Patient-Reported Outcomes Measurement Information System (PROMIS) Outcomes: Higher odds of adverse mental health when physical function is impaired

Abstract

Purpose:

Balance and vestibular disorders have a profound impact on quality of life. Anxiety (ANX) and depression (DEP) are common with dizziness, vertigo, or imbalance. It is unclear whether self-reported anxiety or depression depend on perceived physical function for individuals participating in vestibular rehabilitation (VPT). We hypothesized that individuals with worse physical function would be more likely to report abnormally high anxiety and/or depression levels.

Participants:

170 individuals referred for vestibular rehabilitation (113 female, 57 males, age 63.7 (21))

Methods:

A retrospective chart review extracted age, Patient-Reported Outcome Measurement Information System (PROMIS) scores (ANX, DEP, physical function) and primary diagnoses [benign paroxysmal positional vertigo (, n=42), unilateral vestibular hypofunction (, n=39), bilateral vestibular hypofunction (, n=14), concussion/head injury (, n=7), dizziness/vertigo (, n=37), imbalance (n=31)]. Average PROMIS scores and percentage of abnormal scores were calculated for physical function, ANX, and DEP scores and reported using descriptive statistics. Logistic regression was performed to separately examine the odds of abnormal ANX and DEP based on abnormal physical function while controlling for age and sex, on the entire dataset and on diagnosis subgroups.

Results:

Individuals referred to vestibular physical therapy with self-reported abnormal physical function were more likely to have abnormal ANX (OR 5.1, p< 0.001) or DEP (OR 3.10, p=0.002). Older adults were less likely to have abnormal ANX (OR = 0.96, p = 0.002). For individuals experiencing BPPV (n=42), those with abnormal physical function are more likely to report ANX (OR 9.9, p=0.009). For individuals with UVH (n=39), those with abnormal physical function were more likely to report ANX (OR 10.1, p=0.008) or DEP (OR 9.9, p=0.010).

Conclusion:

Self-reported abnormal physical function corresponds to a higher incidence of abnormal ANX and DEP for individuals referred to VPT. Individuals with benign proximal positional vertigo and unilateral vestibular hypofunction were approximately 10-times more likely to experience abnormally high ANX or DEP compared to all other primary diagnosis. Clinicians seeing patients in VPT should be screening for anxiety and depression especially for those with self-reports of impaired physical function.

Introduction:

Dizziness is among the most common symptomatic complaints with a reported lifetime prevalence between 15% and 35% ^(1,2). Connections between mental health and chronic to episodic bouts of dizziness in both civilian and military populations are well documented in the literature ^(3,4,5) Symptoms of dizziness and vertigo are often associated with peripheral vestibular dysfunction but are also present in conditions like altered mental health ^(6,7,8). Close to half of the patients who present to specialized dizziness clinics also present with significant psychiatric comorbidity such as anxiety and depression disorders ⁽⁹⁾. This relationship also exists across the lifespan. Wiltink et al. ⁽¹⁰⁾ reported comorbid anxiety and/or depression were related to a higher degree of subjective impairment in individuals aged 14–90 with primary diagnosis of dizziness. Furthermore, healthcare utilization for patients who present with dizziness in ambulatory settings is at 8.8 per 1000 visits in the United States ⁽¹¹⁾.

Anxiety disorders and depression are complex and typically individualised in context to the patient and their situation ⁽¹²⁾. A growing body of evidence suggests both functional and structural links between the vestibular system and emotional/cognitive brain areas ⁽¹³⁾. Close relationships between the vestibular and cerebellar systems for the elaboration and coordination of emotional, cognitive, and visceral responses have also been observed ⁽¹⁴⁾.

Mental health conditions mediate the experience of dizziness, even in the absence of vestibular or other central nervous system pathology ⁽¹⁵⁾. Specifically, the parabrachial nucleus is a site of convergence of vestibular information processing and somatic and visceral sensory information processing ^{(16; 17)}. These pathways appear to be involved in movement or in the action of positional avoidance. Both anxiety and conditioned fear are often observable in individuals with dizziness. Reciprocal connections exist between the thalamocortical pathway and limbic pathways responsible for emotional regulation within the cerebellum ^{(18; 19)}. These shared neural pathways likely explain the frequent balance-related anxiety experienced by individuals with vestibular or balance disorders ^{(20; 21)}. A recent meta-analysis of MRI studies confirmed areas of converging signals in the upper brainstem and cortical areas between ascending vestibular pathways and pathways involved in emotional regulation ⁽²²⁾.

Individuals with persistent psychological symptoms, such as anxiety or depression, also appears to have the higher levels of self-reported physical handicap ⁽²³⁾. Up to 68.25% of individuals with dizziness as a primary complaint, can develop acute psychiatric disorder after the onset of vestibular disease ⁽²⁴⁾. For many patients with vestibular disease these psychiatric symptoms tend to resolve once the dizziness and/or vertigo resolves ⁽²⁵⁾. Godemann et al., ⁽²⁶⁾ found that 20% of individuals who presented with acute vertigo had lingering symptoms of dizziness six months later but found no correlation between persistent vertigo complaints and vestibular function. ‘This information highlights that the persistent vertigo experienced by these individuals could result from acquired anxiety triggered by their initial experience with vertigo.

Negative health related quality of life impacts physical function and creates emotional distress in individuals with self-reporting dizziness ⁽²⁷⁾. Individuals who experience dizziness/vertigo in combination with anxiety/depression experience longer recovery times compared to those without anxiety/depression ⁽²⁸⁾. Sensations of dizziness and vertigo impair participation in and performance of activities of daily living, negatively impacting psychological health ⁽⁸⁾. Individuals with acute (lasting at least 24 hours) or chronic (>1 year of persistent dizziness/imbalance) vestibular dysfunction experience both psychological distress and impaired physical performance in relation to their premorbid tasks of daily living ^(29,30). However, it is less clear whether individuals with vestibular disease or balance problems who endorse difficulty with common daily activities are more likely to report abnormally high anxiety or depression. Further, whether this relationship differs across vestibular diagnoses is also unknown. We hypothesize that individuals who present for vestibular rehabilitation and self-report abnormal physical function will be more likely to experience elevated anxiety and depression levels.

Methods

We performed a retrospective chart review on adults referred to a tertiary care outpatient clinic by a team of Otolaryngologists, with primary complaints of dizziness, vertigo or imbalance who also completed computerized patient reported outcome measures for mental health (anxiety, depression), and physical function. We extracted primary treating diagnosis, age, sex, scores within the domains of mental health (anxiety, depression), and physical function for individuals who attended at least one physical therapy visit between Mar 1, 2021, and April 1, 2022. The University of Rochester IRB determined that this retrospective chart review met federal and university criteria for exempt status.

Primary diagnoses identified through chart review for this study were as follows: benign paroxysmal positional vertigo (BPPV), unilateral vestibular hypofunction (UVH), bilateral vestibular hypofunction (BVH), concussion/head injury (mTBI), dizziness/vertigo (dizzy), and imbalance. In total, 170 participants (113 female, 57 males) met the inclusion criteria, see Table 1 for demographic information on this cohort.

Table 1.

Demographics for all individuals including diagnostic counts.

Number of Subjects	170
Male	53 (31.2%)
Female	113 (66.5%)
Average Age	63.7 (21)	Range (20–90)
Primary Diagnosis
Benign Paroxysmal Positional Vertigo (BPPV)	42 (24.7%)
Unilateral Vestibular Hypofunction	39 (22.9%)
Bilateral Vestibular Hypofunction	14 (8.2%)
Concussion/Head Injury	7 (4.1%)
Dizziness / Vertigo	37 (21.7%)
Imbalance	31 (18.2%)

BPPV was diagnosed based on symptoms of vertigo with associated nystagmus according to established diagnostic criteria while in the Dix-Hallpike or Supine Roll test positions ⁽³¹⁾. UVH was diagnosed based on caloric asymmetry > 25% ⁽³²⁾, unilateral VOR gain < 0.68 ⁽³³⁾, or presence of consistent overt saccades during ipsilesional head impulses ⁽³⁴⁾ as many individuals declined video head impulse testing during the COVID-19 pandemic. BVH was diagnosed according to the Barany Criteria based on total bilateral caloric response < 12 degrees/second, or bilateral VOR gain < 0.6, or imbalance and oscillopsia with overt saccades during clinical head impulse testing indicating probable bilateral vestibulopathy ⁽³⁵⁾. Concussion/head injury was defined as a traumatic injury affecting the brain, brought on by a direct blow to the head or indirectly via forces exerted on the body without extended loss of consciousness, consistent with the clinical practice guideline for physical therapist evaluation/treatment of concussion ⁽³⁶⁾. The treating physical therapist was never the first provider to make a diagnosis of concussion/head injury as this diagnosis was made by physicians in the emergency department, neurology clinic, or orthopaedic clinic. A primary diagnosis of dizziness/vertigo was made in the absence of definitive peripheral vestibular dysfunction when the individual’s primary symptom was dizziness or vertigo, examples may include vestibular migraine, mal de débarquement syndrome, or other non-stroke/non-TBI central dizziness. A primary diagnosis of imbalance was made in the absence of definitive peripheral vestibular dysfunction without symptoms of dizziness or vertigo and in the absence of another cause of imbalance such as diabetes, neuropathy, or musculoskeletal injury.

Patient Reported Outcomes Measurement Information System (PROMIS)

The Patient-Reported Outcomes Measurement Information System (PROMIS) measures were used to collect standardized self-reports related to patient symptoms. PROMIS measures patients’ self-reported symptoms using adaptive computerized testing based on item response theory ⁽³⁷⁾. This adaptive design retains an elevated level of score confidence while shortening administration time ⁽³⁷⁾. All PROMIS scores are standardized and expressed as t-scores with a population mean of fifty and a standard deviation of 10 ⁽³⁸⁾. Cut-off criteria for identifying abnormal responses were specified as one half of a standard deviation (>5 points) away from the population mean ⁽³⁹⁾.

The PROMIS physical function scale, is well correlated with other legacy patient reported measures of physical function such as the Lower Extremity Functional Scale (LEFS) and Short Form 36-item Health Survey (SF-36) with significantly fewer questions and was more generalizable ⁽⁴⁰⁾. PROMIS physical functional measures are easily interpreted, have a low time burden for the patient, have low measurement error, and better content validity than traditional outcome measures ⁽⁴¹⁾. The PROMIS subscales assessing mental health conditions, such as depression and anxiety, have also been compared with well-established legacy depression and anxiety screening measures, such as the Patient Health Questionnaire-9 (PHQ-9) depression instrument and Generalized Anxiety Disorder-7 (GAD-7) instrument. Both cut off scores for classification of depression and anxiety, along with severity of the condition, have been observed to crosswalk between those legacy measures and PROMIS T-Scores ^(42,43).

Data Analysis

PROMIS scores were categorized as normal or abnormal based on whether individual scores exceeded one-half standard deviation, (5 points) from the standardized mean score of 50 ⁽⁴⁴⁾. For ANX and DEP the cut-off score for abnormal was 55 (higher scores mean worse symptoms) and for physical function the cut-off score was 45 (lower scores mean greater impairment). The percentage of abnormal scores was calculated for physical function, ANX, and DEP scores and reported using descriptive statistics. Logistic regression was performed to separately examine the odds of abnormal ANX and DEP based on abnormal physical function while controlling for age and sex, on the entire dataset. This analysis was repeated as an exploratory post-hoc analysis on each diagnosis subgroup. Different subgroups did not have consistent sample sizes, and no direct comparison was made across subgroups. All analyses were conducted using STATA version 14, and alpha was specified at 0.05 for the combined logistic regression since they are independent analyses. For post-hoc analyses for each diagnostic category we used a corrected alpha of 0.01.

Results

PROMIS scores for the cohort (average and standard deviation) are presented in Table 2 for the combined cohort and separated by diagnostic subgroups. Abnormal anxiety was reported by 54.6% of individuals. Abnormal depression was reported by 42.1% of individuals. Abnormal physical function was reported by 42.1% of individuals. After separating by diagnoses, 31.0% of individuals with BPPV had anxiety while 26.2% reported depression and 52.4% reported decreased physical function. For individuals with UVH, 46.2% reported anxiety, 41.0% reported depression and 61.5% reported decreased physical function. Of the individuals with BVH, 50% reported abnormal anxiety, 21% reported abnormal depression and 78.6% reported decreased physical function. For individuals with mild traumatic brain injury, 71.4% reported higher than normal anxiety, 57.0% reported depression and 42.9% reported decreased physical function. For individuals, whose primary diagnosis was dizziness, 54.1% reported anxiety, 43.2% reported depression and 56.8% reported decreased physical function. Finally, for individuals with imbalance as a primary diagnosis, 51.6% reported anxiety, 29.0% reported depression and 61.3% reported reduced physical function.

Table 2.

Percentage of individuals reporting abnormal levels of anxiety, depression, and physical function as an entire cohort and also by diagnostic group.

PROMIS Metric	Combined Diagnoses
Anxiety	54.6%
Depression	51.5%
Physical Function	42.1%
	Individual Diagnoses
	BPPV	UVH	BVH	TBI	Dizzy	Imbalance
Anxiety	31.0%	46.2%	50.0%	71.4%	54.1%	51.6%
Depression	26.2%	41.0%	21.0%	57.0%	43.2%	29.0%
Physical Function	52.4%	61.5%	78.6%	42.9%	56.8%	61.3%

Logistic Regression - Combined groups

After controlling for age and sex, individuals in the total sample size who reported abnormal physical function were 5-times more likely to have abnormal anxiety (OR 5.1, p< 0.001) and 3-times more likely to report abnormal depression (OR 3.10, p=0.002), see Table 3. Older adults were less likely to have abnormal anxiety (OR = 0.96, p = 0.002), corresponding to a 9.6 % reduction in the odds of anxiety for each decade.

Table 3.

Logistic regression results.

Dependent Variable	Independent Variables	Odds Ratio	p value	95% CI
ANX	physical function	5.1 *	< 0.001	2.5–10.3
	Sex	1.3	0.503	0.6–2.6
	Age	0.96 *	0.002	0.94–0.99
DEP	physical function	3.1 *	0.002	1.5–6.3
	Sex	1.2	0.601	0.6–2.4
	Age	0.98	0.124	0.96–1.00

^*indicates significant Odds Ratio. n = 170 for each analysis.

phys_fxn – physical function PROMIS score

Logistic Regression – Subgroup analyses

Our subgroup analyses demonstrated that individuals experiencing BPPV (n=42) who reported abnormal physical function, were 10-times more likely to report ANX (OR 9.9, p=0.009), see Table 4 for details. For individuals with UVH (n=39), those with abnormal physical function were 10-times more likely to report either abnormal ANX (OR 10.1, p=0.008) or abnormal DEP (OR 9.9, p=0.010). Significance was observed for several additional analyses; however, after alpha correction these effects became non-significant. Females, with a primary diagnosis of dizziness/vertigo (n=37), were 8-times more likely to report abnormal ANX (OR 7.9, p=0.036) compared to males, but there was no significant relationship between physical function and either ANX or DEP in this cohort. Individuals with a primary diagnosis of imbalance (n=31) and abnormal physical function were 11-times more likely to report abnormal ANX (OR 11.1, p=0.018) or abnormal DEP (OR 10.9, p=0.048). Abnormal physical function was not significantly related to ANX for individuals with BVH (OR 0.64, p=0.79). Logistic regression was not performed in BVH subgroup for dependent variable DEP as there were no individuals who reported normal DEP and normal physical function.

Table 4.

Logistic regressions separated by diagnostic category.

Primary Diagnosis	Dependent Variable	Independent Variables	Odds Ratio	p value	95% CI
BPPV (n = 42)	ANX	physical function	9.9 *	0.009	1.8–55.8
		Sex	1.5	0.583	0.34–6.6
		Age	0.98	0.527	0.00–24.0
	DEP	physical function	3.3	0.1.22	0.73–15.1
		Sex	1.6	0.530	0.37–7.0
		Age	1.01	0.808	0.95–1.07
UVH (n = 39)	ANX	physical function	10.2 *	0.008	1.8–56.4
		Sex	0.34	0.218	0.06–1.9
		Age	0.96	0.143	0.91–1.01
	DEP	physical function	9.9 *	0.010	1.7–56.6
		Sex	0.91	0.911	0.19–4.4
		Age	0.98	0.520	0.93–1.03
Dizziness/Vertigo (n = 37)	ANX	physical function	1.9	0.396	0.42–8.6
		Sex	7.8	0.036	1.1–54.0
		Age	0.96	0.070	0.92–1.0
	DEP	physical function	0.90	0.875	0.24–3.4
		Sex	1.8	0.477	0.36–8.7
		Age	1.0	0.814	0.96–8.9
IMBALANCE (n = 31)	ANX	physical function	11.1	0.018	1.5–81.8
		Sex	1.2	0.878	0.12–11.6
		Age	0.93	0.124	0.84–1.02
	DEP	physical function	11.0	0.048	1.02–118.9
		Sex	2.1	0.547	0.18–25.3
		Age	0.94	0.218	0.86–1.03
BVH (n = 14)	ANX	physical function	0.64	0.791	0.02–16.8
		Sex	0.07	0.141	0.002–2.3
		Age	1.3	0.149	0.92–1.7
	DEP	Not performed - No individuals who reported normal DEP and normal physical function.

^*indicates Odds Ratio significance was recognized with a corrected alpha value ( p < 0.01). Reference group for sex were males. Logistic regression not performed in BVH subgroup for dependent variable DEP as there were no individuals who reported normal DEP and normal physical function.

DISCUSSION

As a specialized subset of physical therapy, vestibular rehabilitation has traditionally existed as a means of adapting or accommodating for conditions impacting dizziness/vertigo, balance, spatial awareness, and visual perception ⁽⁴⁵⁾. The aim of this study was to investigate the prevalence of mental health conditions, specifically anxiety and/or depression, in a population of individuals referred to vestibular physical therapy and determine whether a relationship existed between self-reported physical function and self-reported mental health. Significant associations were found between decreased physical function and increased anxiety/depression for the combined cohort as well for some individual diagnoses, specifically UVH and BPPV. Although trends toward significance were observed for individuals with other diagnoses, those relationships did not hold up under stricter acceptance criteria correcting for multiple comparisons. Our results are consistent with Decker et al., ⁽⁴⁶⁾ that anxiety related to vertigo and psychiatric comorbidity are low for individuals with BVH, supporting the view that a functioning peripheral vestibular system is the prerequisite needed.

The current results demonstrate in this cohort approximately 50% of individuals referred to vestibular rehabilitation will report abnormal ANX/DEP, suggesting a need for increased emphasis on screening for mental health conditions affecting individuals presenting for vestibular physical therapy. The prevalence of mental health disorders continues to increase ^(47,48). The detrimental effects of mental health disorders on health care outcomes highlights the importance of a comprehensive whole-body evaluation of physical health ⁽⁴⁹⁾. Physical therapists have assumed a traditional role of focusing primarily on helping people maintain and/or improve their movement, strength, and functional activity tolerance ⁽⁵⁰⁾. Within the overarching aim to improve physical function is an entrenched goal of improving a person’s quality of life, well-being, and sense of empowerment including psychological and mental health ⁽⁵¹⁾.

The coexistence of dysfunctional mental and physical health conditions has been well documented in other patient cohorts ^(52,53). According to the National Mental Health Commission Equally Well Consensus Statement ⁽⁵⁴⁾, 80% of individuals living with mental health disorders have a coexisting physical health issue. Individuals with chronic physical health impairment frequently develop mental health disorders ⁽⁵³⁾. Functional disability may actually be magnified when mental health disorders and chronic health conditions are both present ⁽⁵⁵⁾. Evidence-based practice with the goal of achieving optimal patient outcomes suggests that neither physical nor mental health should be treated in isolation ^(56,57). Physical therapist should work with or begin to develop multidisciplinary teams including mental health specialists who understand the synergistic relationship between mental and physical health.

The current results indicate that a heightened awareness around identifying mental health status should be placed at the heart of evidence-based physical therapy practice. Screening for mental health conditions should be prioritized in patient populations referred to vestibular rehabilitation. We encourage the physical therapy profession to develop strategies to actively screen for mental health conditions while fostering relationships with mental health providers, facilitating holistic individualized treatment for patients referred to vestibular physical therapy. When mental health concerns are identified, physical therapists should support those individuals with mental health needs by referring to appropriate local/regional mental health specialists to achieve optimal patient-centred outcomes. Specific strategies, such as cognitive behavioural therapy, has been employed and has yielded favourable results ^(58,59).

Certain aspects of individualized vestibular rehabilitation sessions also present with an enormous possibility for growth. Physical therapists, with their unique appreciation for the effect that mental health conditions play on short- and long-term recovery can have an enormous impact on patient outcomes. This study offers insight into the prevalence of mental health conditions in a subset of patient populations referred for vestibular physical therapy. Further studies should determine which referral and/or rehabilitation strategies successfully address abnormal mental health for individuals referred to vestibular physical therapy.

LIMITATIONS

Limitations to this work include the inability to identify independent contributions for underlying conditions with the coexistence of symptoms of dizziness and symptoms associated with mental health conditions. The time since onset of symptoms was not recorded in this sample which may have influence on anxiety or depression levels. Categorizing PROMIS scores as normal/abnormal was a first step to characterize the relationship between self-reports physical function and self-reported mental health. These analyses were not able to examine a dose-response effect related to severity of self-reported impairments in physical function of mental health. This study is not designed to allow for the ability to determine if anxiety/depression leads to reduced physical function or decreased physical function impacts levels of anxiety/depression. Also, the clinical head impulse test utilized during clinical evaluation has high specificity (0.91) but low to moderate sensitivity (0.45) and corresponds to semicircular canal paresis of approximately 42.5–60% ^{(60; 61)}. Thus, some individuals with milder canal paresis who opted out of video head impulse testing during the COVID-19 pandemic may have been misdiagnosed into either the dizziness or imbalance category.

CONCLUSION

Our data indicate, the number of individuals who present with mental health comorbidities such as anxiety or depression presenting to vestibular rehabilitation is significant. Physical therapists that work within this specialized domain of physical therapy practice should be aware of the prevalence and degree of influence physical performance has on mental health. Additional screening measures to identify these mental health factors is warranted.