate impairment in executive function increased this reduction to 0.29. The combination of mild memory impairment and pain medication use (2–7 days per week) decreased the SPPB score by 0.36. This decrease was increased to 0.51 among participants with severe impairment in delayed recall. The IWRT results suggested a similar pattern. Considering that pain can increase the risk for falling and that pain medication synergizes the effect of executive function and memory impairment on balance, nonpharmaceutical approaches or reducing the dosage to the tolerable pain threshold could reduce the risk for poor balance and could thereby prevent falls (10). However, we did not obtain information regarding the type and dosage of participants’ pain medications, which was a limitation of our study. Future studies should explore the interactions between the type and dosage of pain medication and executive function and memory, predicting balance score and falling.
Sleep medications, particularly benzodiazepines, are used to improve sleep quantity and quality, as well as to reduce anxiety. Approximately 12% of community-dwelling older adults use such medications regularly, and >80% continue using sleep medications for >2 years even though benzodiazepines can increase the risk for falls and injuries (29). Our findings suggest that sleep medications increase the reduction in SPPB scores by 0.16. In the long term, sleep medications, particularly benzodiazepines, may cause dependency and may negatively affect memory function, which in turn increases the risk for balance loss (30).
We found that the proportion of depression among community-dwelling older adults increased by 1.4 percentage points every year. Results from previous studies have revealed that mild cognitive impairment is significantly associated with depression (31). Interestingly, we found that depression increased fall risk by 33.0%, another finding that is corroborated by the literature (32). Depression may be a prodromal manifestation of cognitive function, which can subsequently affect physical performance and falling. However, it may also be the case that increased falling leads to symptoms of depression due to decreased independence and functional mobility. This hypothesis is supported by our finding that depression reduced NHATS SPPB scores by 0.26. As such, depression appears to play an important role in cognitive function and physical performance and should therefore be a central component of fall prevention programs for older adults.
Participants from minority groups (i.e., Blacks and White Hispanics) were significantly less likely to have higher NHATS SPPB scores compared with Whites. Although there are some moderating factors between race-ethnicity and SPPB score, such as body mass index, joint pain, level of physical activities, and function, the findings in our study support those of previous research regarding higher risk for low physical performance and risk for falling among people from minority groups compared with Whites (33, 34). Hence, more public health programs targeting these populations, especially in earlier stages of life, to increase physical performance may reduce the fall risk.
A strength of this study was our use of a large representative sample in a longitudinal study while also controlling for the effect of different components of cognitive function in different models. However, cognitive function could confound the reliability of answers on self-reported measures. We also note that participants’ recall bias for episodes of falling could have affected the outcome analysis.
Conclusions
Because clear guidelines for managing mild cognitive impairment are not available, clinical approaches should be focused on risk factors and symptoms to reduce fall risks and further cognitive decline. On the basis of our findings, older adults should be screened for these items during an annual examination by their physicians and therapists. In particular, the CDT, immediate memory screening test, and standardized SPPB assessment could be viewed as early prognostic tools for fall risk. Pain medications should be adjusted or limited for individuals with deficits in cognitive function.
References
1.
Liu-Ambrose TY, Ashe MC, Graf P, et al: Increased risk of falling in older community-dwelling women with mild cognitive impairment. Phys Ther 2008; 88:1482–1491
2.
Holtzer R, Wang C, Lipton R, et al: The protective effects of executive functions and episodic memory on gait speed decline in aging defined in the context of cognitive reserve. J Am Geriatr Soc 2012; 60:2093–2098
3.
Peel NM, Alapatt LJ, Jones LV, et al: The association between gait speed and cognitive status in community-dwelling older people: a systematic review and meta-analysis. J Gerontol A Biol Sci Med Sci 2018; 74:943–948
4.
Persad CC, Jones JL, Ashton-Miller JA, et al: Executive function and gait in older adults with cognitive impairment. J Gerontol A Biol Sci Med Sci 2008; 63:1350–1355
5.
Herman T, Mirelman A, Giladi N, et al: Executive control deficits as a prodrome to falls in healthy older adults: a prospective study linking thinking, walking, and falling. J Gerontol A Biol Sci Med Sci 2010; 65:1086–1092
6.
van Schoor NM, Smit JH, Pluijm SMF, et al: Different cognitive functions in relation to falls among older persons: immediate memory as an independent risk factor for falls. J Clin Epidemiol 2002; 55:855–862
7.
Guralnik JM, Ferrucci L, Pieper CF, et al: Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the Short Physical Performance Battery. J Gerontol A Biol Sci Med Sci 2000; 55:M221–M231
8.
Ward A, Arrighi HM, Michels S, et al: Mild cognitive impairment: disparity of incidence and prevalence estimates. Alzheimers Dement 2012; 8:14–21
9.
McCall WV: Sleep in the elderly: burden, diagnosis, and treatment. Prim Care Companion J Clin Psychiatry 2004; 6:9–20
10.
Ali A, Arif AW, Bhan C, et al: Managing chronic pain in the elderly: an overview of the recent therapeutic advancements. Cureus 2018; 10:e3293
11.
Walker PC, Alrawi A, Mitchell JF, et al: Medication use as a risk factor for falls among hospitalized elderly patients. Am J Health Syst Pharm 2005; 62:2495–2499
12.
National Health and Aging Trends Study: Welcome Researchers. Baltimore, Johns Hopkins University Bloomberg School of Public Health. nhats.org/researcher
13.
Kasper JD, Freedman VA: National Health and Aging Trends Study User Guide: Rounds 1, 2 & 3 Final Release. Baltimore, Johns Hopkins University Bloomberg School of Public Health, 2015
14.
A Profile of Older Americans: 2012. Washington, DC, US Department of Health and Human Services, Administration for Community Living, 2012
15.
Total Number of Medicare Beneficiaries. San Francisco, CA, Kaiser Family Foundation, 2011. [updated 2021]. https://www.kff.org/medicare/state-indicator/total-medicare-beneficiaries/?currentTimeframe=0&sortModel=%7B“colId”:“Location”,“sort”:“asc”%7D
16.
Montaquila J, Freedman VA, Spillman B, et al: National Health and Aging Trends Study—National Study of Caregiving, 2012. www.NHATS.org
17.
Kasper JD, Freedman VA: National Health and Aging Trends Study User Guide: Rounds 1–9 Beta Release. Baltimore Johns Hopkins University Bloomberg School of Public Health, 2020
18.
Kasper JD, Freedman VA, Niefeld MR: Construction of Performance-Based Summary Measures of Physical Capacity in the National Health and Aging Trends Study, NHATS Technical Paper #4. Baltimore, Johns Hopkins University Bloomberg School of Public Health, 2012
19.
Samton JB, Ferrando SJ, Sanelli P, et al: The Clock Drawing Test: diagnostic, functional, and neuroimaging correlates in older medically ill adults. J Neuropsychiatry Clin Neurosci 2005; 17:533–540
20.
Lyness SA, Lee AY, Zarow C, et al: 10-minute delayed recall from the Modified Mini–Mental State Test predicts Alzheimer’s disease pathology. J Alzheimers Dis 2014; 39:575–582
21.
Takayama Y: A delayed recall battery as a sensitive screening for mild cognitive impairment: follow-up study of memory clinic patients after 10 years. J Med Dent Sci 2010; 57:177–184
22.
Gill TM, Williams CS: Development and validation of a functional outcome measure in the National Health and Aging Trends Study. J Gerontol A Biol Sci Med Sci 2018; 73:1111–1118
23.
Kroenke K, Spitzer RL, Williams JBW: The Patient Health Questionnaire–2: validity of a two-item depression screener. Med Care 2003; 41:1284–1292
24.
Plummer F, Manea L, Trepel D, et al: Screening for anxiety disorders with the GAD-7 and GAD-2: a systematic review and diagnostic metaanalysis. Gen Hosp Psychiatry 2016; 39:24–31
25.
Delbaere K, Kochan NA, Close JCT, et al: Mild cognitive impairment as a predictor of falls in community-dwelling older people. Am J Geriatr Psychiatry 2012; 20:845–853
26.
Tinetti ME, Speechley M, Ginter SF: Risk factors for falls among elderly persons living in the community. N Engl J Med 1988; 319:1701–1707
27.
Berg WP, Alessio HM, Mills EM, et al: Circumstances and consequences of falls in independent community-dwelling older adults. Age Ageing 1997; 26:261–268
28.
Leveille SG, Jones RN, Kiely DK, et al: Chronic musculoskeletal pain and the occurrence of falls in an older population. JAMA 2009; 302:2214–2221
29.
de Jong MR, Van der Elst M, Hartholt KA: Drug-related falls in older patients: implicated drugs, consequences, and possible prevention strategies. Ther Adv Drug Saf 2013; 4:147–154
30.
Hanlon JT, Horner RD, Schmader KE, et al: Benzodiazepine use and cognitive function among community-dwelling elderly. Clin Pharmacol Ther 1998; 64:684–692
31.
Tyrovolas S, Koyanagi A, Lara E, et al: Mild cognitive impairment is associated with falls among older adults: findings from the Irish Longitudinal Study on Ageing (TILDA). Exp Gerontol 2016; 75:42–47
32.
Kvelde T, Lord SR, Close JCT, et al: Depressive symptoms increase fall risk in older people, independent of antidepressant use, and reduced executive and physical functioning. Arch Gerontol Geriatr 2015; 60:190–195
33.
Bartley EJ, Hossain NI, Gravlee CC, et al: Race/ethnicity moderates the association between psychosocial resilience and movement-evoked pain in knee osteoarthritis. ACR Open Rheumatol 2019; 1:16–25
34.
Cruz-Almeida Y, Sibille KT, Goodin BR, et al: Racial and ethnic differences in older adults with knee osteoarthritis. Arthritis Rheumatol 2014; 66:1800–1810
Information & Authors
Information
Published In
History
Received: 30 May 2021
Revision received: 26 November 2021
Accepted: 20 January 2022
Published online: 10 June 2022
Published in print: Winter 2023
Keywords
Authors
Competing Interests
The authors report no financial relationships with commercial interests.
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).