Relationship between home environment and energy expenditure of community-dwelling older adults

Abstract

Introduction

Method

The home energy expenditure of 35 healthy older women was determined from retrospective geolocation data and a multi-sensor device measuring energy expenditure (SenseWear Armband®). Barriers at home were identified with the Housing Enabler.

Results

The median was 51 environmental barriers with only 7.5 barriers between the 1st and 3rd quartile, on a total of 161 possible environmental barriers of the Housing Enabler. The number of home environmental barriers was positively and moderately correlated with energy expenditure at home (r_s = 0.47, p = 0.01). No characteristic of the home built environment was identified that could explain the variability in energy expenditure.

Conclusion

Future research should identify the characteristics of the home associated with a lower or higher energy expenditure according to the characteristics of the person. This could be carried out by occupational therapists for the purpose of preventing deconditioning, energy management, promotion of social participation, recommendations for home adaptations or relocation.

Keywords

Environmental barriers accessibility housing aging daily activities occupational therapy

Introduction

High energy expenditure by healthy older individuals has numerous benefits, and housework and exercises done at home are among the most common physical activities. However, there is little knowledge of how characteristics of the home built environment could impact energy expenditure for moderate and vigorous daily activities. The decline in mobility and functional capacities associated with age (Wolinsky et al., 2011) may make it difficult for a significant number of older adults to remain at home (Werngren-Elgstrom et al., 2008); in fact, over 37% of Canadian seniors are limited in their activities (Joubert and Baraldi, 2016). Adopting a physically active life while aging is a key factor in the maintenance of physical capacities, independence and general health (Manini et al., 2006; Paterson and Warburton, 2010). Although the importance of a physically active life is recognized, studies show that between 40% and 60% of North American seniors are not active enough (Hall et al., 2014; Joubert and Baraldi, 2016). These percentages may be even higher as the amount of physical activity reported is typically overestimated by participants (Harvey et al., 2015; Leask et al., 2015). The “home” energy expenditure of healthy elderly people is not yet known, although it is recognized as a valuable variable in estimating the absolute amount of physical activity and its impact on health (Manini et al., 2006). Greater energy expenditure by older individuals is linked to numerous benefits, including a lower risk of premature mortality (Manini et al., 2006), developing chronic medical conditions (Bassuk and Manson, 2005) and mobility limitations (Paterson and Warburton, 2010). In participants aged 70–82 years recruited in 1998/1999, Manini et al. (2009) have shown that “greater energy expenditure from any and all physical activity was significantly associated with reduced risk of developing mobility limitation among men, but not among women.”

There is also little knowledge concerning home environment and physical activity, and how these could impact the energy expenditure of healthy women over 55 years old in realizing their daily activities. It is important for occupational therapists to understand how the home environment could impact the moderate and vigorous daily activities of older individuals.

Literature review

Physical activity can be divided into two categories: exercise (for example, sports) and daily activities. Here, daily activities would then correspond to the energy expended for everything we do that is not sleeping, eating or sports-like exercise (Levine and Kotz, 2005). Energy expenditure from physical activity associated with daily activities (such as doing laundry, getting around) varies from person to person (Donahoo et al., 2004) and is affected by life habits and the physical environment in which occupations are done (Levine and Kotz, 2005). The concept of energy expenditure is often measured by MET (Metabolic Equivalent of Task) (Ainsworth et al., 2011). MET is a unit of measure of the intensity of a physical activity, which corresponds to the ratio of the activity to the resting metabolic rate. It corresponds to sedentary behavior (1.0–1.5 METs), light-intensity (1.6–2.9 METs), moderate-intensity (3–5.9 METs) and vigorous-intensity (>6 MET) activities. For example, light-effort activities would be cleaning a sink and toilet, dusting and vacuuming; moderate-effort activities would be using an electric buffer, going upstairs, mopping, taking out the trash and vacuuming; and vigorous-effort would imply carrying groceries upstairs, moving or carrying furniture, or bicycling. The threshold for moderate or high/vigorous intensity is ≥3 METs. Bassuk and Manson (2005) recommend 0.5 hours per day of moderate-intensity exercise (MET ≥ 3) for achieving health benefits in people with chronic conditions.

Many authors have studied the impact of environmental variables on physical activity and energy expenditure in young and older individuals. At least two systematic reviews (McCormack and Shiell, 2011; Saelens and Handy, 2008) reported a significant positive correlation between physical activity and factors in the urban built environment, such as security, estheticism and the neighborhood built environment. In addition, older people who report various outdoor environmental barriers, such as poor lighting, poor roads and sidewalks, intrusive noise and no resting places are at greater risk of exacerbating the decline in their physical capacities and having difficulty walking long distances (Rantakokko et al., 2012).

The relationship between older individuals’ physical activity and different variables in the urban environment has been widely studied (Annear et al., 2014). Housework and exercises done at home are reportedly the most common physical activities done by older individuals (Lawlor et al., 2002). Daily tasks that require the most energy from older individuals include climbing stairs, vacuuming, grocery shopping and housework (Hall et al., 2014). Since many of these daily activities are done at home, the home environment may influence older individuals’ energy expenditure, especially if the person is restricted in their ability to go outside (for example during winter). That is why the home built environment is considered a determinant of the performance of daily activities (Wahl et al., 2009), of participation in daily activities and social roles (Levasseur et al., 2008; Oswald et al., 2007), and of the risk of falls among frail older individuals (Chase et al., 2012). Three studies (Chad et al., 2005; Cress et al., 2011; Csapo et al., 2009) showed that the level of physical activity is lower among older individuals living in seniors’ residences than their own homes, and these residences have less environmental challenges and distance to cover.

More knowledge is needed in the context of aging about the impacts of the characteristics of the home built environment (Wahl et al., 2009) and physical activity. What we know is that Benzinger et al. (2014) did not find any correlation between the number of physical barriers and older individuals’ physical activity at home. Harvey et al. (2015) also found low intensity activities and a small percentage of time spent on moderate and vigorous activities in older individuals’ sedentary behaviors at home.

Considering that high energy expenditure (moderate-intensity: 3–5.9 METs) by healthy older individuals has numerous benefits, that housework and exercises done at home are among the most common physical activities, and that certain characteristics of the urban built environment are associated with the level of people’s physical activity, it is plausible that there is a relationship between home environmental press and energy expenditure at home.

Conceptual framework and objectives

According to the ecological model of aging (Lawton and Nahemow, 1973), environmental press that is too high (such as physical barriers) for the older person’s competency level (for example their functional capacities or energy reserves) can lead to or risk creating disability. However, if people have a personal competency level that is high enough to respond to the pressures in their environment, they will approach their maximum performance potential. This model also emphasizes the risk for older individuals of under-stimulation in an environment. We hypothesized that greater environmental press (number of physical barriers) is positively correlated with energy expenditure at home among older individuals with a good level of personal competency. With the aim of developing knowledge about the main obstacles in the homes of healthy older individuals in a metropolitan area, we had three specific objectives: (1) Characterize accessibility and home environmental press (number of physical barriers) among healthy older women; (2) investigate the relationship between home environmental press and energy expenditure at home of healthy older women, and (3) identify the environmental characteristics that could explain the variability in energy expenditure in a sample of healthy older women.

Method

Research design

A descriptive/correlational cross-sectional design was used to address the above-mentioned research objectives. Variables were collected in three phases. Firstly, the energy expenditure and mobility variables of healthy older women and men were measured during the summers of 2011 and 2012 by Blamoutier et al. (2017). Secondly, the energy expenditure and mobility variables of healthy older women and dynapenic women were measured during summer of 2014 by Blamoutier et al. (2017). Dynapenia is the age-associated loss of muscle strength that is not caused by neurologic or muscular diseases (Clark and Manini, 2012). Non-dynapenic woman have grip strength over 20kg (Sallinen et al., 2010). Thirdly, variables characterizing the home built environment and accessibility were gathered in summer 2015 at the same participants’ homes in the Montreal area, by author AP. These data collections were subprojects of the Ecological Mobility in Aging and Parkinson’s (EMAP) project (Duval et al., 2017). The EMAP project, a 5-year project, aims to address populations with dyskinesia as well as “healthy elders” to develop norms and assessments in mobility when aging at home.

In the summers of 2011 to 2014, data collection was spread over 14 days, during which participants wore two devices: one device measuring energy expenditure and the other device recording the participant’s geolocation. On the first visit, the devices were handed out, instructions were given, and the consent form and questionnaires were completed. A second visit took place one week later to ensure that the assessment tools were working properly, address any potential problems associated with wearing the devices, and repeat the instructions. Finally, the participants went to the laboratory for a third meeting to return the devices and do the body composition measures. To control for weather conditions, data collection took place between May and October when there was no snow.

Author AP asked the participants in summer 2015 to read and sign the consent form for a new project (relationship between their energy expenditure and home environment); it was approved by Université Laval Research Ethics Committee (# 2015-133/29-05-2015).

Participants and recruitment

In summer 2015, participants who had agreed to be re-contacted from a previous study were recruited again by the first author (AP), to constitute a convenient sample. The original target population was older women (55 years and over) living at home on the Island of Montreal. Participants had been recruited (summer 2011 to 2014) from the bank of participants of the University Research Centre at the Institut de Gériatrie de Montréal (URCIG), by word-of-mouth and through posters at community and sports centers and residential complexes by the fourth author (MB).

In order to recruit the most participants possible from the previous studies who had agreed to be re-contacted, the authors made the choice to include the participants from the two previous studies that had collected data of energy expenditure and mobility variables using the same protocol. Considering the use of two convenience samples having different populations, we selected only healthy women in order to control for health and gender as we needed a homogeneous sample to be able to properly interpret of our results.

The inclusion criteria established by Blamoutier et al. (2017) were: (1) have normal cognitive functioning according to the Montreal Cognitive Assessment (MoCA) (assessed by 4th author MB) (≥26/30) (Nasreddine et al., 2005); (2) not be diagnosed with any illness in the bank of participants of the URCIG or self-reported; (3) live independently in a dwelling in the community on the Island of Montreal and (4) have a body mass index (BMI) below 30. Thus, the degree of personal competency was controlled by deliberately selecting people in good health. Three new inclusion criteria were added by AP, to ensure uniformity in the sample of older individuals in good health and limit biases associated with using retrospective data: (5) be a non-dynapenic woman (grip strength over 20 kg, assessed with a dynameter by the 4th author) (Sallinen et al., 2010); (6) be living in the same home as when they participated in the previous study conducted between 2011 and 2015; and (7) not have had any major work done in the home since they participated in the previous study. For example, major work would include removing walls, adding a ramp, or replacing the bathtub with a non-threshold shower. Minor work would be the use of light assistive technology like a rolling tray to carry dishes and long handle pliers, adding lighting, eliminating carpets and moving furniture. This ensured that the measures taken at the home reflected the living environment. AP contacted participants by phone to explain the new study, verify the inclusion criteria and set up a meeting at their home. Of the 50 potential participants, 35 met the inclusion criteria and completed the assessments. The reasons for exclusion were: had moved (n=6), refused (n=3), could not be contacted (n=3), major work done in the home (n=2) and medical complication (n=1).

Procedure

AP is an occupational therapist with experience in home intervention for older individuals. He also managed a systematic review on tools assessing home built environment during his master in sciences (Patry et al., 2019). Between June and September 2015, he collected data at the participants’ homes; visits lasted 90 to 120 minutes. This was done up to four years after the start of data collection of the previous study.

Assessment and measures used

Physical barriers and functional limitations

Physical barriers and functional limitations were assessed with the Housing Enabler (2nd version) (Iwarsson and Slaug, 2001). The Housing Enabler combines a short interview with detailed observation of the home physical environment and takes about one hour to complete. The assessment tool is a user-friendly piece of software and comprises three sections. Section 1 (personal components) is a 14-item dichotomic grid completed via the interview and direct observation to identify the person’s functional limitations and dependence on mobility devices. For example, it could be presence or absence in Difficulty interpreting information, Severe loss of sight or Incoordination. Section 2 (environmental components) consists of a comprehensive observation of the home physical environment done with a checklist grid comprising 161 physical barriers (outdoor, entrances, indoor). Figure 1 is a screen capture of the “indoor” grid. Section 3 (calculation of accessibility score) addresses calculation of the person–environment fit score (P-E Fit score), using a matrix where the profile of functional limitations (section 1) is juxtaposed with the environmental barriers identified (section 2), resulting in an accessibility score for the dwelling for the person. The P-E Fit score is quantified on a predefined ordinal scale from 0 to 4 based on the relationship between the person’s functional limitations and the home environment barriers, and produces a total accessibility score ranging from 0 to 1832. Score 1832 is a theoretical maximum, to the extent that the person should have all the functional limitations and all possible 161 barriers. A higher P-E Fit score suggests greater accessibility limitations anticipated in the participant’s environment. A person without functional limitations will automatically be assigned a P-E Fit score of 0, meaning no accessibility problems. Environmental barriers will be prioritized or displayed in order of involvement in the calculation of P-E Fit. The Housing Enabler is one of the few comprehensive tools for assessing the home built environment whose psychometric properties have been evaluated, and that also has an electronic version adapted for research. Fänge et al. (2007), Helle et al. (2014) and Iwarsson et al. (2005) point to moderate interrater reliability for the Housing Enabler.

Figure 1.

Screen capture of Housing Enabler (2nd version) – indoors checking example.

To verify if there were some aspects of the environment that might better explain the variability in energy expenditure, categories of the built environment and physical barriers were selected as data complementary to the Housing Enabler. Using a semi-structured interview and observation grid, the following characteristics were documented: type of dwelling; type of access; number of rooms; presence of a long corridor; and number of indoor and outdoor steps. Each participant was interviewed once, and the information was immediately entered into an Excel file.

At-home energy expenditure

At-home energy expenditure was determined from the fusion of GPS receiver and energy expenditure tracker data worn by the participants. First, a GPS receiver was used to identify periods when the person was at home. A GPS receiver recalculates its position as an expression of latitude, longitude and altitude from signals from a constellation of satellites. The time difference between the arrival of the satellites’ signals to the receiver and their known positions in orbit at that specific moment is used to determine the geographic location of the receiver using trigonometry calculations. Two GPS receiver models were used in the EMAP study: the Qstarz GPS Travel Recorder XT and the WIMUGPS (Wireless Inertial Measurement Unit with GPS) (Boissy et al., 2011). Participants were given the GPS device (Qstarz or WIMUGPS) and instructed to (a) wear it at the waist, antenna toward the outside, (b) wear it continuously during waking hours and (c) charge it each night. The recorded GPS data was processed using Matlab (version 14). Raw data was filtered and interpolated to obtain a continuous chronological series. Temporal analysis of the series was done to group the GPS data into “clusters,” that is, areas in which the device’s wearer spent continuous time over the period of a day. The home cluster was defined as the cluster where the person spent the most time during each of the days. It was confirmed that the clusters did in fact correspond to the home by comparing the distance between the cluster’s geographical data and the participant’s exact address using Google Maps®. All clusters within a radius of more than 150 meters from the dwelling’s coordinates were removed from the analysis of the results.

Energy expenditure was measured with the SenseWear Armband® (BodyMedia Inc., Pittsburgh, PA). This portable device uses a two-axis accelerometer, skin temperature sensor, heat flux sensor and ambient temperature sensor. This instrument showed good criterion validity for estimating energy expenditure compared to a criterion method, doubly labeled water (St-Onge et al., 2007). The participants wore the armband on the triceps on their right arm day and night, except during baths or showers, for 14 consecutive days. Using the participant’s personal and anthropometric data (age, sex, weight, height, smoker), the SenseWear Armband® measures energy expenditure (kilocalories) in real time. Other information provided by the armband is minute-by-minute intensity level or MET and period of time for which intensity was moderate or high (≥3 METs).

All the time periods spent at home that were identified by the GPS data were compiled and rounded to the minute. Data points corresponding to when the participant was at home were time stamped and used to segment at-home energy expenditure data. To address objective 2, total energy expenditure, time spent on moderate or vigorous physical activity (≥3 METs) and energy expenditure associated with moderate or vigorous physical activity were compiled for these time periods (rounded to the minute) using SenseWear 8.1 software. Thus, each minute spent at home was linked to measures of physical activity (energy expenditure, intensity, and time spent on moderate or vigorous physical activity) and averaged over the recording period.

Data analysis

Descriptive analyses (mean, standard deviation, frequency) were done for the participants (age, BMI, education, marital status), and also for the characteristics of the built environment (frequency, median, quartiles), the main physical barriers encountered (frequency, score), and Housing Enabler scores (mean, standard deviation). To address objective 1, characterization of accessibility and home environmental press (extracting number and types of physical barriers) among healthy older women was compiled with the Housing Enabler software.

Descriptive analyses of energy expenditure were done (mean, standard deviation). To obtain an approximation of the level of intensity of the activities when participants were at home, the approximate mean MET was calculated with the formula: METs = (kcal/min × 200)/(body weight (kg) × 3.5) (Humphrey, 2006), using the participants’ mean weight and energy expenditure per minute.

To address objective 2, the relationship between energy expenditure variables at home and the characteristics and obstacles of the home environment were analyzed using Spearman’s correlation in a bivariate analysis. Two energy expenditure variables were used for the analysis: mean energy expenditure at home (kcal/min) and mean energy expenditure in moderate or vigorous physical activity at home (≥3.0 METs). To verify our main hypothesis, namely that there is a correlation between the number of physical barriers at home and energy expenditure at home, the correlations had to have a p < 0.05 to be considered significant. The correlation between the participants’ energy expenditure, age and BMI was analyzed using Pearson’s correlation. Partial correlations with energy expenditure measures and environmental characteristics were used to control for the participants’ age.

To address objective 3 (to identify if physical barriers were associated with the variability in mean energy expenditure at home), the six environmental barriers in the participants’ homes generating the highest P-E Fit score were selected. The Long corridor (indoor corridor ≥ 6 meters) item was also added as a variable of interest (not because of a physical barrier but because it might better explain the variability in energy expenditure). The Mann–Whitney non-parametric test was used to verify if there was a statistically significant difference in mean energy expenditure at home depending on whether the environmental barrier was present versus absent, using the Bonferroni correction in a bivariate analysis (threshold p ≤ 0.007). Descriptive and statistical analyses were done with SPSS [IBM SPSS Statistics 23.0].

Results

Participant profile

After applying the inclusion and exclusion criteria, 35 healthy women between 55 and 78 years old with a mean age of 66 years were assessed (Table 1). The majority had studied at university (71%) and 26% were married or in a common-law relationship. The mean BMI was 24.1 ± 2.8, which is within norms. Limitations in stamina (11%) and reduced spine and/or lower extremity function (9%) were the two most prevalent functional limitations in our sample. Only 17% of the women reported falls related to the home environment, and 11% had made minor adaptations to address those issues (for example installing a handrail or non-slip strips).

Table 1.

Participant profile of healthy senior non-dynapenic women and their dwellings (n=35).

Variables
No. of women	35
Mean age ‒ years (SD)	66 (4.69)
Mean body mass index ‒ kg/m² (SD)	24.1 (2.8)
Mean energy expenditure (EE) ‒ kcal/min (SD)	1.67 (0.31) n=27
Mean total minutes of EE measured at home (SD)	3453.5 (1982.3) n=27
Education
High school ‒ n (%)	7 (20)
College/certificate ‒ n (%)	3 (9)
Bachelor’s ‒ n (%)	11 (31)
Master’s/doctorate ‒ n (%)	14 (40)
Marital status
Married/common-law ‒ n (%)	9 (26)
Divorced/widowed ‒ n (%)	14 (40)
Single ‒ n (%)	12 (34)
Functional limitations¹
None	28
Limitations in stamina ‒ n (%)	4 (11)
Reduced spine and/or lower extremity function ‒ n (%)	3 (9)
Loss of upper extremity function ‒ n (%)	2 (6)
Difficulty in moving head ‒ n (%)	2 (6)
Visual impairment ‒ n (%)	2 (6)
Reduced upper extremity function ‒ n (%)	1 (3)
Loss of hearing ‒ n (%)	1 (3)
Reported falls at home ‒ n (%)	6 (17)
Home adaptations to prevent falls (handrail, visual indicator, anti-slip strips, use of alternate access)	4 (11)
Dwelling
Apartment/duplex/triplex ‒ n (%)	19 (54)
Residential complex ‒ n (%)	9 (26)
House ‒ n (%)	7 (20)
Characteristics of dwellings²
No. of indoor steps ‒ median (q1–q3)	14 (2–25.75)
No. of environmental barriers ‒ median (q1–q3)³	51 (47–54.5)
No. of rooms⁴ ‒ median (q1–q3)	4 (3.5–5)
No. of outdoor steps ‒ median (q1–q3)	4 (1.25–8.75)
No. of floors ‒ median (q1–q3)	1 (1–1)
Access to main rooms
Regular staircase ‒ n (%)	18 (51)
Half-spiral staircase ‒ n (%)	7 (20)
Door step ‒ n (%)⁵	4 (11)

¹Number of participants with one of 14 functional limitations and mobility dependences according to the Housing Enabler (2nd version).²(q1-q3) represents interquartile range.³Number of environmental barriers in dwelling (outdoor, entrances, indoor) measured with the Housing Enabler (2nd version); range 0–161 (out of 161 possible environmental barriers).⁴Bathroom and shed excluded.⁵2 steps or less.

Home built environment

In Table 1, the participant profile includes general dwelling information. Most participants were living in an apartment (54%), 26% in a multi-habitation residential complex and 20% in a house. The median number of environmental barriers was 51, with only 7.5 barriers between the first and the third quartile, from a total of the 161 possible environmental barriers of the Housing Enabler. A median of 14 indoor steps was observed, with a wide variance (2–25.75 steps), compared to smaller medians for number of rooms (4), outdoor steps (4), and floors or levels (1). Finally, 51% of the dwellings were accessed via regular staircases, 20% via half-spiral staircases and 11% via doorsteps. Table 1 includes a profile of the participants’ dwellings.

Table 2 shows the 10 environmental barriers in healthy women’s dwellings generating the highest P-E Fit scores, their frequency and mean P-E Fit score. Stairs as the only access had the highest P-E Fit score (47) and was present in 80% of the participants’ dwellings; high curbs scored 16 and occurred 71% of the time, while irregular/uneven surfaces scored 13 and were observed for all participants. Concerning the women’s profile for accessibility, 28 participants had no functional limitation (none) and seven participants had 1–4 types of functional limitations (Table 1). Participants with no functional limitation were automatically given a P-E Fit score of 0. This is why the mean P-E Fit score for our sample of healthy older women was very low (28 scored “none”), namely 12.3 (SD ± 32.2), as shown in Table 2, indicating few accessibility problems.

Table 2.

Environmental barriers and person–environment fit scores (n=35).

Environmental barriers: total and their frequency (%)
Stairs the only route (no lift/ramp)	47 (80)
High curbs (more than 4cm)	16 (71)
Doors that do no stay in open position/ close quickly	15 (37)
High, low and/or irregular heights of risers	15 (57)
Irregular/uneven surface	13 (100)
Letterbox difficult to reach	13 (49)
Heavy doors without automatic opening	13 (34)
Stair treads with narrow depth (less than 26cm) or irregular depth	12 (23)
Wall-mounted cupboards and shelves placed high	12 (34)
No grab bar at shower/bath and/or toilet	12 (71)
Person–environment fit score^1,2
Mean person–environment fit score (SD)	12.3 (32.2)³

¹Person–environment Fit (P-E Fit) is quantified on a predefined ordinal scale from 0 to 4 points based on the relationship between a person and his/her home environment, and produces a total accessibility score ranging from 0 to 1832, with higher scores denoting more accessibility problems. No functional limitation for a person, P-E Fit=0.

²Person–environment fit score according to the Housing Enabler is very low (higher scores indicate greater person–environment ﬁt problems); the range of functional limitations was 0–1832.

³The result here is the 10 environmental barriers generating the highest person–environment fit scores. As 28 women had P-E Fit score 0 (see Table 1), it justifies such a low score but with a large standard deviation.

Relationship between home built environment and energy expenditure

Participants who did not have at least 600 minutes of energy expenditure data at home were excluded from the analyses of the results. Unfortunately, we therefore could not use data for 8 of the 35 participants. Reasons for not having minimum data were discomfort with the SenseWear Armband® (n=5), a technical problem (n=2) or not having spent enough time at home (n=1).

For the 27 participants for whom we had enough data on their energy expenditure at home, the mean energy expenditure was 1.67 kilocalories per minute, SD ± 0.31. The energy expenditure data collected at home for each participant varied between 655 and 9893 minutes. There was a noteworthy variability in total energy expenditure time measured at home (mean: 3453 ± 1982 minutes). The approximate mean MET was 1.53, which corresponds to light activities (range from 1.5 to 3 METs) (Ainsworth et al., 2011). Only 5.8% of the total energy expenditure time measured at home involved moderate or vigorous physical activity (≥3.0 METs). The number of environmental barriers at home was significantly correlated with participants’ energy expenditure at home (r_s = 0.469, p = 0.014) (see Table 3). The relationship between participants’ age and energy expenditure was not significant according to bivariate analyses. There was no relationship or trend between participants’ BMI and energy expenditure at home. Among the barriers measured, neither the number of rooms nor the number of outdoor and indoor steps was significantly correlated with energy expenditure at home. No significant relationship was found between mean daily energy expenditure in moderate or vigorous physical activity at home (≥3.0 METs) and the number of environmental barriers, rooms, or outdoor and indoor steps. Participants’ age and BMI were not significantly correlated with mean daily energy expenditure in physical activity at home. When controlling for age, the correlation between mean energy expenditure at home and the number of environmental barriers was significant (r_s = 0.485, p = 0.012). The number of rooms and number of outdoor and indoor steps were not significantly correlated with either energy expenditure variable after controlling for age.

Table 3.

Correlation between mean energy expenditure and number of environmental barriers in home built environment and characteristics of the home and participants (n=27).¹

	Energy expenditure (EE) in kcal/min
	EE at home²	EE in physical activity³	R ^4,5
Variables
Environmental barriers⁶	0.469 (0.014)*	0.308 (0.118)	Spearman
Number of rooms	0.266 (0.180)	0.288 (0.145)	Spearman
Number of outdoor steps	0.166 (0.409)	−0.247 (0.215)	Spearman
Number of indoor steps	0.213 (0.286)	0.081 (0.687)	Spearman
Age	−0.335 (0.088)	−0.265 (0.181)	Pearson
Body mass index (BMI)	0.209 (0.294)	−0.321 (0.103)	Pearson
Variables – controlling for age
Environmental barriers¹	0.485 (0.012)*	0.299 (0.138)	Spearman
Number of rooms	0.226 (0.268)	0.256 (0.207)	Spearman
Number of outdoor steps	0.260 (0.199)	−0.200 (0.327)	Spearman
Number of indoor steps	0.112 (0.586)	−0.010 (0.962)	Spearman

¹n=27: participants who did not have at least 600 minutes of energy expenditure data at home were excluded from the analyses of the results. This requirement was not met for 8 of the 35 participants (due to discomfort with the SenseWear Armband® n=5, technical problem n=2, not enough time at home n=1).

²EE at home: mean energy expenditure at home (kcal/min). ≥ 3.0 METs * p value < 0.05.

³EE in physical activity: mean energy expenditure in moderate or vigorous physical activity at home (≥ 3.0 METs).

⁴Spearman’s rho (r_s) was used for correlation between energy expenditure and environmental barriers; p value in brackets (2-tailed).

⁵Pearson’s rho (r) was used for correlation between energy expenditure and participants’ characteristics; p value in brackets (2-tailed).

⁶Number of environmental barriers in dwelling (outdoor, entrances, indoor) measured with the Housing Enabler (2nd version), range 0–161.

Association between environmental barriers and energy expenditure at home

Table 4 compares the mean energy expenditure at home between participants for whom a particular environmental barrier was present and participants for whom it was absent. No significant difference was found between the groups. The mean rank of energy expenditure at home was higher in the presence of six of the seven environmental barriers.

Table 4.

Comparison of mean rank of energy expenditure at home of two groups: presence or absence of environmental barrier (n=27).

	Groups
	Prevalence (n)	Mean rank (presence)	Mean rank (absence)	p value¹
Stairs the only route	22	8.8	15.2	0.11
High curbs	18	13.2	14.4	0.74
Doors that do not stay in open position	10	13.4	15.0	0.64
High, low or irregular height of risers	16	15.1	13.3	0.58
Letterbox difficult to reach	15	11.0	16.4	0.08
Heavy doors without automatic openings	9	12.7	16.6	0.25
Presence of a long corridor²	14	12.2	15.6	0.28

¹Mann–Whitney U-test, using Bonferonni correction in bivariate analysis and threshold p value < 0.007.

²Length of corridor inside dwelling over 6 meters.

Discussion and implications

In this study, the relationship between energy expenditure at home and home environmental press (number of physical barriers) of healthy older women was investigated. This study confirmed our hypothesis that there is a statistically significant positive, albeit moderate correlation between home environmental press and energy expenditure at home. This positive relationship between the two variables conforms to the ecological model of aging (Lawton and Nahemow, 1973), according to which, if people have sufficient personal competency to respond to the pressures in their environment, they will approach their maximum performance potential. Our findings indicate that homes with more physical barriers result in increased energy expenditure but not more moderate or vigorous activity. The level of low intensity activities overall and small percentage of time spent on moderate and vigorous activities was predictable, given older individuals’ sedentary behaviors at home (Harvey et al., 2015). These data suggest that the majority of the time recorded at home was spent on daily activities and housework.

We did not find an association between any of the physical characteristics and a decrease or increase in physical activity. Our sample size and the low prevalence of some physical barriers limited our ability to detect a potential relationship. The portrait of home accessibility generated must be interpreted in light of the fact that the majority of participants did not have any functional limitations.

Thus, it appears that total physical barriers are a better indicator of energy expenditure at home than each of the main barriers considered separately. Although no barrier was individually associated with energy expenditure at home, the correlation between energy expenditure at home and the number of barriers at home was moderate and statistically significant. One possible explanation is that some physical barriers make a small contribution to increasing energy expenditure among healthy older women living at home, but that it is the interaction between these barriers that is the factor of interest associated with the variability in older women’s energy expenditure at home. However, because of the descriptive/correlational design used in this cross-sectional study, we could not determine which physical barriers caused the increase in energy expenditure. Our results differ from those of Benzinger et al. (2014), who did not find any correlation between the number of physical barriers and older individuals’ physical activity at home. However, this difference could be attributable to their use of a questionnaire to assess physical activity, which does not measure in real time the potential impact of the environment when executing activities. Our results complement the evidence from three studies (Chad et al., 2005; Cress et al., 2011; Csapo et al., 2009) conducted with people living in seniors’ residences where there were fewer environmental challenges and less activity and muscle strength than among seniors living at home. Although these studies suggest that a premature relocation to a living environment with fewer challenges could contribute to the deconditioning of older individuals, to our knowledge, no study has been able to determine which environmental characteristics cause a decrease in physical activity. This will be discussed in future research.

Strengths and limitations

This is the first study to link objective measures of the physical characteristics of the home environment with objective measures of energy expenditure at home. The variables stem from the ecological model of aging that has been used in numerous studies in gerontology and is one of the few models that take into account the positive and negative effects of the challenges of the home built environment on older people. The combined use of the SenseWear Armband® and a GPS device provided data on energy expenditure at home in an ecological context. Some authors support the use of objective energy expenditure measures in research to increase the validity of the results (Hall et al., 2014; Harvey et al., 2015). The Housing Enabler produces a reliable and objective assessment of the home built environment and accessibility. To increase generalization and the validity of the results in the context of the home environment and person–environment fit, it is necessary to use tools with proven psychometric properties (Wahl et al., 2009). Our study thus supports the importance of using objective measures when assessing the interaction between variables in the physical environment and mobility.

Our study has some limitations. The main limitation is the loss of data during collection, primarily related to wearing the SenseWear Armband® (skin irritation, discomfort, technical problems) when our sample size was already quite small. Despite this (n=27 with complete data for energy expenditure instead of n=35), we achieved satisfactory heterogeneity in terms of the type of dwelling and neighborhood in an urban setting with a target population of healthy older women. We expect these heterogeneous data may be transferable to Montreal area, but there is no available norm for healthy women of 55 to 78 years old living there. A second limitation is that, to our knowledge, there are no assessment tools that specifically measure the impact of the home built environment on energy expenditure or physical activity. As mentioned above, various reasons justified using the Housing Enabler, but when the results are interpreted it must be taken into account that not all the physical barriers identified had a presumed impact on energy expenditure (for example switch too high). Third, the use of convenience samples, with the time lag between measures of energy expenditure and physical characteristics, could affect the validity of the results. However, the authors controlled for changes in the environment through the exclusion criteria. Fourth, with the selection of healthy older women as the population within the convenience samples, the measure of the P-E Fit was not as clinically significant as with frail elderly, for example. The interpretation of the results concerning the impact of the home environmental barriers on accessibility at home of older individuals was weakened by the choice of the inclusion criteria. On the other hand, a low P-E Fit confirms Lawton’s model, which demonstrates that healthy elderly people, in our case non-dynapenic woman, are less likely to experience environmental pressure from their homes.

Implications for clinical practice

The aim of health professionals is to promote their clients’ health. As experts in the relationship between person, occupation and environment, occupational therapists are responsible for determining the impacts of the environment on their clients’ occupational participation (CAOT, 2012). In this respect, our study helps to develop knowledge on how a change in the home built environment influences clients’ energy expenditure. However, some points must be made concerning the interpretation of the results in energy expenditure and built home environment. An increase in energy expenditure is not necessarily beneficial for everyone in all situations. In fact, energy management is a significant component for rehabilitation and homecare clinicians. For people with mobility or endurance limitations (such as multiple sclerosis or chronic obstructive pulmonary disease), an increase in energy expenditure in their daily activities associated with environmental barriers could be synonymous with a decrease in occupational participation (Clarke et al., 2011; Pho et al., 2012). This leads occupational therapists to consider the potential impacts of an intervention (such as a home adaptation) or recommendations (for example relocation) on their clients’ energy expenditure.

In occupational therapy reasoning, there are factors that need to be considered when making home modifications (Bridge, 2010; Fagan and Sabata, 2016; Millikan, 2012; Russell et al., 2018; Stark et al., 2009) or recommendations for relocation (Millikan, 2012). These authors suggest documenting the home environment (type of building, access, bedroom, lounge, toilet, laundry, kitchen, dining, bathroom, back yard) and functional performance in all daily activities (independent, nil issue, difficulty, pain, unable, with human assistance, with assistive technologies).

When undertaking their assessment, occupational therapy home assessors should consider the physical features (such as access) and sub-features (such as street frontage, driveway front access, back access, side access, internal access), characteristics (such as gutter/curbing, stairs/steps, ramp, driveway, lift/elevator, lighting) and sub-characteristics relevant to their client (for example material such as concrete, wood, tiles; slope/gradient; ownership such as torrens/strata/council), and the impact on the client (for example on their mobility) (Millikan, 2012).

This demonstrates the level of analysis that the occupational therapist should consider before making a recommendation for all physical features. This is the realm of the clinical reasoning process for occupational therapists, and where the intersection between form/protocol, function and context occurs. Careful examination and analysis of the client, their abilities (including energy expenditure) and their environment is crucial before making recommendations about the retrofit or modification of a person’s environment. A list of problems or primary issues identified should be outlined for all physical features in parallel to a list of recommendations. For example, in Millikan (2012: 38) one of the primary issues identified is a primary access, because the client is unable to use it when he’s not well (multiple sclerosis). There is no access for emergency egress/emergency services. And, finally, the client is at high risk of falls at any time using this access point. One of four recommendations formulated for access is the installation of a suitable access point from the street level to the house level, followed by a justification. In some cases the relevant body may decide that the cost of modifications is untenable and will recommend that the client relocate to a more suitable property (Millikan, 2012). Russell et al. (2018, Conclusion section) have stated that the home modification process protocol potentially should:

Provide a systematic approach to the process of modifying the home;

Ensure that ethical and professional practice is followed by enabling occupational therapists to verbalize and visualize their role in the process, and reduce the complexity of the current process by identifying the key questions, actions and outcome of each phase;

Improve the effectiveness and efficiency of practice by ensuring that practitioners collect the right information, at the right time;

Ensure that the person has choice and control through their involvement in all phases of the process;

Guide professional reasoning based on a conceptual model of practice;

Ensure consistency of occupational therapy practice by accommodating regional, legislative and regulatory differences between practice settings;

Ensure that financial constraints and other contextual issues within practice become a design consideration and not a barrier for accessing funding for a modification.

In summary, there should be a balance between benefits of energy expenditure and risks in the home environment, but this finding requires more research, and also individualized and deep clinical reasoning, when making recommendations for older individuals.

Implications for future research

It would be interesting to document the effects of an occupational therapy intervention, such as a home adaptation, on energy expenditure in a cohort study (tracking older individuals in time). We found a significant association in a sample of healthy older individuals, but the impacts of the built environment and home modifications could be clinically more significant for people with mobility problems. Indeed, the ecological model of aging suggests that people with less personal competency to handle the pressures of the built environment are more sensitive to changes in their environment. These aspects of this exploratory study constitute the main elements of interest for clinicians and researchers. Specifically, if we can identify the characteristics of the home built environment associated with lower or higher energy expenditure according to the person’s characteristics, there are many potential implications for homecare and rehabilitation clinicians, such as the prevention of deconditioning, energy management, fostering social participation, and recommendations for home adaptations or relocation.

Conclusion

The number of physical barriers at home is positively correlated with energy expenditure at home among healthy older women. Based on our results, it is reasonable to hypothesize that changes in the built environment of older people’s homes could have an impact on their energy expenditure. However, we were unable to identify any physical barrier that could explain the variability in older individuals’ energy expenditure. To establish a possible causal link, experimental studies with a longitudinal design are required, including with populations with mobility problems.

Key findings

The more physical barriers the home contains, the higher the energy expenditure among healthy older people (for example stairs the only route, high curbs, doors that do not stay in open position).

Total physical barriers could be a better indicator of energy expenditure at home than each of the barriers considered separately.

What the study has added

This study has contributed to knowledge concerning how the home built environment could impact the energy expenditure of healthy older individuals: total physical barriers seem a better indicator than one in particular.

Footnotes

Acknowledgments

We would like to thank Catherine Lavigne-Pelletier for her precious help for the phone recruitment and support with previous data regarding the armband (Ecological Mobility in Aging and Parkinson’s (EMAP) project). We also would like to thank Jean Leblond, the statistician at the CIRRIS, for all data analysis. Finally, thank you to Nicolas Robitaille for technical support regarding the use of Matlab at the CIRRIS.

Research ethics

Ethical approval was obtained from Université Laval Research Ethics Committee (# 2015-133/29-05-2015).

Consent

All participants provided written informed consent to be interviewed for the study.

Declaration of conflicting interests

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

Funding

The authors disclosed receipt of the following financial support for the research. The first author (AP) was supported by a student grant from the Canadian Institutes of Health Research (CIHR-AMG-100485) and the Centre for Interdisciplinary Research in Rehabilitation and Social Integration (CIRRIS) located in Quebec City, Canada. The Principal Investigator of the CIHR grant is the third author (CD).

Contributorship

Alexandre Patry researched literature, recommended the conceptual framework for the study, applied for ethical approval, and contributed to the development of the data. Christian Duval, Claude Vincent, Patrick Boissy and Margaux Blamoutier contributed to the methodology of the project and the primary statistical analysis plan. Alexandre Patry carried out the final statistical analysis with the support of Simon Brière, and all authors interpreted the data. Alexandre Patry wrote the first draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version.

References

Ainsworth

Haskell

Herrmann

et al . (2011) 2011 compendium of physical activities: A second update of codes and MET values. Medicine and Science in Sports and Exercise 43(8): 1575–1581.

Annear

Keeling

Wilkinson

et al . (2014) Environmental influences on healthy and active ageing: A systematic review. Ageing and Society 34(4): 590–622.

Bassuk

Manson

(2005) Epidemiological evidence for the role of physical activity in reducing risk of type 2 diabetes and cardiovascular disease. Journal of Applied Physiology 99(3): 1193–1204.

Benzinger

Iwarsson

Kroog

et al . (2014) The association between the home environment and physical activity in community-dwelling older adults. Aging Clinical and Experimental Research 26(4): 377–385.

Blamoutier

Boissy

Brière

et al . (2017) Is a decrease of grip strength associated with community mobility restriction in dynapenic older women? Journal of Cachexia, Sarcopenia and Muscle-Clinical Reports 3(2): 1–9.

Boissy

Brière

Hamel

et al . (2011) Wireless inertial measurement unit with GPS (WIMU-GPS)—Wearable monitoring platform for ecological assessment of lifespace and mobility in aging and disease. In: Engineering in Medicine and Biology Society, EMBC, 2011 Annual International Conference of the IEEE. IEEE, pp.5815–5819. Available at : https://ieeexplore.ieee.org/document/6091439/authors#authors

Bridge

(2010) Home modification: Occupation as the basis for an effective practice. In: Curtin M, Molineux M and Supyk-Mellson J (eds) Occupational Therapy and Physical Dysfunction. Enabling Occupation, 6th ed. pp.409–429. Edinburgh: Churchill Livingstone/Elsevier.

Canadian Association of Occupational Therapists (2012) Profile of occupational therapy practice in Canada. Ottawa, ON: CAOT, 1–33. Availabe at : https://www.caot.ca/document/3653/2012otprofile.pdf

Chad

Reeder

Harrison

et al . (2005) Profile of physical activity levels in community-dwelling older adults. Medicine and Science in Sports and Exercise 37(10): 1774–1784.

10.

Chase

Mann

Wasek

et al . (2012) Systematic review of the effect of home modification and fall prevention programs on falls and the performance of community-dwelling older adults. American Journal of Occupational Therapy 66(3): 284–291.

11.

Clark

Manini

(2012) What is dynapenia? Nutrition 28(5): 495–503.

12.

Clarke

Ailshire

Nieuwenhuijsen

et al . (2011) Participation among adults with disability: The role of the urban environment. Social Science & Medicine 72(10): 1674–1684.

13.

Cress

Orini

Kinsler

(2011) Living environment and mobility of older adults. Gerontology 57(3): 287–294.

14.

Csapo

Gormasz

Baron

(2009) Functional performance in community-dwelling and institutionalized elderly women. Wiener klinische Wochenschrift 121(11–12): 383–390.

15.

Donahoo

Levine

Melanson

(2004) Variability in energy expenditure and its components. Current Opinion in Clinical Nutrition & Metabolic Care 7(6): 599–605.

16.

Duval

Speechley

Boissy

et al . (2017) Ecological mobility in aging and Parkinson’s. Available at: http://mobilityinaging.com/ (accessed 16 January 2017).

17.

Fagan

Sabata

(2016) Home modifications and occupational therapy – Fact sheet developed for the American Occupational Therapy Association. Available at: www.aota.org/∼/media/Corporate/Files/AboutOT/Professionals/WhatIsOT/RDP/Facts/HomeMod-Occ-Therapy.pdf (accessed 10 January 2019).

18.

Fänge

Risser

Iwarsson

(2007) Challenges in implementation of research methodology into community-based occupational therapy—The Housing Enabler example. Scandinavian Journal of Occupational Therapy 14(1): 54–62.

19.

Hall

Morey

Dutta

et al . (2014) Activity-related energy expenditure in older adults: A call for more research. Medicine and Science in Sports and Exercise 46(12): 2335–2340.

20.

Harvey

Chastin

Skelton

(2015) How sedentary are older people? A systematic review of the amount of sedentary behavior. Journal of Aging and Physical Activity 23(3): 471–487.

21.

Helle

Nygren

Slaug

et al . (2014) The Nordic Housing Enabler: Inter-rater reliability in cross-Nordic occupational therapy practice: Previously published in Scandinavian Journal of Occupational Therapy 2010; 17: 258–266. Scandinavian Journal of Occupational Therapy 21(sup1): 71–79.

22.

Humphrey

(2006) Clinical applications: The exercise caloric challenge. ACSM's Health & Fitness Journal 10(2): 40–41.

23.

Iwarsson

Slaug

(2001) The Housing Enabler. An Instrument for Assessing and Analysing Accessibility Problems in Housing. Nävlinge and Staffanstorp: Veten & Skapen HB & Slaug Data Management.

24.

Iwarsson

Nygren

Slaug

(2005) Cross-national and multi-professional inter-rater reliability of the Housing Enabler. Scandinavian Journal of Occupational Therapy 12(1): 29–39.

25.

Joubert

Baraldi

(2016) La santé des Québécois: 25 indicateurs pour en suivre l’évolution de 2007 à 2014. Résultats de l’Enquête sur la santé dans les collectivités canadiennes, Québec, Institut de la statistique du Québec, 143 p. [25 indicators to monitor from 2007 to 2014. Results of the survey on health in Canadian communities, Quebec, Quebec Statistical Institute, 143 p.]. [online]. Available at : http://www.stat.gouv.qc.ca/statistiques/sante/etat-sante/sante-globale/sante-quebecois-2007-2014.pdf

26.

Lawlor

Taylor

Bedford

et al . (2002) Is housework good for health? Levels of physical activity and factors associated with activity in elderly women. Results from the British Women's Heart and Health Study. Journal of Epidemiology and Community Health 56(6): 473–478. http://doi.org/10.1136/jech.56.6.473

27.

Lawton

Nahemow

(1973) Ecology and the aging process. In: Eisdorfer C and Lawton MP (eds) Psychology of Adult Development and Aging. Washington, DC: American Psychological Association, pp.619–674.

28.

Leask

Harvey

Skelton

et al . (2015) Exploring the context of sedentary behaviour in older adults (what, where, why, when and with whom). European Review of Aging and Physical Activity 12(1): 1–8.

29.

Levasseur

Desrosiers

St-Cyr Tribble

(2008) Do quality of life, participation and environment of older adults differ according to level of activity? Health and Quality of Life Outcomes 6(1): 30–40.

30.

Levine

Kotz

(2005) NEAT–non‐exercise activity thermogenesis–egocentric & geocentric environmental factors vs. biological regulation. Acta Physiologica Scandinavica 184(4): 309–318.

31.

McCormack

Shiell

(2011) In search of causality: A systematic review of the relationship between the built environment and physical activity among adults. International Journal of Behavioral Nutrition and Physical Activity 8(1): 125–135.

32.

Manini

Everhart

Patel

et al . (2006) Daily activity energy expenditure and mortality among older adults. Journal of the American Medical Association 296(2): 171–179.

33.

Manini

Everhart

Patel

et al . (2009) Activity energy expenditure and mobility limitation in older adults: Differential associations by sex. American Journal of Epidemiology 169(12): 1507–1516.

34.

Millikan

(2012) Environmental Assessment & Modification for Australian Occupational Therapists. Sydney: Home Modification Information Clearinghouse, University of New South Wales. April 2012 [online]. Available at: http://www.homemods.info/Download.ashx?File=f380cdd3ef328029f147ae91c9ab4cd7 (accessed 9 February 2019).

35.

Nasreddine

Phillips

Bédirian

et al . (2005) The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. Journal of the American Geriatrics Society 53(4): 695–699.

36.

Oswald

Wahl

Schilling

et al . (2007) Relationships between housing and healthy aging in very old age. The Gerontologist 47(1): 96–107.

37.

Paterson

Warburton

(2010) Physical activity and functional limitations in older adults: A systematic review related to Canada’s Physical Activity Guidelines. International Journal of Behavioral Nutrition and Physical Activity 7(38): 1–22.

38.

Patry

Vincent

Duval

et al . (2019) A systematic review of standardised home assessments tools addressing accessibility. Canadian Journal of Occupational Therapy. In press.

39.

Pho

Tanner

Roth

et al . (2012) Nursing strategies for promoting and maintaining function among community-living older adults: The CAPABLE intervention. Geriatric Nursing 33(6): 439–445.

40.

Rantakokko

Iwarsson

Mänty

et al . (2012) Perceived barriers in the outdoor environment and development of walking difficulties in older people. Age and Ageing 41(1): 118–121.

41.

Russell

Ormerod

Newton

(2018) The development of a design and construction process protocol to support the home modification process delivered by occupational therapists. Journal of Aging Research 2018(4904379): 1–13. https://doi.org/10.1155/2018/4904379.

42.

Saelens

Handy

(2008) Built environment correlates of walking: A review. Medicine and Science in Sports and Exercise 40(7 Suppl): S550.

43.

Sallinen

Stenholm

Rantanen

et al . (2010) Hand‐grip strength cut points to screen older persons at risk for mobility limitation. Journal of the American Geriatrics Society 58(9): 1721–1726.

44.

Stark

Landsbaum

Palmer

et al . (2009) Client-centred home modifications improve daily activity performance of older adults. Canadian Journal of Occupational Therapy 76(1): 235–245.

45.

St-Onge

Mignault

Allison

et al . (2007) Evaluation of a portable device to measure daily energy expenditure in free-living adults. American Journal of Clinical Nutrition 85(3): 742–749.

46.

Wahl

Fänge

Oswald

et al . (2009) The home environment and disability-related outcomes in aging individuals: What is the empirical evidence? The Gerontologist 49(3): 355–367.

47.

Werngren-Elgstrom

Carlsson

Iwarsson

(2008) Changes in person-environmental fit and ADL dependence among older Swedish adults. A 10-year follow-up. Aging Clinical and Experimental Research 20(5): 469–478.

48.

Wolinsky

Bentler

Hockenberry

et al . (2011) Long-term declines in ADLs, IADLs, and mobility among older Medicare beneficiaries. BMC Geriatrics 11(43): 1–12.