Why Stay Home? Temporal 
Association of Pain, Fatigue and 
Depression with Being at Home
Authors: Craig Ravesloot, Bryce Ward, Tannis 
Hargrove, Jennifer Wong, Nick Livingston, Linda 
Torma, and Catherine Ipsen
NOTICE: this is the author’s version of a work that was accepted for publication in Disability & 
Health Journal. Changes resulting from the publishing process, such as peer review, editing, 
corrections, structural formatting, and other quality control mechanisms may not be reflected in 
this document. Changes may have been made to this work since it was submitted for publication. 
A definitive version was subsequently published in Disability & Health Journal, [VOL# 9, ISSUE# 
2, (November 2015)] DOI# 10.1016/j.dhjo.2015.10.010
Ravesloot, Craig, Bryce Ward, Tannis Hargrove, Jennifer Wong, Nick Livingston, Linda Torma, 
and Catherine Ipsen. "Why Stay Home? Temporal Association of Pain, Fatigue and Depression 
with Being at Home." Disability & Health Journal 9, no. 2 (November 2015). 
DOI: 10.1016/j.dhjo.2015.10.010
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
Why stay home? Temporal a
and depression with
Craig Ravesloota,*, Bryce Wardb, Tannis Ha
Livingstona, Linda Tormac, and Catherine Ip
a The Rural Institute on Inclusive Communities, USA
b The Bureau of Business and Economic Research, USA
ple w
outside the home. Existing studies of these secondary con-
ditions have focused little attention on their impact on daily
life. Thus, the impact of these secondary conditions on
daily life remains largely unknown.
jects and within subjects over time.
Pain is a secondary condition associated with a variety of
primary disabling conditions including spinal cord injury
(SCI),3 multiple sclerosis (MS),4 arthritis,5 traumatic brain
injury (TBI),6 amputation7 and cerebral palsy.8 Self-reportFor most people, participation in daily life involves leav-
ing home. People leave home to work, recreate, socialize, 
and shop, among other things.1 For people with disabilities, 
participation is a complex personal choice that is valued for 
many reasons.2 Secondary conditions like pain, fatigue and
depression may limit people’s engagement in activity
We address this gap in the literature. We obtained data
from a mixed impairment sample using a longitudinal elec-
tronic diary technique known as Ecological Momentary
Assessment (EMA). Using these data, we describe the rela-
tionship between pain, fatigue, depression and time spent at
home. We examine these relationships both between sub-like pain, fatigue and depression may increase the difficulty of leaving home. Despite decades of research on these secondary conditions, 
little is known about how they are associated with being at home.
Objective: We used Ecological Momentary Assessment data to examine within subject fluctuation in these secondary conditions to
examine their effect on the likelihood that participants remain at or return home.
Methods: Participants (n 5 139) were recruited from a population based sampling frame to complete an Ecological Momentary
Assessment that queried their location and experience with secondary conditions six times a day for two weeks.
Results: Between subjects secondary condition ratings averaged across time periods indicated that pain and depression were associated 
with the share of measurement periods that respondents reported being at home. Within subject results indicated that a standard unit in-
crease in pain, fatigue and depression was associated with being home one to two days later. Within day results indicated that increases 
in pain and fatigue were associated with increased likelihood of being home later, but increases in depression were associated with lower 
likelihood of being home later.
Conclusion: These results suggest there may be a complicated relationship among these secondary conditions and community partic-
ipation with effects observed both across and within days. One interpretation suggests that secondary condition severity is tempered by 
adjusting participation. These results may have implications for intervening on these secondary conditions. c Montana State University, School of Nursing, USA
Abstract
Background: Community participation is important to most peossociation of pain, fatigue
being at home
rgrovea, Jennifer Wonga, Nick 
sena
ith disabilities despite the fact that common secondary conditions measures of pain interference among people with disabilities
indicate that pain can affect social activities, recreation,
each EMA prompt, participants reported where they were
(e.g., at home), what they were doing (e.g., watching tele-mood, work, self-care, communication, learning new infor-
mation/skills, and enjoyment of life.9,10 Not surprisingly,
pain is associated with lower life satisfaction, poorer phys-
ical and mental health and level of handicap.11,12
Fatigue is also a common secondary condition reported by
people with MS,13 stroke,14 arthritis,15 Parkinson’s disease,16
TBI17 and SCI.18 It is associated with a reduction in daily ac-
tivity engagement, including social and recreational activiti-
es,19e21 but may not be associated with engagement in major
life activities like employment and education.22
Depression is a secondary condition that affects people
diagnosed with diabetes related disability,23 Parkinson’s’
disease,24 epilepsy,25 SCI,26 MS,27 as well as those who
have experienced a stroke.28 Endemic to depression is the
effect it has on engaging in activity and, not surprisingly,
depression is associated with a decrease in participation.23,29
When people make the decision to leave home, they
consider the costs and the benefits of going out. For people
experiencing pain, fatigue or depression the relative benefits
and costs of leaving home may change. To go out, among
other things (e.g., money), people need sufficient time, en-
ergy and motivation to engage in the planned activity. Pain,
fatigue, or depression may decrease the resources available
and make it more challenging to leave home. For instance,
these conditions may increase the time and effort to com-
plete self-care activities, and thus reduce the resources
available to pursue activities outside the home.30 Alterna-
tively, pain, fatigue, or depression may increase the ‘‘price’’
of activity outside the home. For instance, it may take more
time and effort for someone with these conditions to com-
plete the same activity.31 In either case, individuals experi-
encing greater pain, fatigue, or depression face more
constraints to engaging in activities outside the home.32
Cross-sectional studies are inadequate for understanding
the relationship between pain, fatigue, depression and com-
munity engagement. Standard between subjects analyses
cannot rule out the possibility that unobserved characteris-
tics drive observed correlations between pain, fatigue, or
depression and one’s propensity to remain home. Between
subjects analyses also ignore within-person fluctuations in
pain, fatigue, or depression. Within person fluctuations
matter. They may drive an observed between subject rela-
tionship. Furthermore, they can illuminate whether people
go home when they experience an increase in pain, fatigue,
or depression. To develop the most effective interventions
to increase community engagement, researchers need longi-
tudinal data that is sensitive to daily fluctuations and indi-
vidual differences.
Ecological Momentary Assessment (EMA), a tech-
nique subsumed under Experience Sampling Methodol-
ogy, collects data in situ, which facilitates investigation
of status fluctuations and individual differences.33,34 It
typically uses electronic devices to collect data in the per-
son’s usual environment. It is a reliable and valid method
for collecting data on a variety of variables (e.g., pain and
emotional state) and reduces retrospective recall biasvision), and questions about their experience in the moment
(e.g., who they were with, how satisfied they were with the
activity, their pain level). To indicate the level of pain they
were experiencing, they used a 10-point rating scale
anchored at each end (0 5 No pain, 10 5 Pain as bad asassociated with autobiographic recall errors.34,35 The
aim of the current study was to use EMA to explore the
within-person temporal relationship between pain, fatigue,
and depression on the amount of time people with impair-
ments spend at home.
Methods
Participants
We recruited 525 participants in a small Western US city
from an initial pool of 10,000 randomly selected house-
holds to complete a paper-and-pencil survey. Invitation let-
ters and pre-coded, anonymous, postage paid post-cards
were sent to each household. Interested participants who re-
turned the post card and met inclusion criteria (i.e.,
endorsed one or more American Community Survey
[ACS] questions)36 were sent a survey using Dillman’s pro-
cedures for materials development (e.g., cover letter con-
tent) and follow up.37 We recruited 149 of the survey
respondents to complete the EMA by telephoning all indi-
viduals who gave permission on the survey to contact them
for another study. The current study sample included 139
participants who responded to at least half of the 84 sched-
uled EMA prompts over the 14 day study period. Data were
collected in 2013 and 2014.
Respondents included in the current analytic sample were
between the ages of 19 and 75 (M 5 55.7, SD 5 11.2) and
were predominantly female (61.8%), Caucasian (97.1%),
and college educated (73.3% reported some college or more
education). The majority were not employed (64.5%) and
reported median household income between $20,000 and
$30,000 (Table 1).
Measures
We used the paper-and-pencil survey to collect a stan-
dard set of demographic variables including age, gender, in-
come, employment status and educational attainment. We
also collected the six American Community Survey
(ACS) disability screener questions and the type of adaptive
equipment respondents normally use.
We collected the EMA data using four-inch touch screen
tablet computers (i.e., Samsung Galaxy Player 4.0) that
were programmed to prompt participants using a random
schedule within 2-h intervals (delivered between 9:00 am
and 9:00 pm), six times daily, for 14 consecutive days. Atyou can imagine). Respondents used a five-point rating
scale (i.e., 0 5 Not at all, 1 5 A little, 2 5 Somewhat,
Relationship status
Married 69 50.0No longer married 35 25.4
Never married 18 13.0
Member of unmarried couple 9 6.5
Education
Grade 12 or GED or less 37 26.6
Some college or tech training or associate’s 58 41.7
Bachelor’s degree or higher 44 31.6
Income
!$10,000 21 15.3Table 1
Demographic characteristics of respondents (n 5 139)
Variable n %
Gender
Male 53 38.1
Female 86 61.8
Ethnicity
White 135 97.1
Non-white 4 2.93 5 Quite a bit, 4 5 Very much) to indicate their current
level of fatigue and depressed mood. Lastly, we obtained
consent from 148 subjects to use the Global Positioning
System (GPS) receiver for each EMA device to collect
the geographic coordinates of the individual’s location.38
This data was later used to validate the degree to which
self-reported location matched GPS coordinates.
Procedures
All procedures were reviewed and approved by the Insti-
tutional Review Board at the University of Montana. Those
recruited to complete the EMA attended a 90-min orienta-
tion where they provided informed consent for the study
including opt-out consent for allowing the GPS receiver
of the device to be activated. Then, we instructed partici-
pants on how to interpret each question and available re-
sponses, and how to enter their responses on the EMA
effects and controlled for time of day (dummy variables
$10,001e40,000 68 49.6
$40,001e80,000 36 26.3
$80,001þ 12 8.8
Employment
Full-time O30 h/week 31 22.5
Part-time 1e30 h/week 18 13.0
Not employed 89 64.5
Volunteering
Weekly volunteer 15 11.0
Occasional volunteer 52 38.0
Does not volunteer 70 51.0
Impairment
Deaf/hard of hearing 23 16.5
Blind or difficulty seeing 13 9.0
Concentration, memory, or decision making difficulty 52 37.9
Difficulty walking or climbing stairs 96 70.0
Difficulty dressing or bathing 41 29.9
Difficulty running errands 49 35.5
Equipment used
Needs physical assistance from others 35 25.4
Uses a walker, cane, crutch, or brace 36 26.1
Manual or power wheelchair users 11 7.9for each of the 6 measurement periods e 2 h blocks be-
tween 9 am and 9 pm e each day) and day of the week
(a dummy variable equal to one if the measurement period
occurred on a weekend). We explored both across and
within day lags. For across day lags, we used the average
of all available measurements from the previous day and re-
gressed this mean on the probability of being at home
today. For within day lags, we examined whether within-
day lags were associated with participants being at home
during any given time period that could be evaluated given
the lag structure. For example, a one period lag within thedevice. We also gave each participant a training guidebook
they could take home and refer to later.
For people with visual impairments, we offered a larger
format device (i.e., Samsung Galaxy Tab 7.0 seven-inch
tablet); all participants chose the smaller device and used
personal magnifiers to view content. The sample did not
include any participants who were totally blind. For indi-
viduals with dexterity problems, who might have had diffi-
culty responding to a touch screen, we offered a stylus and/
or the larger format device. All of these participants chose
to use a stylus and the smaller device.
Data handling and analytic strategy
We entered all survey data into a Microsoft Excel39
spreadsheet programmed with input value constraints that
did not allow out-of-range entries and checked the data
input accuracy. We reviewed EMA data uploaded from
the devices for accuracy and converted the raw data into
Excel files. All data were imported into Stata40 for statisti-
cal analysis and standardized to improve interpretability for
variables with different scales.
We divided our analysis into between subject and within
subject analyses. For our between subjects analyses, we
averaged individual EMA responses for all available mea-
surement periods and combined them with demographic in-
formation obtained from the pre-EMA survey. We then
examined descriptive statistics, correlations, and linear
regression results. In our between subject regression anal-
ysis, we regressed the share of observations at home
(measured as a standardized score) on standardized scores
for a subjects’ average pain, fatigue, and depressed mood
(across the 14-day EMA measurement window) controlling
for sex, age, ethnicity, and education. Confidence intervals
for this regression were computed using heteroskedastic-
robust standard errors.
For our within subject analyses, we examined the rela-
tionship between lagged (i.e., self-reported symptoms pre-
ceding subsequent ‘‘at home’’ observations) individual
reports of pain, fatigue, and depressed mood and the likeli-
hood that an individual is at home. We estimated these re-
lationships using logistic regression with individual fixedday could be evaluated with five time periods, while a three
period lag could only be evaluated with the three periods.
Results
structed was 0.92 ( p ! 0.01). We used the subject’s EMA
increased by 14%.
Table 2
Frequency of observations by location
Location Frequency (%) n (%)
Business or store 615 (6) 123 (88)
Church or religious place 80 (!1) 33 (24)
Gym or exercise facility 84 (!1) 39 (28)
Health care facility 168 (2) 59 (42)
Home 7132 (70) 139 (100)
Office building 274 (3) 48 (35)
Outside 733 (7) 110 (79)
Restaurant or bar 145 (1) 63 (45)
School or education 188 (2) 31 (22)
Someone else’s home 259 (3) 65 (47)
Transportation vehicle 371 (4) 96 (69)
Venue such as a theater 34 (!1) 18 (13)
Other location 126 (1) 49 (35)
Frequency is the total number of observations in each location.
n 5 the number of individuals with at least one observation in each
location.
Table 3
Linear regression of between subject predictors for the proportion of time
spent at home
Variables
Proportion of time at home
b p 95% CI
Constant 0.58 0.69 1.82, 0.65
Age 0.01 0.24 0.01, 0.03
Gender 0.03 0.86 0.32, 0.38
Caucasian 0.15 0.68 0.58, 0.88
Some college 0.16 0.38 0.54, 0.21
College 0.35 0.13 0.79, 0.10
Average pain 0.19* 0.04 0.01, 0.37
Average fatigue 0.14 0.29 0.39, 0.12
Average depression 0.22* 0.03 0.02, 0.42
R2 0.10
Variables entered represent average values across all 84 EMA measure-
ment periods. Gender is coded 1 for women, 0 for men; Caucasian is coded
1, Caucasian 1, other ethnicities 0; Some College and College variables areresponse as the dependent variable in these analyses because
there were some obvious anomalies in the GPS location
coordinates (e.g., coordinates in uninhabited areas).aWith re-
gard to the independent variables, the average raw scale pain
rating across all observations was 3.86 (SDBetween 5 2.08;
SDWithin 5 1.38), average depression was 1.73
(SDBetween 5 0.90; SDWithin 5 0.56) and average fatigue
was 2.54 (SDBetween5 0.84; SDWithin 5 0.80).
The distribution of locations reported by total observa-
tions across all respondents and by the number of respon-
dents is included in Table 2. Additionally, we examined
the extent to which people reported electronic or computer
based activity at home, which could affect interpretation of
results. Of 596 observations of socializing, 343 were at
home which included 191 instances of talking on the phone
(n5 77), 46 instances of social networking (n5 16) and 30
instances of other electronic communication (n 5 15).
Additionally, there were 388 instances of computer use
for activities such as gaming or shopping (n 5 74). Com-
bined, these social and computer based activities at home
accounted for 10% of the total observations at home.
At the between subject level (i.e., measured EMA re-
sponses averaged over 84 measurement periods), the corre-
lations between the proportion of time respondents spent at
home and their level of pain (r 5 0.19), fatigue (r 5 0.05)
and depressed mood (r5 0.17) were statistically significant
( p ! 0.05) for pain and depressed mood. Taken together
and regressed on the proportion of time spent at home, after
adjusting for basic demographic variables, average pain and
depressed mood were both positively associated with the
proportion of time individuals reported being home, while
fatigue was not positively associated with the proportion
of time individuals reported being at home (Table 3).
Next, we examined the effects of within subject changes
in pain, fatigue and depressed mood on the likelihood of
being home within and across days, after accounting for
variability due to time periods within the same day (i.e.,
time periods 1e6) and weekend versus weekday activity
(Tables 4e6).
Pain
Table 4 includes the effect of changes in pain level on
the likelihood the individual is at home. This table shows
a We programmed devices to use the coordinates of a WIFI signal if a
GPS satellite lock was unavailable (i.e., when the device was in a largeOverall, respondents completed 87% of the 84 possible
study prompts. They reported that they were at home for
70% of those prompts. The Spearman rank order correlation
between subjects ‘‘at home’’ response to the EMA question
‘‘Where are you?’’ and the GPS coordinate variable we con-building with no windows). Location anomalies were associated with the
WIFI location coordinates.that individuals who experienced more pain yesterday or
more pain earlier in the day (up to 6 h prior) are more likely
to be at home in the current period. Specifically, for a one
standard deviation unit increase in average pain yesterday,
the odds that an individual is at home increases by an
average of 36%. A one standard-unit increase in pain three
periods earlier in the day increases the odds that an individ-
ual is at home currently by 18%, on average. Lastly, within-
period pain results indicate that individuals reporting cur-
rent pain are 24% less likely to be at home.
Fatigue
Results for fatigue are different (Table 5). There were
no observed effects of fatigue on the likelihood of being
at home measured within the period, or lagged one day.
However, for a one standard-unit increase in fatigue one
period prior, the odds of being home in the current periodin reference to high school education or less.
*p ! 0.05.
amount of time people spent at home and on the likelihood
that people were at home at any given time during the
Table 4
Pain measured contemporaneously, lagged one day, and lagged between
periods within days regressed on being at home
Pain
At home
OR 95% CI
Current pain 0.76** 0.61e0.93
Mean pain yesterday 1.36** 1.08e1.72
Pain lagged 3 periodsa 1.18* 1.01e1.38
Time period 4 (ref.)b 1
Time period 5 1.77*** 1.40e2.23
Time period 6 3.32*** 2.34e4.07
Weekend 1.10 0.87e1.40
Time period variables represent measurement periods within the day.
Weekend is a binary weekday (coded 0), weekend day (coded 1) variable.
Values presented here represent standardized coefficients.
*p ! 0.05; **p ! 0.01; ***p ! 0.001.
a Only home status at periods 4, 5, and 6 were covaried given the effect of
within day pain lagged up to 3 periods.
b Time periods 5 and 6 were entered as covariates (time period 4 refer-
ence) to account for within day activity and to isolate the effects ofDepressed mood
As with fatigue, there was no effect of within-period
depressed mood on the likelihood of being at home in the
current period (Table 6). However, results indicate that
average depressed mood scores yesterday, and one mea-
surement period prior (up to 2 h) were related to the odds
of being at home in the current period. On average, individ-
uals who reported a one standard-unit increase in average
depressed mood yesterday were 28% more likely to report
being at home in the current period. On the other hand, for
a one standard-unit increase in depressed mood one period
prior, individuals were 15% less likely to report being at
home. These results suggest that individuals who experi-
contemporaneous pain, pain lagged one day and pain within the day.ence increases in average prior-day depressed mood tend
to spend more time at home the following day, but are less
Table 5
Fatigue measured contemporaneously, lagged one day, and lagged between
periods within days regressed on being at home
Fatigue
At home
OR 95% CI
Current fatigue 0.92 0.82e1.03
Mean fatigue yesterday 1.11 0.94e1.31
Fatigue lagged 1 period 1.14** 1.03e1.27
Time period 2 (ref.)a 1
Time period 3 0.92 0.76e1.10
Time period 4 1.23 0.97e1.56
Time period 5 2.20*** 1.68e2.89
Time period 6 3.98*** 2.82e5.61
Weekend 1.26* 1.02e1.55
Time Period variables represent measurement periods within the day.
Weekend is a binary weekday (coded 0), weekend day (coded 1) variable.
Values presented here represent standardized coefficients.
*p ! 0.05; **p ! 0.01; ***p ! 0.001.
a Time periods 3 through 6 are accounted for (period 2 reference) to
observe the effects of contemporaneous fatigue, and fatigue lagged one
day, and fatigue lagged one period prior.study. Between subject differences in the proportion of time
people spent at home over the course of the study were
observed only for pain and depressed mood. Within subjectlikely to remain home if their depressed mood increases
within a given day.
Discussion
We used EMA to explore the temporal association be-
tween pain, fatigue, and depressed mood on both the
Table 6
Depression measured contemporaneously, lagged one day, and lagged
between periods within days regressed on being at home
Depression
At home
OR 95% CI
Current depressed mood 1.10 0.93e1.32
Mean depressed mood yesterday 1.28* 1.02e1.62
Depressed mood lagged 1 period 0.85* 0.73e0.99
Time period 2 (ref.)a 1
Time period 3 0.93 0.78e1.11
Time period 4 1.26* 1.00e1.59
Time period 5 2.32*** 1.78e3.02
Time period 6 4.15*** 2.99e5.77
Weekend 1.24* 1.01e1.53
Time Period variables represent measurement periods within the day.
Weekend is a binary weekday (coded 0), weekend day (coded 1) variable.
Values presented here represent standardized coefficients.
*p ! 0.05; ***p ! 0.001.
a Time periods 3 through 6 are accounted for (period 2 reference) to
observe the effects of contemporaneous fatigue, and fatigue lagged one
day, and fatigue lagged one period prior.results showed that pain and depressed mood were associ-
ated with the likelihood that an individual was at home both
across and within days. Fatigue was only related to the like-
lihood of being home within a day. These results are consis-
tent with prior research that has linked higher levels of pain
and fatigue to lower activity levels. Higher levels of activity
following higher levels of depressed mood within the day
were contrary to our expectations. These results highlight
the ecological fallacy of using between subject results to
infer within subject effects.41
The between subject results for pain and depressed mood
were consistent with those reported in cross-sectional
research, with higher average pain and depressed mood
associated with a greater proportion of time spent at home.
For example, pain interferes with the activities of people
with MS and SCI.9,10 However, for this sample there was
not an association between fatigue and the proportion of
time spent at home. One possibility for this difference
may be that while fatigue may affect the quality of partici-
pation, it has a smaller effect on the quantity as indicated
by cross-sectional modeling. Alternatively, there may be
opposing effects such that people with higher fatigue are
observation that is consistent with instrumentation effects.either more likely to stay home or people who are more
active report more fatigue, but are less likely to be home.
Within subject effects are evident for all three vari-
ables. Depressed mood and pain have effects across days.
This may reflect the fact that interventions and self-
management strategies for these conditions experienced
by our sample have limited short-term effects when
compared with the effect that resting or sleep have on
reducing fatigue. Within days, people have less pain when
they are at home; however, increases in pain during the
early part of a day increases the likelihood that an individ-
ual is at home later in the day. These results suggest that
when people notice an increase in pain they go home (or
remain home). Fatigue shows a similar pattern within the
day e people report lower fatigue at home and increases
in fatigue earlier in the day are associated with being at
home later in the day. For these data, fatigue has an imme-
diate effect with the likelihood of being home, evident
within the next 2 h, while pain has both an immediate
and lagged effect with the likelihood of being home,
which is evident for as long as 6 h later. This indicates that
pain affects being home over a longer period than fatigue,
which could reflect that rest improves fatigue, which re-
duces its longer-term effects on being home (e.g., after
resting, people may go out again).
In contrast to pain and fatigue, we observed a different
within day-effect for depressed mood. Increases in
depressed mood ratings within the day were associated with
decreased probability that individuals would be home one
period later. This could reflect a self-management strategy
used by respondents, wherein perceived increases in
depressed mood prompt going out to help manage mood.
This would be consistent with behavioral activation treat-
ment42 for depression that has shown that mood improves
when activity levels increase.43 Another interpretation
might be that depression scores increase in anticipation of
planned or scheduled activities outside the home.
These results may reflect the tension that people experi-
ence between being out in the community and managing
their pain, fatigue and depressed mood. On the one hand, in-
crease in depressed mood is associated with getting out of
the house within the next few hours; on the other hand, being
out increases pain and fatigue.29,44 This may set up a
dysfunctional cycle of symptom management that ultimately
has the potential to decrease community engagement. Alter-
natively, because pain and fatigue may be reduced over time
through exercise and physical activity that increases physical
capacity, these results suggest that any subsequent increase
in activity may help people manage depressed mood.45
Lastly, if these results reflect depression self-management,
then they may reflect the importance of increasing physical
capacity that supports community engagement without
increasing pain and fatigue.
Daily activity in the community comes at a price and
sooner or later, everyone goes home. Time spent away from
home may be understood along a continuum based onThird, the lagged structure of the data analytic approach
is a limitation. While useful for examining associations
over time, the lagged structure limits our ability to model
both the sequential effects of secondary conditions on each
other and on being at home simultaneously given the sam-
ple size of this study. Consequently we chose to analyze
pain, fatigue, and depression as predictors independently.
Missing data represents another potential source of bias
in this study. If the pattern of missing data is related to
the temporal pattern observed in the independent variables,
this artifact could account for these results. We examined
the data for these patterns but did not find evidence of this
effect. Fourth, in retrospect, the omission of an EMA vari-
able about pain medication use is another study limitation.
These results may reflect both analgesic and fatigue associ-opportunity, preference, and constraint parameters. At one
end, total lack of opportunity to access the community is
associated with spending no time away from home. Howev-
er, with opportunity comes choice, which is governed by
preferences and constraints. For people with disabilities
and associated secondary conditions, preferences may be
shaped by available opportunities and constraints that exert
a greater influence on their choices than those without dis-
abilities. In this framework, pain and fatigue are constraints
on community engagement that may increase risk for
depressed mood. Research on symptom clustering has sug-
gested that pain, fatigue and depression move together when
examined with cross-sectional data.46,47 The results of this
study may reflect the balance people work to achieve be-
tween pain, fatigue, depression and community engagement.
Limitations
There are a number of limitations to interpreting these
results. First, the study sample was limited in geographic
coverage and not demographically representative of the
US population of people with disability, which limits the
generalizability of these results. Further, without a compar-
ison sample of people without disability, we do not know if
these results are specific to people with disabilities or
would also be observed in the general population. Second,
while use of EMA may reduce recall error, repeated mea-
sures may prompt instrumentation effects. For example, rat-
ing secondary conditions six times a day may have
increased secondary condition salience for respondents,
which may have prompted behavior (e.g., raised awareness
of depressed mood prompting people to leave home). We
examined the data for these effects and while we found a
slight increase in the value of pain ratings over the course
of the study (ETA 5 0.14, p ! 0.01), we did not find
any evidence of trends in either fatigue or depression. Some
participants made anecdotal comments that rating their pain
helped them understand its effects on their daily life, anated with pain medication use. Finally, while our results
suggest that relationships exist between pain, fatigue,
strictly exogenous and therefore our analysis may suffer
from some degree of endogeneity or omitted variable bias
tion to understand better sequential effects of secondary
conditions on participation. Alternatively, future research
Leaving home is important to living a full life and an
important aspect of community engagement. Results from
this study indicate that secondary conditions play an impor-
tant role in that choice. Even more, these results suggest
there may be opposing effects of community participation
on secondary conditions. Individuals may leave home to
manage depressed mood, but return home when pain and
fatigue increase. As such, this study suggests that the most
effective interventions may address a symptom cluster of
pain, fatigue and depression.
Acknowledgments
The authors would like to thank our research advisors
David Gray, Mark Jensen, Mike Mayer and Glen White
for their invaluable contributions to the development of
project materials and procedures. We also acknowledge
the contributions of Lillie Greiman and research assistants
Lorraine Adams, Seterra Burleson, Cathy Berendts, Alex-
andra Schiwal and Sandra Skogley, who contributed to data
collection and management for the project.
References
1. Flood SM, Moen P. Healthy time use in the encore years: do work,could examine how specific activities (e.g., visiting
friends, shopping alone, doing volunteer work) affect sec-
ondary conditions and participation. Lastly, future studies
may use EMA to evaluate interventions (e.g., ecological
momentary interventions) to increase community partici-
pation and to examine the sequential effects of these sec-
ondary conditions as people make choices to increase
their involvement in their community.
Conclusion(e.g., pain medication use).
Modern technology that allows researchers to collect
EMA data is now ubiquitous in much of the world. Future
research may capitalize on these measurement techniques
to extend participation research that will allow investiga-
tors to examine with much greater resolution personal
and environmental factors that affect participation (Living-
ston et al, 2015). For example, with more subjects, this line
of research could begin to model the temporal effects of
these secondary conditions on each other and on participa-depression and one’s propensity to be at home, the variation
in secondary conditions used to estimate our results is notresources, relations, and gender matter? J Health Soc Behav.
2015;56(1):74e97.2. Hammel J, Magasi S, Heinemann A, Whiteneck G, Bogner J,
Rodriguez E. What does participation mean? An insider perspective
from people with disabilities. Disabil Rehabil. 2008;30(19):
1445e1460.
3. Saulino M. Spinal cord injury pain. Phys Med Rehabil Clin N Am.
2014;25(2):397e410.
4. Foley PL, Vesterinen HM, Laird BJ, et al. Prevalence and natural his-
tory of pain in adults with multiple sclerosis: systematic review and
meta-analysis. Pain. 2013;154(5):632e642.
5. Jakobsson U, Hallberg IR. Pain and quality of life among older people
with rheumatoid arthritis and/or osteoarthritis: a literature review. J
Clin Nurs. 2002;11(4):430e443.
6. Lucas S, Hoffman JM, Bell KR, Dikmen S. A prospective study of
prevalence and characterization of headache following mild traumatic
brain injury. Cephalalgia. 2014;34(2):93e102.
7. Nikolajsen L, Jensen TS. Phantom limb pain. Br J Anaesth.
2001;87(1):107e116.
8. Van Der Slot WM, Nieuwenhuijsen C, Van Den Berg-Emons RJ, et al.
Chronic pain, fatigue, and depressive symptoms in adults with spastic
bilateral cerebral palsy. Dev Med Child Neurol. 2012;54(9):836e842.
9. Raichle KA, Osborne TL, Jensen MP, Cardenas D. The reliability and
validity of pain interference measures in persons with spinal cord
injury. J Pain. 2006;7(3):179e186.
10. Osborne TL, Raichle KA, Jensen MP, Ehde DM, Kraft G. The reli-
ability and validity of pain interference measures in persons with mul-
tiple sclerosis. J Pain Symptom Manage. 2006;32(3):217e229.
11. Putzke JD, Richards JS, Dowler RN. The impact of pain in spinal cord
injury: a caseecontrol study. Rehabil Psychol. 2000;45(4):386e401.
12. Osborne TL, Turner AP, Williams RM, et al. Correlates of pain inter-
ference in multiple sclerosis. Rehabil Psychol. 2006;51(2):166e174.
13. Induruwa I, Constantinescu CS, Gran B. Fatigue in multiple
sclerosisda brief review. J Neurol Sci. 2012;323(1):9e15.
14. Lerdal A, Bakken LN, Kouwenhoven SE, et al. Poststroke fatigueda
review. J Pain Symptom Manage. 2009;38(6):928e949.
15. Murphy SL, Smith DM, Clauw DJ, Alexander NB. The impact of
momentary pain and fatigue on physical activity in women with oste-
oarthritis. Arthritis Care Res. 2008;59(6):849e856.
16. Friedman JH, Brown RG, Comella C, et al. Fatigue in Parkinson’s dis-
ease: a review. Mov Disord. 2007;22(3):297e308.
17. Bushnik T, Englander J, Wright J. The experience of fatigue in the first
2 years after moderate-to-severe traumatic brain injury: a preliminary
report. J Head Trauma Rehabil. 2008;23(1):17e24.
18. Craig A, Tran Y, Wijesuriya N, Middleton J. Fatigue and tiredness
in people with spinal cord injury. J Psychosom Res. 2012;73(3):
205e210.
19. Michael KM, Allen JK, Macko RF. Fatigue after stroke: relationship
to mobility, fitness, ambulatory activity, social support, and falls effi-
cacy. Rehabil Nurs. 2006;31(5):210e217.
20. Flachenecker P, Kumpfel T, Kallmann B, et al. Fatigue in multiple
sclerosis: a comparison of different rating scales and correlation to
clinical parameters. Mult Scler. 2002;8(6):523e526.
21. Gagnon C, Mathieu J, Jean S, et al. Predictors of disrupted social
participation in myotonic dystrophy type 1. Arch Phys Med Rehabil.
2008;89(7):1246e1255.
22. Cantor JB, Ashman T, Gordon W, et al. Fatigue after traumatic brain
injury and its impact on participation and quality of life. J Head
Trauma Rehabil. 2008;23(1):41e51.
23. Von Korff M, Katon W, Lin E, et al. Potentially modifiable factors
associated with disability among people with diabetes. Psychosom
Med. 2005;67(2):233e240.
24. Arun MP, Bharath S, Pal PK, Singh G. Relationship of depression,
disability, and quality of life in Parkinson’s disease: a hospital-based
case-control study. Neurol India. 2011;59(2):185e189.
25. Fiest KM, Dykeman J, Patten SB, et al. Depression in epilepsy: a sys-
tematic review and meta-analysis. Neurology. 2013;80(6):590e599.26. Saunders LL, Krause J, Focht K. A longitudinal study of depression in
survivors of spinal cord injury. Spinal Cord. 2012;50(1):72e77.
27. Chwastiak L, Ehde DM, Gibbons LE, Sullivan M, Bowen JD,
Kraft GH. Depressive symptoms and severity of illness in multiple
sclerosis: epidemiologic study of a large community sample. Am J
Psychiatry; 2002:1862e1868.
28. Ayerbe L, Ayis S, Wolfe CD, Rudd AG. Natural history, predictors and
outcomes of depression after stroke: systematic review and meta-anal-
ysis. Br J Psychiatry. 2013;202(1):14e21.
29. Wilkie R, Peat G, Thomas E, Croft P. Factors associated with partic-
ipation restriction in community-dwelling adults aged 50 years and
over. Qual Life Res. 2007;16(7):1147e1156.
30. Mitra S. The capability approach and disability. J Disabil Policy Stud.
2006;16(4):236e247.
31. World Health Organization. International Classification of Func-
tioning, Disability, and Health. Geneva: World Health Organiza-
tion; 2001.
32. Knaggs JD, Larkin KA, Manini TM. Metabolic cost of daily activities
and effect of mobility impairment in older adults. J Am Geriatr Soc.
2011;59(11):2118e2123.
33. Csikszentmihalyi M, Larson R, Prescott S. The ecology of adolescent
activity and experience. J Youth Adolesc. 1977;6(3):281e294.
34. Stone AA, Shiffman S. Ecological momentary assessment (EMA) in
behavorial medicine. Ann Behav Med. 1994;16(3):199e202.
35. Piasecki TM, Hufford MR, Solhan M, Trull TJ. Assessing clients in
their natural environments with electronic diaries: rationale, benefits,
limitations, and barriers. Psychol Assess. 2007;19(1):25e43.
36. United States Census Bureau. American Community Survey S1801.
Disability Characteristics, http://factfinder.census.gov/servlet/STTable?
_bm5y&-qr_name5ACS_2005_EST_G00_S1801&-geo_id501000US
&-ds_name5ACS_2005_EST_G00_&-_lang5en&-format5&-CONT
EXT5st; 2005. Accessed 19.05.08.
37. Dillman D. Mail and Internet Surveys. The Tailored Design Method.
2nd ed. New Jersey: John Wiley & Sons, Inc; 2007.
38. Google I. Configure access points with Google location feature. Up-
dated 2015. https://support.google.com/nexus/answer/1725632?hl5en;
Accessed 29.05.15.
39. Microsoft. Microsoft Excel. Microsoft Corporation; 2013.
40. StataCorp. Stata Statistical Software: Release 14; 2015.
41. Monaco J. Ecological Fallacy. Encyclopedia of Behavioral Medicine;
2013:644.
42. Hopko DR, Lejuez C, Ruggiero KJ, Eifert GH. Contemporary behav-
ioral activation treatments for depression: procedures, principles, and
progress. Clin Psychol Rev. 2003;23(5):699e717.
43. Jacobson NS, Dobson KS, Truax PA, et al. A component analysis of
cognitive-behavioral treatment for depression. J Consult Clin Psychol.
1996;64(2):295.
44. Malone LA, Vogtle LK. Pain and fatigue consistency in adults with ce-
rebral palsy. Disabil Rehabil. 2010;32(5):385e391.
45. Rimmer JH, Chen M, McCubbin JA, Drum C, Peterson J. Exercise
intervention research on persons with disabilities: what we know
and where we need to go. Am J Phys Med Rehabil. March 01,
2010;89(3):249e263.
46. Motl RW, Suh Y, Weikert M. Symptom cluster and quality of life in
multiple sclerosis. J Pain Symptom Manage. 2010;39(6):1025e1032.
47. Ravesloot C, Seekins T, Walsh J. A structural analysis of secondary
conditions experienced by people with physical disabilities. Rehabil
Psychol. 1997;42(1):3e16.
r a g
erin
ot a
Preliminary results were presented at the Association of Programs for Rural Independent Living annual meeting. Disclosures: The contents of this paper were developed unde
number H133G110077 awarded to Craig Ravesloot and Cath
policy of the Department of Education, and readers should n
Albuquerque, NM 2014.rant from the Department of Education, NIDRR grant 
e Ipsen. However, those do not necessarily represent the 
ssume endorsement by the Federal Government.