Longitudinal Association Between Subjective Fatigue and Future Falls in Community-Dwelling Older Adults: The Locomotive Syndrome and Health Outcomes in the Aizu Cohort Study (LOHAS)

Abstract

Objective:We examined the longitudinal association between the severity of fatigue and falls in community-dwelling older adults. Method: Subjective fatigue was assessed using the Short Form 36 Health Survey (SF-36) Vitality subscale and classified into four categories by quartile (mildest, mild, moderate, severe). The main outcome was the incidence of any falls during the 2-year follow-up period. Results: Of the 751 participants, 236 (31.4%) experienced falls during the 2-year period. In multivariable logistic regression analysis with adjustment for possible confounding factors, the adjusted odds ratios (and 95% confidence intervals) for mild, moderate, and severe categories (vs. mildest category) of 1.60 (0.94-2.75), 1.87 (1.12-3.11), and 2.15 (1.23-3.76), respectively (p for trend = .007). Discussion: Our results suggest that the severity of fatigue is associated with the risk of subsequent falls for community-dwelling older adults even after adjustment for possible confounding factors.

Keywords

fall fatigue older adults population-based study

Introduction

Falls are a common problem facing older adults in an aging society, with 30% of those aged over 65 years said to experience falls at least once annually (Tinetti & Williams, 1997). Approximately 80% to 90% of bone fractures in older adults are caused by falls (Morrison, Fan, Sen, & Weisenfluh, 2013). Furthermore, falls lead to serious public health problems such as hospitalization, institutionalization, and an increase in comorbidities (Gill, Murphy, Gahbauer, & Allore, 2013; Kannus, Sievänen, Palvanen, Järvinen, & Parkkari, 2005).

The incidence of falls is associated with multiple and complex factors, and their prevention in elderly people is difficult. Previous studies have commonly identified gait disorder, cognitive impairment, and medications as risk factors for falls (Hendrich, Bender, & Nyhuis, 2003; Toba, Kikuchi, Iwata, & Kozaki, 2009). A close association with social and environmental factors is also generally accepted; indeed, the Fall Risk Index (FRI-21), a self-administered questionnaire to assess the risk of falls, includes examination of social factors and environmental factors (Matsubayashi, 2011). Moreover, Tinetti, Williams, and Mayewski (1986) reported a positive association between the number of disabilities and the risk of falls. There is, therefore, an important clinical need to comprehensively identify unknown risk factors for falls in older adults.

Fatigue is defined as a feeling of low energy, and subjective fatigue is an important symptom in older adults. In fact, 20% to 50% of older adults living in the community are said to complain of being fatigued (Yu, Lee, & Man, 2010). In addition, some definitions of frailty, as exemplified by the Cardiovascular Health Study scale established by Fried and colleagues, include subjective fatigue as a key component (Fried et al., 2001; Malmstrom, Miller, & Morley, 2014). In several studies, fatigue was associated with restricted activity, disability (Vestergaard et al., 2009), and an increase in mortality in older adults (Hardy & Studenski, 2008). One study demonstrated an association between falls and the Short Form 36 Health Survey (SF-36) Vitality subscale score (Burns, Byles, Mitchell, & Anstey, 2012). However, the participants of that study were only women aged over 55 years, and some important confounding factors such as falls history, which is one of the most important confounding factors associated with falls (Bryant et al., 2012; Yamashita, Noe, & Bailer, 2012), were not considered in their analysis. It is, therefore, unknown whether a relationship between the severity of fatigue and falls would remain after adjustment for these important confounding factors.

In this study, we examined whether the severity of fatigue is independently associated with the incidence of falls during a 2-year follow-up period, using data for older community-dwelling adults in the Locomotive Syndrome and Health Outcome in Aizu Cohort Study (LOHAS), a population-based prospective cohort study in Japan.

Method

Study Participants

This study protocol was approved by the Ethics Committee of Kyoto University Graduate School of Medicine and Fukushima Medical University, and all participants provided written informed consent. LOHAS is a cohort study that started in 2008 and involves residents aged 40 to 80 years who participated in annual health check-ups in two communities (Tadami and Minamiaizu) in Fukushima Prefecture, Japan. The two communities are located next to each other, and their features are very similar. Inclusion criteria were age over 60 years in 2008 and participation in all health check-ups during 2008-2010 (three times). We did not set exclusion criteria. Details of LOHAS have been reported elsewhere (Otani et al., 2012).

Fatigue Assessment

The main assessment item was fatigue, assessed using the Japanese version of the SF 36 (version 2) subscale Vitality (Fukuhara, Bito, Green, Hsiao, & Kurokawa, 1998), which is commonly used to validate instruments designed to assess fatigue in the general population and patients with diseases (Choi et al., 2014; Hewlett, Dures, & Almeida, 2011; Lindeberg, Ostergren, & Lindbladh, 2006). A previous study showed that the Japanese version of the SF-36 subscale Vitality had good internal consistency (Cronbach’s α = .77; Fukuhara & Suzukamo, 2011). The Vitality subscale consists of four questions about the condition of the participant over the previous month: (a) felt full of pep, (b) had a lot of energy, (c) felt worn out, or (d) felt tired. Five responses to each question were possible, ranging from “all of the time” to “none of the time.” We can calculate a total score in the range of 0 to 100. In the present study, we calculated the score of subjective fatigue, which was calculated by subtracting the Vitality subscale score from 100, with a higher score reflecting more severe fatigue. Subjective fatigue was classified into four categories (mildest, mild, moderate, severe) in accordance with score quartile.

In addition, we used only two items in the four questions that directly ask about fatigue, namely, “felt worn out” and “felt tired,” to calculate the score of the two-item version of subjective fatigue, giving a range of 2 to 10. We also classified the two-item version of subjective fatigue into four categories (mildest, mild, moderate, severe).

Falls Assessment

We assessed the incidence of falls in the past 1 year via a self-reported questionnaire at each annual health check-up in 2009 and 2010. In the present study, the incidence of falls was defined as the experience of at least one fall during the 2-year follow-up period.

Assessment of Other Potential Confounding Factors

As potential confounding factors, we assessed sociodemographic and clinical characteristics including age and sex, depression symptoms, sleep disorder, comorbidities, amount of physical activity, and fall history, as obtained from a self-reported questionnaire, as well as body mass index (BMI) and physical function, as measured by local nurse practitioners or clinicians. Depression symptoms were assessed with the 10-item version of the Center for Epidemiological Studies Depression Screening Index (CES-D), which evaluates the number and frequency of symptoms (Andresen, Malmgren, Carter, & Patrick, 1994), with symptoms treated as a continuous variable. A previous study reported that the 10-item version of CES-D showed good internal consistency (Cronbach’s α = .83; Irwin, Artin, & Oxman, 1999). Sleep disorder was assessed with one question, “How do you rate your sleep quality in the past month?” with participants required to choose from the four responses of “very good,” “good,” “poor,” and “very poor.” We identified those who chose “poor” or “very poor” as having a sleep disorder. Because the previous studies reported that the number of chronic medical conditions were closely associated with increased risk of falls (Afrin et al., 2016; Lawlor, Patel, & Ebrahim, 2003), we noted the number of comorbidities, including hypertension, diabetes mellitus, kidney disease, hepatic disease, cardiovascular disease, cerebrovascular disease, respiratory disease, and anemia, and divided the result into three categories (number of comorbidities: 0, 1, or ≥2). Physical function was assessed with handgrip strength (Hairi et al., 2010; Legrand et al., 2014) and one-leg standing time (Hurvitz, Richardson, Werner, Ruhl, & Dixon, 2000). The hand with the stronger handgrip strength was assessed. One-leg standing time was divided into two categories according to whether the subject could stand on one leg for more than 30 s. The average amount of physical activity in 1 week was calculated according to the algorithm of the International Physical Activity Questionnaire (Craig et al., 2003) in metabolic equivalents (METs) and then divided into four categories by quartile (Q1, the least amount of physical activity; Q2, Q3, and Q4, the greatest amount of physical activity in ascending order). Fall history was defined as the experience of at least one fall in the past year at baseline. BMI was divided into three categories (<25 or 25 to 29.9 or ≥30 kg/m²).

Statistical Analysis

The primary analysis was conducted in participants with no missing data. All confounding factors were described by each fatigue state, and values were expressed as the mean and standard deviation (SD) for continuous variables, and number and proportion (%) for categorical variables. We calculated the adjusted odds ratio (OR) as an indicator of effect using logistic regression analysis. The dependent variable was the incidence of any fall in the past 2 years. The independent variable was the score of subjective fatigue. To assess the linearity of association between falls and fatigue severity, we performed the fractional polynomial method to verify whether nonlinear models were shown to be superior to a linear model in term of the goodness of fit or not (Royston, 2000; Williams, 2011). In addition, we included categories of subjective fatigue (reference: mildest) as dummy variables and performed a test of linear trends across four fatigue categories in accordance with the method of previous studies (Hu et al., 1999). To investigate if the relationship between the incidence of falls and fatigue severity was consistent when we used the two-item version of subjective fatigue, which is a simpler method of assessment, the raw score of and the categorized two-item version of subjective fatigue (reference: mildest category) were employed as an independent variable, and the same analyses as above were repeated. All potential confounding factors were adjusted for in each analysis.

Sensitivity Analysis

We also performed a multiple imputation procedure using the chained equations method for 208 (21.7%) participants with at least one missing confounding factor based on the assumption of missing at random. The missing values were imputed using the existing score of subjective fatigue and other confounding factors. We created 20 imputed datasets and then integrated them into one dataset. The statistical model was the same as that in the primary analysis for the integrated dataset.

All analyses were conducted using Stata version 13.1 (StataCorp LP, College Station, TX). A p value <.05 (two-tailed) was considered statistically significant.

Results

51 of the 1,010 participants in each health check-up had missing data for the SF-36 Vitality subscale at baseline or for falls in either health check-up at during the follow-up periods (Figure 1). After excluding 208 (21.7%) participants with at least one missing datum for confounding factors, 751 participants were included in the primary analysis.

Figure 1.

Flow chart of study participants.

Characteristics of study participants at baseline are described in Table 1. Mean age was 69.9 years (SD = 4.9 years) and the proportion of females was 61.7% (n = 463). The median values of the score of the subjective fatigue for the four categories of mildest, mild, moderate, and severe were 34.1, 43.7, 50.2, and 59.8, respectively. Baseline characteristics for the 1,010 participants who were initially included and the 751 participants in the main analysis were similar (see the appendix).

Table 1.

Characteristics of Study Participants According to Four Categories of Subjective Fatigue.

	TotalN = 751	Four categories of subjective fatigue
	TotalN = 751	Mildest (0-12.5)n = 181	Mild (18.75-25)n = 169	Moderate (31.25-43.75)n = 213	Severe (50-100)n = 188
	M [SD], n (%)
Age	69.9 [4.9]	70.6 [5.1]	69.6 [5.0]	70.1 [4.7]	69.3 [4.9]
Sex (male)	288 (38.3)	82 (45.3)	72 (42.6)	77 (34.2)	57 (29.3)
BMI
<25 kg/m²	19 (2.5)	6 (3.3)	3 (1.8)	5 (2.4)	5 (2.7)
25-29.9	467 (62.2)	110 (60.8)	107 (63.3)	132 (62.0)	118 (62.8)
≥30	265 (35.3)	65 (35.9)	59 (34.9)	76 (35.7)	65 (34.6)
CES-D score	4.1 [3.9]	2.0 [2.6]	3.4 [2.7]	4.4 [3.3]	6.6 [4.8]
Sleep disorder	99 (13.2)	7 (3.9)	23 (13.6)	32 (15.0)	37 (19.7)
Standing on one leg (<30 sec)	268 (28.4)	57 (26.2)	57 (27.7)	81 (29.8)	73 (29.6)
Grip strength (kg)	30.9 [9.2]	32.6 [9.7]	31.8 [9.1]	30.6 [9.0]	29.0 [8.6]
Amount of physical activity
Q1 (Mets ≤2180)	188 (25.0)	42 (23.2)	40 (23.7)	59 (27.7)	47 (25.0)
Q2 (2181-6684)	187 (24.9)	42 (23.2)	53 (31.4)	45 (21.1)	47 (25.0)
Q3 (6685-14597)	185 (24.6)	44 (25.3)	34 (20.1)	61 (28.6)	46 (24.5)
Q4 (≥14598)	191 (25.4)	53 (29.3)	42 (24.9)	48 (22.5)	48 (25.5)
Number of comorbidities
0	333 (44.3)	87 (48.1)	73 (43.2)	94 (44.1)	79 (42.0)
1	306 (40.8)	68 (37.6)	74 (43.8)	87 (40.9)	77 (41.0)
≥2	112 (14.9)	26 (14.4)	22 (13.0)	32 (15.0)	40 (17.0)
Fall history	141 (18.8)	21 (11.6)	28 (16.6)	49 (23.0)	43 (22.9)

Note. BMI = body mass index; CES-D = Center for Epidemiological Studies Depression Screening Index; Mets = metabolic equivalents.

Falls during the 2-year follow-up period were reported by 31.4% of all participants (n = 236). The incidence of falls in the four categories of mildest, mild, moderate, and severe was 21.6%, 28.4%, 36.0%, and 38.3%, respectively (Figure 2).

Figure 2.

Incidence of falls during the 2-year follow-up period according to the four categories of subjective fatigue.

Longitudinal Association Between Subjective Fatigue and Future Falls

In the multivariable logistic regression analysis with adjustment for confounding factors, the score of subjective fatigue independently associated with future falls, as the adjusted OR was 1.42 (per 1 SD increase, 95% confidence interval [CI] = [1.16, 1.73], p = .001; Table 2). The fractional polynomial method demonstrated that any nonlinear models were not shown to be superior to a linear model in term of the goodness of fit. In the analysis including categories of subjective fatigue, participants with severe fatigue had the greatest likelihood of falls (Figure 3); using the mildest fatigue category as reference, the adjusted ORs (95% CI) for mild, moderate, and severe fatigue were 1.60 [0.94, 2.75], p = .084; 1.87 [1.12, 3.11], p = .017; and 2.15 [1.23, 3.76], p = .007 (p for trend = .007), respectively. ORs and 95% CIs of other factors associated with future falls are shown in Table 2.

Table 2.

Adjusted Odds Ratio of Falls in Models With Subjective Fatigue and Two-Items Version of Subjective Fatigue.

	Falls during the 2-year follow-up
	Adjusted OR (95% CI)
	Subjective fatigue	p value	Subjective fatigue (two-items version)	p value
Score of subjective fatigue (per 1 SD)	1.42 [1.16, 1.73]	.001	1.47 [1.22, 1.79]	<.001
Age (per year)	1.04 [1.00, 1.08]	.072	1.04 [1.00, 1.08]	.076
Sex
Female (vs. male)	1.03 [0.56, 1.90]	.92	1.06 [0.56, 2.01]	.86
Standing on one leg standing
<30 sec (vs. ≥30)	1.56 [1.05, 2.34]	.030	1.49 [0.99, 2.25]	.056
Grip strength (per kg)	1.00 [0.97, 1.04]	.86	1.01 [0.97, 1.04]	.72
CES-D score (per point)	1.00 [0.95, 1.05]	.95	1.01 [0.96, 1.06]	.77
Sleep disorder
Present (vs. absent)	1.24 [0.74, 2.09]	.41	1.24 [0.74, 2.09]	.42
Number of comorbidities
0	Ref		Ref
1	0.74 [0.51, 1.09]	.125	0.75 [0.51, 1.10]	.144
≥2	1.01 [0.60, 1.68]	.97	1.04 [0.62, 1.74]	.89
Amount of physical activity
Q1	Ref		Ref
Q2	1.14 [0.70, 1.85]	.60	1.02 [0.63, 1.67]	.92
Q3	1.23 [0.76, 1.99]	.40	1.06 [0.65, 1.74]	.80
Q4	0.83 [0.51, 1.36]	.46	0.75 [0.45, 1.24]	.26
BMI
<25 kg/m²	Ref		Ref
25-29.9	0.83 [0.29, 2.40]	.73	0.81 [0.28, 2.32]	.70
≥30	0.69 [0.23, 2.04]	.59	0.65 [0.22, 1.92]	.44
Fall history
Present (vs. absent)	6.43 [4.23, 9.75]	<.001	6.36 [4.17, 9.70]	<.001

Note. OR = odds ratio; CI = confidence interval; CES-D = Center for Epidemiological Studies Depression Screening Index; BMI = body mass index.

Figure 3.

Adjusted OR of falls in models with subjective fatigue and two-items version of subjective fatigue.

Longitudinal Association Between the Two-Item Version of Subjective Fatigue and Falls

Using the two-item version of subjective fatigue, the score of subjective fatigue independently associated with future falls, as the adjusted OR was 1.48 (per 1 SD increase, 95% CI = [1.22, 1.79], p < .001; Table 2). In addition, in the analysis including categories of subjective fatigue, participants with severe fatigue had a greater likelihood of any fall (Figure 3). Using the mildest category as reference, adjusted ORs (95% CI) for mild, moderate, and severe fatigue were 1.73 [1.05, 2.86], p = .030; 2.20 [1.39, 3.48], p = .001; and 3.10 [1.58, 6.08], p = .001, respectively (p for trend < .001). ORs and 95% CIs of other factors associated with future falls are shown in Table 2.

Sensitivity Analysis

Sensitivity analysis using the multiple imputation approach showed a consistent association between subjective fatigue and falls: the adjusted OR of the score of subjective fatigue was 1.32 (per 1 SD increase, 95% CI = [1.10, 1.57], p = .002), and when using the mildest category as reference, the adjusted ORs (95% CI) of mild, moderate, and severe categories were 1.33 [0.82, 2.16], p = .25; 1.67 [1.06, 2.62], p = .025; and 1.75 [1.07, 2.86], p = .031, respectively (p for trend = .020).

Discussion

In this longitudinal study of community-dwelling older adults, subjective fatigue was independently associated with future falls during a 2-year period, even after adjustment for potential confounding factors. The association showed a positive degree of fatigue-dependent association. These results suggest a causal relationship between fatigue and falls. Moreover, the robustness of the results was confirmed in the sensitivity analysis.

The SF-36 Vitality subscale has been used to assess the general population as well as patients with some diseases. The subscale conceptualizes fatigue as a single continuum from full of energy to fatigue (Hewlett et al., 2011). The Vitality subscale contributes to both the mental and physical component scores of SF-36, and is associated equally with the two components. Evaluation of general fatigue, therefore, includes both physical and mental aspects. However, it was also suggested that the vitality was more strongly related to mental health than to the physical health (Burns et al., 2012). In fact, a previous study of 2,832 community-dwelling older adults aged more than 65 years suggested an association between fatigue as measured with the SF-36 Vitality subscale and cognitive and executive functions (Lin, Chen, Vance, Ball, & Mapstone, 2013). From our present results, one possible reason for the observed relationship is that more severe fatigue was associated with future falls because of lower executive and attentional functions, which are understood to be risk factors for falls (Mirelman et al., 2012).

In our population, the smallest difference in the median value of the fatigue score was about 10 points between the mildest and mild categories. When we compared this difference with patients with other diseases, we found that it was larger than that between patients with and without kidney disease (Bjorner et al., 2007). Accordingly, the differences between each fatigue category in our study were clinically important, meaning that the classification was valid for assessing the association with clinical outcome.

The several previous studies of an association between fatigue and falls in community-dwelling older adults suffered from a number of serious problems. First, one previous study (Olsson Möller et al., 2013) demonstrated that fatigue was associated with future falls in those aged more than 80 years only, and not in those under 80. In contrast, we found an association even in the population including those aged under 80 years (99.6%). This discrepancy may have arisen from the different scales used to assess fatigue. We precisely evaluated fatigue using a continuous 0 to 100 point scoring system in the validated and widely used SF-36 subscales, but the previous study assessed this using a single question which asked about the presence or absence of fatigue. Second, the other previous study (Burns et al., 2012) demonstrated that the SF-36 Vitality subscale score was associated with the likelihood of falls in a large cohort, which consisted of females only. We included males (38.3%) among our study participants, and observed consistent results of the previous study. These results suggest that our findings might be more applicable to older adults in general than these previous studies. Third, in addition, the previous study (Burns et al., 2012) did not consider some potential confounding factors, including fall history—which has been recognized as the most important predictive factor for future falls—sleep disorders or comorbidities. Therefore, the observed association in the previous study between the SF-36 Vitality subscale score and falls was not sufficiently adjusted. In contrast, we observed a significant association even after adjustment for these relevant confounding factors, suggesting that our study was the first to demonstrate a valid association between fatigue and falls.

With regard to other factors entered in the multivariable model, results showed that objectively measured physical function and fall history were significantly associated with future falls, supporting the results of the previous study. In contrast, some known confounding factors, such as sex and depression symptoms, did not show a significant association with future falls. These results suggest that the characteristics of study participants in the present and these previous studies differed.

The strength of the present study is that the results were obtained from a relatively large cohort, which supports the generalizability of the observed relationship. Moreover, we adjusted for already-known potential confounding factors, which is essential for validating the results. Assessment using the two-item version of the fatigue score might be suitable for detecting the risk of future falls. This two-items version is brief and less burdensome, allowing the presence of fatigue to be more easily and conveniently assessed in clinical practice. We believe that our findings have impact on clinical practice because fatigue can be treated by cognitive behavior therapy, group exercise, and graded exercise therapy (Alexander et al., 2010; Whiting et al., 2001). The more effective fall prevention program may be implemented by including these treatments of fatigue in community-dwelling older adults.

However, our study has several limitations. First, the study is limited by our use of the SF-36 Vitality subscale to assess fatigue. This subscale is commonly used worldwide, but whether the relationship between fatigue and falls can be demonstrated using other scales is unknown. Second, we did not have data for some unmeasured potential confounding factors, including use of medications such as antihypertensives and sedatives. We, therefore, tried to reduce the effect of unmeasured confounding factors by adjusting for depression symptoms and sleep disorder, which are closely associated with medications like antihypertensives or sedatives. However, we could not consider other important confounding factors such as visual impairment and impairment due to disability. Third, data of confounding factors were missing for 208 (21.7%) participants. This might have resulted in selection bias in the present study, although sensitivity analysis using multiple imputation showed a similar association between fatigue and falls. Fourth, our assessment of falls was based on questionnaire asking their experience of falls during the past 1 year. This approach can cause recall bias, which affects the accuracy of falls measurement. Last, results from the longitudinal analysis revealed that while subjective fatigue possibly affects future falls, the prevention of falls in older people by improving fatigue cannot be guaranteed. As a general limitation of observational studies, adjustment of unknown confounding factors highly associated with the relationships observed in the study was not possible.

Conclusion

The relationship between fatigue and falls was positively associated in community-dwelling older adults. The association between fatigue and falls remained even after adjustment for potential confounding factors. Further studies are needed to confirm whether an intervention to reduce fatigue can prevent falls in older adults.

Footnotes

Appendix

Table A2.

Univariate Analysis.

	Fallers (n = 236)	Non fallers (n = 515)	p value
	M [SD], n (%)		p value
Score of subjective fatigue	37.6 [21.4]	29.1 [19.7]	<.001
Score of the 2-item version of subjective fatigue	5.1 [1.9]	4.2 [1.8]	<.001
Age (years)	70.6 [5.1]	69.6 [4.8]	.017
Sex (male)	82 (34.8)	206 (40.0)	.169
Standing on one leg standing (<30 s)	83 (35.2)	122 (23.7)	.001
Grip strength (kg)	29.8 [9.0]	31.7 [9.2]	.002
CES-D score	4.7 [4.2]	3.9 [3.7]	.007
Sleep disorder	39 (16.5)	60 (11.7)	.067
Number of comorbidities			.29
0	109 (46.2)	224 (43.5)
1	87 (38.9)	219 (42.5)
≥2	40 (17.0)	72 (14.0)
Amount of physical activity (Mets)			.47
Q1	59 (25.0)	129 (25.1)
Q2	57 (24.2)	130 (25.2)
Q3	66 (28.0)	119 (23.1)
Q4	54 (22.9)	137 (26.6)
BMI			.58
<25 (kg/m²)	6 (2.5)	13 (2.5)
25-29.9	153 (64.8)	314 (61.0)
≥25	77 (32.6)	188 (36.5)
Fall history	95 (40.3)	46 (8.9)	<.001

Note. Unpaired t test or Wilcoxon rank sum test for continuous variables (score of subjective fatigue, score of the two-items version of subjective fatigue, age, grip strength, CES-D score). Pearson’s chi-square test for categorical variables (sex, standing on one leg, sleep disorder, number of comorbidities, BMI, fall history). BMI = body mass index; CES-D = Center for Epidemiological Studies Depression Screening Index; Mets = metabolic equivalents.

Acknowledgements

The authors thank the staff of the public offices of Tadami and Minami-Aizu for their assistance in locating participants and scheduling examinations. The authors also thank the participants of the LOHAS.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Afrin

Honkanen

Koivumaa-Honkanen

Lukkala

Rikkonen

Sirola

. . . Kröger

(2016). Multimorbidity predicts falls differentially according to the type of fall in postmenopausal women. Maturitas, 91, 19-24. doi:10.1016/j.maturitas.2016.05.004

Alexander

N. B.

Taffet

G. E.

Horne

F. M.

Eldadah

B. A.

Ferrucci

Nayfield

. . . Studenski

(2010). Bedside-to-Bench conference: Research agenda for idiopathic fatigue and aging. Journal of the American Geriatrics Society, 58, 967-975. doi:10.1111/j.1532-5415.2010.02811.x

Andresen

E. M.

Malmgren

J. A.

Carter

W. B.

Patrick

D. L.

(1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). American Journal of Preventive Medicine, 10, 77-84.

Bjorner

J. B.

Wallenstein

G. V.

Martin

M. C.

Lin

Blaisdell-Gross

Tak Piech

. . . Mody

S. H.

(2007). Interpreting score differences in the SF-36 vitality scale: using clinical conditions and functional outcomes to define the minimally important difference. Current Medical Research and Opinion, 23, 731-739. doi:10.1185/030079907X178757

Bryant

M. S.

Rintala

D. H.

Hou

J. G.

Rivas

S. P.

Fernandez

A. L.

Lai

E. C.

. . . Protas

E. J.

(2012). The relation of falls to fatigue, depression and daytime sleepiness in Parkinson’s disease. European Neurology, 67, 326-330. doi:10.1159/000335877

Burns

R. A.

Byles

Mitchell

Anstey

K. J.

(2012). Positive components of mental health provide significant protection against likelihood of falling in older women over a 13-year period. International Psychogeriatrics, 24, 1419-1428. doi:10.1017/S1041610212000154

Choi

Tate

J. A.

Hoffman

L. A.

Schulz

Ren

Donahoe

M. P.

. . . Sherwood

P. R.

(2014). Fatigue in family caregivers of adult intensive care unit survivors. Journal of Pain and Symptom Management, 48, 353-363. doi:10.1016/j.jpainsymman.2013.09.018

Craig

C. L.

Marshall

A. L.

Sjöström

Bauman

A. E.

Booth

M. L.

Ainsworth

B. E.

. . . Oja

(2003). International Physical Activity Questionnaire: 12-country reliability and validity. Medicine and Science in Sports and Exercise, 35, 1381-1395. doi:10.1249/01.MSS.0000078924.61453.FB

Fried

L. P.

Tangen

C. M.

Walston

Newman

A. B.

Hirsch

Gottdiener

. . . Cardiovascular Health Study Collaborative Research Group. (2001). Frailty in older adults: Evidence for a phenotype. The Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 56, M146-M156.

10.

Fukuhara

Bito

Green

Hsiao

Kurokawa

(1998). Translation, adaptation, and validation of the SF-36 Health Survey for use in Japan. Journal of Clinical Epidemiology, 51, 1037-1044.

11.

Fukuhara

Suzukamo

(2011). Manual of SF-36v2 Japanese version. Kyoto: Institute for Health Outcomes & Process Evaluation Research.

12.

Gill

T. M.

Murphy

T. E.

Gahbauer

E. A.

Allore

H. G.

(2013). Association of injurious falls with disability outcomes and nursing home admissions in community-living older persons. American Journal of Epidemiology, 178, 418-425. doi:10.1093/aje/kws554

13.

Hairi

N. N.

Cumming

R. G.

Naganathan

Handelsman

D. J.

Le Couteur

D. G.

Creasey

. . . Sambrook

P. N.

(2010). Loss of muscle strength, mass (sarcopenia), and quality (specific force) and its relationship with functional limitation and physical disability: The Concord Health and Ageing in Men Project. Journal of the American Geriatrics Society, 58, 2055-2062. doi:10.1111/j.1532-5415.2010.03145.x

14.

Hardy

S. E.

Studenski

S. A.

(2008). Fatigue predicts mortality in older adults. Journal of the American Geriatrics Society, 56, 1910-1914. doi:10.1111/j.1532-5415.2008.01957.x

15.

Hendrich

A. L.

Bender

P. S.

Nyhuis

(2003). Validation of the Hendrich II Fall Risk Model: A large concurrent case/control study of hospitalized patients. Applied Nursing Research, 16, 9-21. doi:10.1053/apnr.2003.YAPNR2

16.

Hewlett

Dures

Almeida

(2011). Measures of fatigue: Bristol Rheumatoid Arthritis Fatigue Multi-Dimensional Questionnaire (BRAF MDQ), Bristol Rheumatoid Arthritis Fatigue Numerical Rating Scales (BRAF NRS) for severity, effect, and coping, Chalder Fatigue Questionnaire (CFQ), Checklist Individual Strength (CIS20R and CIS8R), Fatigue Severity Scale (FSS), Functional Assessment Chronic Illness Therapy (Fatigue) (FACIT-F), Multi-Dimensional Assessment of Fatigue (MAF), Multi-Dimensional Fatigue Inventory (MFI), Pediatric Quality Of Life (PedsQL) Multi-Dimensional Fatigue Scale, Profile of Fatigue (ProF), Short Form 36 Vitality Subscale (SF-36 VT), and Visual Analog Scales (VAS). Arthritis Care & Research, 63(Suppl. 11), S263-S286. doi:10.1002/acr.20579

17.

F. B.

Sigal

R. J.

Rich-Edwards

J. W.

Colditz

G. A.

Solomon

C. G.

Willett

W. C.

. . . Manson

J. E.

(1999). Walking compared with vigorous physical activity and risk of type 2 diabetes in women: A prospective study. Journal of the American Medical Association, 282, 1433-1439.

18.

Hurvitz

E. A.

Richardson

J. K.

Werner

R. A.

Ruhl

A. M.

Dixon

M. R.

(2000). Unipedal stance testing as an indicator of fall risk among older outpatients. Archives of Physical Medicine and Rehabilitation, 81, 587-591.

19.

Irwin

Artin

K. H.

Oxman

M. N.

(1999). Screening for depression in the older adult: Criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D). Archives of Internal Medicine, 159, 1701-1704. doi:10.1001/archinte.159.15.1701

20.

Kannus

Sievänen

Palvanen

Järvinen

Parkkari

(2005). Prevention of falls and consequent injuries in elderly people. The Lancet, 366, 1885-1893. doi:10.1016/S0140-6736(05)67604-0

21.

Lawlor

D. A.

Patel

Ebrahim

(2003). Association between falls in elderly women and chronic diseases and drug use: Cross sectional study. British Medical Journal, 327, 712-717. doi:10.1136/bmj.327.7417.712

22.

Legrand

Vaes

Matheï

Adriaensen

Van Pottelbergh

Degryse

J.-M.

(2014). Muscle strength and physical performance as predictors of mortality, hospitalization, and disability in the oldest old. Journal of the American Geriatrics Society, 62, 1030-1038. doi:10.1111/jgs.12840

23.

Lin

Chen

D.-G.

Vance

D. E.

Ball

K. K.

Mapstone

(2013). Longitudinal relationships between subjective fatigue, cognitive function, and everyday functioning in old age. International Psychogeriatrics, 25, 275-285. doi:10.1017/S1041610212001718

24.

Lindeberg

S. I.

Ostergren

P.-O.

Lindbladh

(2006). Exhaustion is differentiable from depression and anxiety: Evidence provided by the SF-36 vitality scale. Stress, 9, 117-123. doi:10.1080/10253890600823485

25.

Malmstrom

T. K.

Miller

D. K.

Morley

J. E.

(2014). A comparison of four frailty models. Journal of the American Geriatrics Society, 62, 721-726. doi:10.1111/jgs.12735

26.

Matsubayashi

(2011). 21-item fall risk index and frailty in the community-dwelling elderly. Igaku No Ayumi, 239, 445-449.

27.

Mirelman

Herman

Brozgol

Dorfman

Sprecher

Schweiger

. . . Hausdorff

J. M.

(2012). Executive function and falls in older adults: New findings from a five-year prospective study link fall risk to cognition. PLoS ONE, 7(6), e40297. doi:10.1371/journal.pone.0040297

28.

Morrison

Fan

Sen

S. S.

Weisenfluh

(2013). Epidemiology of falls and osteoporotic fractures: A systematic review. Clinicoeconomics and Outcomes Research, 5, 9-18. doi:10.2147/CEOR.S38721

29.

Olsson Möller

Midlöv

Kristensson

Ekdahl

Berglund

Jakobsson

. (2013). Prevalence and predictors of falls and dizziness in people younger and older than 80 years of age—A longitudinal cohort study. Archives of Gerontology and Geriatrics, 56, 160-168. doi:10.1016/j.archger.2012.08.013

30.

Otani

Takegami

Fukumori

Sekiguchi

Onishi

Yamazaki

. . . The LOHAS Research Group. (2012). Locomotor dysfunction and risk of cardiovascular disease, quality of life, and medical costs: Design of the Locomotive Syndrome and Health Outcome in Aizu Cohort Study (LOHAS) and baseline characteristics of the study population. Journal of Orthopaedic Science, 17, 261-271. doi:10.1007/s00776-012-0200-5

31.

Royston

(2000). A strategy for modelling the effect of a continuous covariate in medicine and epidemiology. Statistics in Medicine, 19, 1831-1847.

32.

Tinetti

M. E.

Williams

C. S.

(1997). Falls, injuries due to falls, and the risk of admission to a nursing home. The New England Journal of Medicine, 337, 1279-1284. doi:10.1056/NEJM199710303371806

33.

Tinetti

M. E.

Williams

T. F.

Mayewski

(1986). Fall Risk Index for elderly patients based on number of chronic disabilities. The American Journal of Medicine, 80, 429-434.

34.

Toba

Kikuchi

Iwata

Kozaki

(2009). “Fall Risk Index” helps clinicians identify high-risk individuals. Japan Medical Association Journal, 52, 237-242.

35.

Vestergaard

Nayfield

S. G.

Patel

K. V.

Eldadah

Cesari

Ferrucci

. . . Guralnik

J. M.

(2009). Fatigue in a representative population of older persons and its association with functional impairment, functional limitation, and disability. The Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 64, 76-82. doi:10.1093/gerona/gln017

36.

Whiting

Bagnall

A. M.

Sowden

A. J.

Cornell

J. E.

Mulrow

C. D.

Ramírez

(2001). Interventions for the treatment and management of chronic fatigue syndrome: A systematic review. Journal of the American Medical Association, 286, 1360-1368.

37.

Williams

J. S.

(2011). Assessing the suitability of fractional polynomial methods in health services research: A perspective on the categorization epidemic. Journal of Health Services Research & Policy, 16, 147-152. doi:10.1258/jhsrp.2010.010063

38.

Yamashita

Noe

D. A.

Bailer

A. J.

(2012). Risk factors of falls in community-dwelling older adults: Logistic regression tree analysis. The Gerontologist, 52, 822-832. doi:10.1093/geront/gns043

39.

D. S. F.

Lee

D. T. F.

Man

N. W.

(2010). Fatigue among older people: A review of the research literature. International Journal of Nursing Studies, 47, 216-228. doi.org/10.1016/j.ijnurstu.2009.05.009