Daily and Weekly Associations among Pain Intensity, Self-Reported Activity Limitations, and Objectively Assessed Physical Activity in Youth with Acute Musculoskeletal Pain

Abstract

Monitoring recovery during acute pain episodes is useful for identifying youth at-risk for pain persisting. Subjective and objective measures can assess function post-injury, but associations among these different measures and pain patterns in the acute period are unknown. To fill this gap, we examined associations among self-reported activity limitations, objectively measured physical activity, and pain intensity in 176 youth (age 11–17, 46% male) seeking healthcare for acute musculoskeletal pain. Participants completed 7-day electronic diaries rating daily pain intensity and activity limitations (Child Activity Limitations Interview, CALI-9) while concurrently wearing an Actiwatch to record physical activity. Results revealed youth reported pain on 47.8% of days with average intensity of 33.4 (0–100). Averaged across the week, between-participant analyses showed greater activity limitations were associated with lower mean (r_Active= −.204, r_Routine= −.159, p<.05) and peak activity (r_Active= −.291, r_Routine= −.184, p<.05). Same-day correlations between CALI scores and physical activity measures within participants were not significant. Linear mixed effects models revealed higher daily pain intensity was associated with greater self-reported activity limitations on Routine (β=.23, p<.001) and Active CALI-9 subscales (β=.07, p<.001). Conversely, higher daily pain was associated with higher activity on actigraphy, specifically higher mean activity (β=.46, p<.01), more activity bouts (β=.013, p<.01), more time in light activity (β=.04, p<.01), and less sedentary time (β=−.04, p<.01). Taken together, self-reported activity limitations and objective physical activity represent two distinct, yet related, aspects of physical functioning associated with pain. Future work should examine how physical activity and activity limitations change longitudinally and predict pain persistence.

Perspective:

This study examined daily associations between pain intensity, self-reported activity limitations, and objectively assessed physical activity in youth during the acute recovery period following a musculoskeletal injury. Self-reported activity limitations and objective physical activity represent two distinct, yet related, aspects of physical functioning that are associated with pain.

Introduction

Acute musculoskeletal (MSK) pain is common among youth¹², with up to 30% developing persistent musculoskeletal pain post-injury¹⁰. Youth experiencing pain often report activity limitations¹⁹ including difficulties engaging in physical (e.g., sports) and routine activities (e.g., schoolwork) due to pain¹⁵. Given activity limitations are an important indicator of quality of life and daily functioning^{2, 20}, they can be vital for monitoring recovery during an acute pain episode and for identifying youth at-risk for pain persisting or poor functional outcomes. Indeed, previous research in youth with acute postoperative pain demonstrates that youth who engaged in physical activity following surgery were less likely to develop chronic post-surgical pain 4 months later¹⁸.

Appropriate tools are needed for assessing and tracking pain-related disability in youth with acute MSK pain. Validated self-report measures for assessing pain-related impairment in youth are available (e.g., Functional Disability Inventory [FDI]²²; PROMIS Pain Interference²¹; and Child Activity Limitations Interview [CALI]¹⁵). Key limitations of these measures are the retrospective format (participants report on symptoms in past 1–2 weeks or month) and lack of relevancy of items for broad samples (e.g., running length of a football field)¹¹. To fill this gap, the CALI-9 was developed to briefly assess pain-related activity limitations via a daily diary. The CALI-9 includes subscales for assessing impairment in active (sports, running, walking) and routine (sleeping, eating, schoolwork) areas and uses an easily interpretable scoring algorithm (range 0–100; higher scores indicate higher impairment)¹¹. The CALI-9 daily diary has previously been used to assess pain outcomes in youth with chronic pain¹⁴, but not with youth experiencing acute MSK pain.

Actigraphy has been used to objectively assess physical activity in youth with acute and chronic pain. In youth with chronic pain, higher pain intensity is related to lower levels of next day activity^{17, 23}. Further, youth who engaged in less physical activity two weeks following spinal fusion surgery were more likely to develop chronic post-surgical pain at 4-months than those with higher physical activity¹⁸. Research has not examined objectively measured physical activity in youth with acute MSK injuries, particularly associations among daily physical activity and pain. While lower levels of physical activity are common and often expected during acute recovery, the association between daily fluctuations in objectively measured physical activity and pain symptoms is unknown.

One conceptual model of physical functioning in youth chronic pain proposes three broad dimensions: perceived disability (e.g., self-reported activity limitations), physical activity (e.g., vigorous sports/exercise), and physiological function (e.g., functional capacity)²³. The model proposes that associations between actual and perceived limitations across these dimensions impact youths’ perceptions of overall physical functioning. Applying components of this framework to youth with acute MSK pain and knowing how physical activity relates to self-reported activity limitations during the recovery period is important. These data may be used to inform decisions on using self-report versus objective measures to assess physical functioning, given the unique burdens associated with collecting each form of monitoring.

To fill these research gaps, this study examined associations among two domains of physical function outlined in Wilson & Palermo’s (2012) conceptual model in a sample of youth with acute musculoskeletal injuries to 1) examine associations among perceived disability (activity limitations assessed via CALI-9 diary) and physical activity (measured by actigraphy) at the daily level and averaged across the week, and 2) examine associations between pain intensity and both perceived disability and physical activity. We hypothesized that higher activity limitations would be associated with lower levels of physical activity. Further, we hypothesized that higher pain intensity would be associated with higher activity limitations and lower physical activity.

Method

Participants and Procedures

All study procedures were approved by the Institutional Review Board. Participants were treatment-seeking youth with acute MSK pain ages 11–17 years old and a participating parent or guardian taking part in a larger ongoing longitudinal project examining pain, sleep and psychosocial outcomes following acute MSK injury. The sample for this study were youth participating in the larger longitudinal study who completed sufficient daily diary and actigraphy data. Potentially eligible youth were identified though an automated alert system in the electronic medical record (EMR) when they received care for acute pain at an emergency department or outpatient clinic at one of two academic medical centers in the northwestern United States. Parents/guardians of potentially eligible youth were contacted via phone call to conduct an eligibility screening and interested/eligible dyads reviewed and signed consents/assents electronically.

Inclusion criteria for youth were: 1) age between 11–17 years old, 2) primary reason for seeking treatment was new onset MSK pain in trunk, back, neck or lower extremities (pain duration <30 days), and 3) and access to a smartphone to complete daily electronic diaries. Exclusion criteria included: 1) co-occurring major medical condition (e.g., diabetes, cancer), 2) previous injury at pain site in the past two years or any history of surgery at their pain location, 3) currently being treated for another chronic MSK condition (e.g., physical therapy for a pre-existing pain problem), 4) the acute MSK pain was related to serious pathology, such as an infection or disease process, 5) currently pregnant, 6) history of major surgeries or major hospitalizations (>7 day duration), 7) psychiatric hospitalization within the last year, and/or 8) major cognitive impairment or intellectual disabilities that would prevent independent completion of study activities. Additionally, both youth and parents needed fluency in verbal and written English to participate.

Within 1 month following pain onset, youth wore an Actiwatch for 7 days and completed 7 days of daily diaries each evening. Youth and parents also completed a questionnaire via REDCap providing demographic information.

Previous publications from this dataset have examined screening for parent pain-related risk factors⁹, screening for youth psychosocial risk factors⁸, and the role of perceived injustice in youth with acute pain⁶. While physical activity data was collected as part of the larger ongoing longitudinal project, it was not a primary study aim and therefore no previous publications from this data set have utilized actigraphy or daily diary data.

Measures

Demographic and Clinical Measures

Survey data was collected via REDCap questionnaires. Parents reported on family demographics (parent education, household income, and health insurance status) and child pain location, date of injury, and pain etiology. Date of injury and pain location were confirmed by electronic medical record data. Youth self-reported gender identity. Youth height and weight were collected at an in-person study visit and used to calculate body mass index (BMI).

Previous Activity Levels and Sports Engagement

Youth reported on activity engagement in the 3 months prior to injury. Specifically, youth indicated the average number of hours per week they participated in “an organized sport or physical activity over the last 3 months”. The six response options ranged from 0–1 hours to 6 or more hours per week. Youth who reported engaging in <6 hours per week were categorized as “low previous activity” and youth who engaged in ≥6 hours per week were categorized as “high previous activity” in analyses.

Youth were also asked about sports participation in the last year. The item lists 25 sports (e.g., basketball, golf, soccer, other) and youth selected all sports they engaged in. A total score tabulating the number of sports per participant was calculated.

Daily Diaries

Daily diaries assessing pain intensity, medication use, and activity limitations were sent to the youth’s smartphone at 7 pm each evening during the assessment period. If the diary was not completed at 7 pm, four subsequent hourly reminders were sent out.

Pain Intensity

Youth reported on the average pain intensity they experienced that day on a Numerical Rating Scale (NRS) of 0 (no pain) to 100 (worst pain possible) using a sliding scale to indicate a response.

Medication Use

On the daily diaries, youth reported if they took any medication for pain that day by selecting “yes” or “no”. If medication was taken, they indicated whether they took over the counter medication (e.g., ibuprofen, acetaminophen) or prescription medication (e.g., oxycodone, morphine) for pain. For analyses, a binary variable: yes/no pain medication that day was coded and included as a covariate.

Activity Limitations

Youth self-reported on their pain-related activity limitations each day using the CALI-9¹¹. On this 9-item measure, participants reported how “difficult or bothersome” activities were that day because of pain using a 5-point scale of 0 (not very difficult) to 4 (extremely difficult). The 3-item CALI-9 Active subscale measures perceived difficulty engaging in vigorous physical activities (e.g., sports, running, walking 1 or 2 blocks) because of pain. The 6-item CALI-9 Routine subscale assesses perceived impairment in routine activities due to pain, such as eating, hanging out with friends, and schoolwork. Subscale scores range from 0–50 with higher scores indicating greater impairment. The diary measure has been previously validated in a chronic pain sample¹⁴. Both the CALI-9 Active (α= 0.9) and CALI-9 Routine (α= 0.8) subscales showed good to excellent reliability in this sample.

Objectively Measured Physical Activity: Actigraphy

Daily physical activity was monitored using an Actiwatch Spectrum¹ worn by youth on their non-dominant wrist for a 7-day period. Movement detection was monitored in 1-minute epochs; this was the pre-determined epoch length in the original study because actigraphy monitoring was being conducted to assess sleep patterns as well. Participants were instructed to remove the Actiwatch while bathing and swimming, and to push a button on the Actiwatch to indicate sleep and wake times. Daytime interval was manually scored by identifying times of sleep onset and offset where 10 continuous minutes below the sleep threshold set by the software indicated sleep onset. Periods of time where the Actiwatch was not worn, as indicated by the participant or by 10+ consecutive minutes where activity counts were 0, were excluded from the total daytime interval. Movement was measured as the mean number of activity counts per 1-minute epoch (mean activity), highest level of activity counts in a 1-minute epoch (peak activity), proportion of the daytime interval spent in sedentary (≤40 counts), light (40–1420 counts), and moderate to vigorous activity (≥1420 counts)^{1, 13, 17–18, 23}. Number of activity bouts in a day were also calculated. Consistent with previous research¹⁸, an activity bout was operationalized as a period of sustained activity where over the course of at least 10 minutes, 80% of epochs were higher than 500 activity counts per minute. As such, activity bouts did not include sedentary activities, but included a range of light physical activity (e.g., walking home from school) and more vigorous activity (e.g., playing soccer). Activity thresholds were chosen in an effort to adequately assess sleep patterns, be consistent with previous research^{13, 17–18, 23}, and were originally chosen based on Philips Respironics thresholds for wrist actigraphy in youth^{1, 18}.

Statistical Analyses

Sample characteristics were summarized using descriptive statistics. Daily pain intensity and self-reported activity limitations from diaries and objective activity from Actiwatch devices were described using mean, standard deviation, and range. The relation between self-reported activity limitations and objective physical activity measures was examined using repeated measure correlations (RMC) to capture same-day associations within individuals. This analysis was selected as most correlation measures assume independence, but this assumption is violated by having multiple days of diary data per participant. Repeated measure correlation is an application of ANCOVA that allows assessment of the relationship between two continuous variables (i.e., physical activity and activity limitations) after grouping the data by participant and controlling for between participant variance^{3, 4, 7}. As values approach −1 or 1, the RMC demonstrates homogeneity of the relationship between these two variables across participants and strength of the association. While both Pearson correlations and RMC use pairwise deletion to handle missing data, RMC often has increased power over aggregating data to use traditional correlation methods^{3, 4, 7}. Pearson correlations were then used to examine the association between weekly means of self-reported activity limitations and objective physical activity measures to determine the strength of the relationship between activity limitations and activity levels. Correlation methods were chosen over other modeling strategies to focus on the association between self-reported activity limitations and objective activity without implying or inferring a directional relationship. The use of RMC examines daily associations within individuals, while average associations using Pearson correlations examine associations between individuals.

Mixed-effects models (MEM) examined the association between daily pain intensity from diaries and same-day self-reported activity limitations and physical activity variables. Outcome variables included CALI Active scores, CALI Routine scores, mean physical activity, peak activity, time sedentary, time in light activity, and number of activity bouts. Pearson correlations were used to examine the association between continuous covariates and primary study variables (pain intensity, CALI Active subscale, CALI Routine subscale, mean activity, peak activity, time sedentary, time in light activity, and number of activity bouts). For dichotomous covariates, nonparametric Wilcoxon rank sum tests with continuity corrections were used given that the primary study variables were not normally distributed. The MEMs used random intercepts and accounted for missing data using maximum likelihood. Unadjusted models predicted each outcome with only average pain intensity. Adjusted models used youth age, gender, BMI, daily medication use, previous activity prior to injury (<6 versus ≥6 hours), and days since injury as additional covariates, due to the potential relation of these variables with youth pain outcomes. All analyses were completed using R v4.2.2¹⁶. Repeated measures correlations used the “rmcorr” package³ and mixed effects models used the “lme4” package⁵. Deidentified data and code for these analyses are available upon request.

Results

Descriptives

Of the 216 participants that provided data for the study, 211 youth completed ≥ 3 days of diaries and 190 participants completed ≥ 3 actigraphy days. The final sample for analysis was 176 youth ≥3 overlapping diary and actigraphy days (M= 5.8 days, SD= 1.23, range= 3–7). Sixty-four (34%) participants completed diary and actigraphy on all 7 days, with an additional 28% having overlapping diary and actigraphy data on 6 out of 7 days. There was a total of 1146 completed diaries across the 176 participants.

Participants

Participant and family demographics are reported in Table 1. The 176 youth were 53% female, predominantly non-Hispanic (77%), and an average age of 14.4 (SD= 1.9) years old. Youth in this sample identified as White (66%), multi-racial (14%), Asian (6%), Black/African American (5%), Native American/Alaskan Native (1%), or as another racial category not listed (5%). Nearly half of the families had a household income higher than $150k, with 17% of the sample having a household income ≤$50k. The majority of youth (82%) reported having lower limb pain (leg, knee, ankle, or foot). Injuries were primarily sports-related (66%) and at the time of data collection, participants were on average 21.7 days (SD= 7.1, range: 5–35) from pain onset.

TABLE 1.

Participant demographic characteristics.

Characteristics (N = 176)	M (SD) or %	Range
Age (years)	14.4 (1.9)	11.0 – 18.0
Sex in medical records
Female	53%
Male	47%
Child reported gender
Female	53%
Male	46%
Non-binary	1%
Child race
White	66%
Asian	6%
Black/African American	5%
Native American/Alaskan Native	1%
Multi-racial	14%
Other	5%
Missing	3%
Child ethnicity
Hispanic or Latino	18%
Not Hispanic or Latino	77%
Missing	2%
Child’s medical insurance
Private	76%
Medicaid/State	21%
Other	2%
Missing	2%
BMI (kg/m2)	22.9 (6.1)	14.6 – 49.2
Time since start of pain (days)	21.7 (7.1)	5.0 – 35.0
Time since medical visit (days)	17.2 (7.2)	3.0 – 31.0
Pain location
Back	6%
Neck	2%
Shoulder	1%
Sternum	1%
Rib	1%
Hip	7%
Knee	20%
Leg	9%
Ankle	34%
Foot	19%
Cause of pain
Sports	66%
Non-sport	31%
Unknown	3%
Activity during 3month prior to injury
Low (<6 hours)	48%
High (≥6 hours)	52%
Annual household income
≤ $25k	7%
$25–50k	10%
$50–80k	13%
$80–120k	13%
$120–150k	10%
≥ $150k	46%
Missing	2%
Highest parental education level
≤ High school	6%
Vocational school/some college	13%
College degree	43%
Graduate/professional degree	37%
Missing	1%

Notes: Percents may not add up to 100% due to missing data and/or rounding.

Pre-Injury Activity Engagement

Approximately half of the participants (52%) reported spending 6 or more hours per week in an organized sport or physical activity over the past three months. Youth reported participating in an average of 1.9 (SD=1.4) sports over the past year. The most commonly reported sports were soccer (n=50), basketball (n=41), track and field (n=37), and cross country (n=25).

Pain and Medication Use

Participants reported pain on an average of 47.8% (SD= 35.9) of the days during the diary period. For participants who reported experiencing pain NRS >0, daily pain intensity averaged 33.4 (SD= 20.7, range= 0.0–84.4). Pain intensity was not associated with age (r= .020, p> .05) or gender (W= 3540, p= .343). Youth reported using medication on an average of 10.2% of diary days (SD= 20.5).

Self-Reported Activity Limitations

Table 2 presents diary data. Participants reported moderate-high levels of pain-related activity limitations on the CALI-9 Active subscale (M= 40.3, SD= 31.5). All items on the Active subscale were frequently endorsed; difficulty running (74.1% of total diary days), difficulty with sports (70.3%), and difficulty walking 1–2 blocks (49.5%). Overall, youth reported experiencing fewer activity limitations in routine activities but there was a notable range of reported impairment (M= 8.6, SD= 12.2, range= 0–68.5). Most frequently endorsed items on the CALI-9 Routine subscale were difficulty doing things with friends (34.3%), being up all day without a nap or rest (26.5%), and riding in the school bus or car (21.8%). CALI-9 scores were not associated with age (r= .043–.047, p> .05) or gender (W= 3355–3519, p= .439).

TABLE 2.

Daily pain and activity descriptions.

Characteristics (N = 176)	M (SD) or %	Range
Days with pain (%)	47.8 (35.9)	0.0 – 100.0
Average daily pain	33.4 (20.7)	0.0 – 84.4
Days with medication (%)	10.2 (20.5)	0.0 – 100.0
Subjective activity limitations
CALI Routine	8.6 (12.2)	0.0 – 68.5
CALI Active	40.3 (31.5)	0.0 – 100.0
Objective activity
Mean activity (counts/minute)	311 (110)	79 – 959
Peak activity (counts/minute)	2299 (698)	1116 – 5857
Number of bouts	4.4 (2.7)	0.0 – 13.3
Duration of bouts (minutes)	94.1 (70.3)	0.0 – 501.0
Percent of time sedentary	21.9 (9.4)	2.9 – 58.3
Percent of time in light activity	76.2 (8.6)	41.7 – 92.7
Percent of time in MVPA	2.0 (2.7)	0.0 – 22.5

Notes: CALI = Child Activity Limitations Interview; MVPA = moderate-vigorous physical activity

Physical Activity Data

Data on physical activity assessed via actigraphy is presented in Table 2. Participants spent the most time in light activity during the day (M= 76.2%, SD= 8.6), followed by sedentary activity (M= 21.9%, SD= 9.4), and very little time in moderate-vigorous physical activity (M= 2.0%, SD= 2.7). Age was negatively associated with average physical activity, with older age associated with lower mean activity (r= −.392, p< .001), lower peak activity (r= −.310, p< .001), and fewer activity bouts (r= −.349, p< .001). Physical activity was also associated with gender such that males had higher mean activity (W= 5013, p< .001), higher peak activity (W= 5264, p< .001), and higher number of activity bouts (W= 4795, p< .001) than females. Males also spent less time in sedentary activities than females (W= 2776, p= .001).

Aim 1: Associations Among Self-Reported Activity Limitations and Objectively Assessed Physical Activity

Repeated measures correlations were used to examine daily associations between physical activity measures and self-reported activity limitations on both active and routine subscales (see Table 3). Contrary to hypotheses, correlations were not significant (r= −.05–0.04, p> .05) indicating that at the daily level, self-reported activity limitations in both active and routine domains were not related with objectively assessed physical activity (actigraphy) at the daily level. This means that at the daily level, there was significant heterogeneity with some participants reporting having low physical activity and reporting corresponding activity limitations and others engaged in physical activity but reported low or no activity limitations on their diaries.

TABLE 3.

Repeated measures correlations and Pearson correlations of averages between subjective and objective activity with 95% confidence intervals.

Objective activity measure	Repeated Measures Correlations		Pearson Correlations
	CALI Routine	CALI Active	CALI Routine	CALI Active
	rrm (95% CI)	rrm (95% CI)	r (95% CI)	r (95% CI)
Mean activity (counts/minute)	0.00 (−0.06, 0.07)	−0.02 (−0.08, 0.05)	−0.16* (−0.30, −0.01)	−0.20** (−0.34, −0.06)
Peak activity (counts/minute)	0.04 (−0.02, 0.11)	0.03 (−0.03, 0.10)	−0.18* (−0.32, −0.04)	−0.29*** (−0.42, −0.15)
Number of activity bouts	0.02 (−0.04, 0.09)	−0.05 (−0.12, 0.01)	−0.07 (−0.22, 0.08)	−0.11 (−0.25, 0.04)
Percent of time sedentary	0.01 (−0.05, 0.08)	−0.03 (−0.10, 0.03)	0.18* (0.03, 0.32)	0.08 (−0.07, 0.22)
Percent of time in light activity	−0.01 (−0.08, 0.05)	0.04 (−0.03, 0.10)	−0.14 (−0.29, 0.00)	−0.02 (−0.17, 0.13)

Notes:

^*p<0.05;

^**p<0.01;

^***p<0.001

However, consistent with our hypotheses, when physical activity assessed via actigraphy and self-reported activity limitations were averaged for each participant across the diary period, Pearson correlations were significant such that higher self-reported activity limitations on the Routine subscale were significantly associated with lower mean physical activity levels (r= −.159, p< .05), lower peak activity (r= −.184, p< .05), and higher amount of time spent in sedentary activity (r= .176, p< .05). Similarly, greater self-reported activity limitations on the Active subscale averaged across the diary period were significantly associated with lower mean physical activity levels (r= −.204, p< .01) and lower peak activity (r= −.291, p< .001).

Aim 2: Associations among Pain Intensity and both Self-Reported Activity Limitations and Objectively Assessed Physical Activity

Results of mixed effects linear models examining associations among daily pain intensity and both self-reported activity limitations and physical activity are presented in Table 4. Coefficients are presented for both unadjusted and adjusted models, along with their 95% CI. A full model summary with covariates is presented in Supplementary Table 1.

TABLE 4.

Summary of linear mixed models examining average daily pain (0–100) as a predictor of same day activity.

Dependent variable	Unadjusted model	Adjusted model
	β (95% CI)	β (95% CI)
CALI Routine (0–100)	0.08*** (0.05, 0.10)	0.07*** (0.04, 0.09)
CALI Activity (0–100)	0.23*** (0.18, 0.28)	0.19*** (0.15, 0.24)
Mean activity (counts/minute)	0.46** (0.17, 0.74)	0.48***(0.20, 0.77)
Peak activity (counts/minute)	2.40 (−0.08, 4.84)	2.27 (−0.19, 4.73)
Number of activity bouts	0.01** (0.00, 0.02)	0.01*** (0.01, 0.02)
Percent of time sedentary (0–100)	−0.04** (−0.07, −0.01)	−0.04** (−0.07, −0.01)
Percent of time in light activity (0–100)	0.04* (0.01, 0.06)	0.04* (0.01, 0.06)

Notes: All models use random intercepts. Adjusted models controlled for age, sex, BMI, time since start of pain, daily medication use, and activity levels prior to injury.

^*p<0.05;

^**p<0.01;

^***p<0.001

Consistent with our hypothesis, mixed effects linear models showed that daily pain intensity was significantly associated with same-day self-reported activity limitations. On days where participants reported having higher pain intensity, they also reported having more limitations in both routine (β= .08, SE= .01) and active (β= .23, SE= .02) activities. For every 10-point increase in pain intensity (on a scale of 0–100) there was a 0.7-point increase in CALI Routine (range= 0–68.5) scores and a 2.3-point increase in CALI Active (range= 0–100) scores.

Contrary to our hypothesis, on days where participants reported having higher pain intensity, their mean physical activity levels were higher (β= .46, SE= .14), they had a higher number of activity bouts (β= .013, SE= .004), spent more time in light activity (β= .04, SE= .01), and spent less time in sedentary activity (β= −.04, SE= .01). A 10-point increase in pain intensity (on a scale of 0–100) was associated with an increase of 4.5 counts/minute in mean activity (range= 79.3–958.9), a 0.1 increase in number of activity bouts (range= 0–13.3), a 0.4% increase in time spent in light activity (range= 41.7–92.7), and a 0.4% decrease in time spent in sedentary activities (range= 2.9–58.3). These associations remained significant when controlling for age, gender, BMI, daily medication use, activity prior to injury, and time since start of pain.

Discussion

This study was the first to examine associations among two domains of physical function (self-reported activity limitations and objectively assessed physical activity) outlined in Wilson & Palermo’s (2012) conceptual model, in youth experiencing acute MSK pain. Results revealed self-reported activity limitations were associated with physical activity averaged across the week, such that youth who reported more pain-related impairment engaged in lower levels of objectively assessed physical activity. However, daily associations among self-reported activity limitations and physical activity were not significant. Failure to detect daily associations among self-report activity limitations and objective activity is likely related to the high heterogeneity across participants. For some youth, daily associations were very strong, and for others, there was no association between daily actigraphy data and diary reports of activity limitations on certain days. Further, daily associations were examined within individuals, while average associations were examined between individuals. This means that daily associations examined day-to-day variation within individual participants, while average associations across the week captured rank ordering effects between participants. Thus, while individual daily variation in self-reported activity limitations and objectively measured physical activity was not significant, participants who had higher self-reported activity limitations overall engaged in lower levels of objectively assessed physical activity overall.

Analyses also examined daily relationships between pain and activity. On days where youth reported higher pain intensity on evening diaries, youth had higher activity assessed via actigraphy during the day, while also reporting greater daily activity limitations. The fact that pain was associated with both higher activity and higher activity limitations suggest that at least a portion of youth are continuing physical activity engagement despite reporting difficulties on the CALI-9. Activity during the day may also have led to increased pain reports that evening and increased reports of activity engagement on the CALI-9, which assesses difficulty in activity engagement rather than actual level of activity participation. Many youth with acute pain may also still be able to engage in Routine activities (e.g., eating, schoolwork) even if more vigorous activities (e.g., running, walking 1–2 blocks) are impacted by pain; this may be particularly relevant for this sample of youth with majority lower extremity MSK injuries.

Importantly while the magnitude of the association between daily pain intensity and self-reported activity limitations and objectively measured physical activity was small, it may have important implications for recovery and pain persistence, as these effects may have a cumulative effect over time. This study looked at a critical period of subacute recovery (<30 days from injury), where even small associations between pain, physical activity, and activity limitations may cumulate over time to impact recovery and pain persistence.

Findings also support Wilson and Palermo’s (2012) conceptual framework that perceived activity limitations and objectively assessed physical activity are related, yet distinct, constructs of physical functioning and there are potentially important reasons to assess both. While both perceived activity limitations and physical activity were related to pain intensity in this sample of youth with acute MSK pain, researchers should thoughtfully consider which assessment of functioning best fits their research question or study sample. The decision to focus on self-reports of activity limitations or actigraphy to assess physical activity in youth post-injury will depend on whether researchers believe perceived disability or movement is more directly related to their study questions and constructs. Future research should also consider the impact of physiological functioning (e.g., fitness, functional capacity) on pain outcomes, which is the third broad dimension of physical functioning proposed by Wilson and Palermo (2012) and is not examined in the current study. Understanding how physiological functioning is associated with perceived disability and physical activity may also have important implications for youth outcomes during the post-injury recovery period.

This study also contributes to the limited data on objectively measured physical activity in an acute MSK injury sample. Prior to this study, little was known regarding how an acute MSK injury impacts daily physical activity levels. Specifically, our findings showed youth spent over 75% of their day in light activities, and 21% of their time in sedentary activities. These youth are more active than youth who were two weeks post-surgery (69% light, 31% sedentary activity)¹⁸. Youth in our sample also had lower levels of mean (310.6 counts/minute) and peak (2299.3 counts/minute) activity than previous studies of youth with chronic pain (mean= 436, peak= 2899¹⁷; mean= 464.9, peak= 3082.0²³) and comparison samples of pain-free youth (mean= 507, peak= 3637¹⁷; mean= 517.8, peak= 3676.4²³). Although higher physical activity in our sample was associated with pain intensity, previous research has shown that higher physical activity is associated with more positive pain outcomes in samples of youth with chronic pain^{17, 23}, as well as more positive outcomes over time in post-surgical samples¹⁸. It is therefore unclear whether higher physical activity during the acute pain period may be associated with fewer activity limitations over time, despite being associated with higher pain intensity in the acute recovery period. Given the current study focused on associations between pain intensity and physical activity within one month following injury, examining how physical activity in the acute pain period predicts recovery and future physical activity over longer time periods is an important next step.

This study has several limitations that are important when contextualizing these findings. First, this study only examined associations between self-reported activity limitations, physical activity, and pain intensity at one time point during the acute recovery period. The one-week snapshot of functioning captured in this study may not reflect functioning across the entire acute pain period. Further, participants were asked to recall their average pain intensity during the day at one point in the evening, rather than rating pain when it was experienced. Future research could use ecological momentary assessment to minimize recall bias and better capture variations in pain throughout the day. Moreover, while all youth participated in assessment within one month of injury, there was variation in time since injury. It is possible that the variation in time since injury may impact these results, particularly given some youth who reported minimal pain and activity limitations. Inclusion of a control group without pain would also allow for more robust comparison of how physical activity and activity limitations may differ during the acute pain period following musculoskeletal injury.

While inclusion criteria specified that participants were generally healthy (e.g., no major comorbid medical conditions), there was no assessment of physical functioning prior to injury. It is possible that physical functioning and previous engagement in physical activity prior to injury may impact the relationship between pain, physical activity, and activity limitations during the recovery period. While the current study used activity levels prior to injury as a covariate, the data was collected via a single item. Future research should examine how additional factors such as cardiovascular fitness, physical activity participation in the context of organized team sports, or other social factors (e.g., income, peer supports, social status) impact recovery trajectories over time.

In this study, youth were asked to remove their Actiwatch when swimming. This may have resulted in an underestimation of physical activity for some youth. At the same time, although the Actigraphy thresholds used in this study were consistent with previous research^{13, 17–18,23}, the low threshold for light activity may have resulted in an overestimation of physical activity. This study used software-defined activity thresholds¹, and using other criteria to define light and sedentary activity may be more sensitive. Continuous measures of physical activity, rather than the use of activity thresholds, may also allow for a closer examination of the relationship between physical activity and self-reported activity limitations. In addition to examining same-day associations between pain and activity, future research could leverage cross-lagged models to examine how pain impacts next-day activity and vice versa. Future research may also benefit from differentiating between groups of participants who experience strong correlations between their physical activity and self-reported activity limitations on a given day and those with no association between objective physical activity and self-reported activity limitations. Finally, this sample was heterogeneous in age and race and ethnicity. While age was included as a covariate in mixed linear models, race and ethnicity variables were not examined due to limited sample size within groups.

Taken together, results suggest both self-report and objective measures provide important data on activity engagement and pain-related disability in the acute pain period. While both self-reported activity limitations and objectively assessed physical activity were associated with daily pain intensity, they represent two distinct, yet related, facets of physical functioning. This study serves to expand our understanding of perceived activity limitations and objectively assessed physical activity in a sample of youth with acute MSK injuries, and may inform future research assessing recovery and predicting pain persistence over time.

Highlights.

• Same-day correlations between activity limitations and physical activity were not significant

• Averaged across the week, greater activity limitations were associated with lower activity

• Higher daily pain intensity was associated with greater self-reported activity limitations

• Conversely, higher daily pain was associated with higher activity on actigraphy

• Activity limitations and activity are two distinct, yet related, aspects of physical functioning

Data Availability

The data, code, and materials necessary to reproduce the analyses presented here are publicly accessible and are available from the first author upon reasonable request.