Health-Related Quality of Life in Older Adults at Risk for Disability

Abstract

Background

The number of older adults living in the U.S. continues to increase, and recent research has begun to target interventions to older adults who have mobility limitations and are at risk for disability. The objective of this study is to describe and examine correlates of health-related quality of life in this population subgroup using baseline data from a larger intervention study.

Methods

The Lifestyle Interventions and Independence for Elders–Pilot study (LIFE–P) was a randomized, controlled trial that compared a physical activity intervention to a non-exercise educational intervention with 424 older adults at risk for disability. Baseline data (collected April–December 2004; analyzed in 2006) included demographics, medical history, the Quality of Well-Being Scale (QWB-SA), a timed 400 m walk, and the Short Physical Performance Battery (SPPB). Descriptive HRQOL data are presented. Hierarchical linear regression models were used to examine correlates of HRQOL.

Results

The mean QWB-SA score for the sample was 0.630 on an interval scale ranging from 0.0 (death) to 1.0 (asymptomatic, optimal functioning). The mean of 0.630 is 0.070 lower than a comparison group of healthy older adults. The variables associated with lower HRQOL included white ethnicity, more comorbid conditions, slower 400-m walk times, and lower SPPB balance and chair stand scores.

Conclusions

Older adults who are at risk for disability had reduced HRQOL. Surprisingly, however, mobility was a stronger correlate of HRQOL than an index of comorbidity, suggesting that interventions addressing mobility limitations may provide significant health benefits to this population.

Impaired mobility, defined as the ability to walk safely and independently,¹ has been shown to predict subsequent broader disability involving independent daily living activities.^{2, 3} Using these findings, researchers identified a subgroup of older adults that are at risk for developing disability.^4-7 These older adults are characterized by a sedentary lifestyle and impaired mobility. They walk more slowly and have reduced strength and balance. They are considered “at risk for disability” because they have reduced mobility but can still perform daily living activities.

Mobility and daily living are important elements of the broader concept of health-related quality of life (HRQOL) ^{8, 9} and most measures of generic HRQOL include questions about mobility.^10-13 The HRQOL of older adults is usually described in association with specific diseases, demographic characteristics, and/or healthy epidemiological samples,^14-16 but few, if any, studies describe the HRQOL of older adults who share functional limitations. The objective of this study is to describe and examine correlates of HRQOL in older adults considered at risk for disability.

METHODS

This manuscript uses baseline questionnaire data (collected April–December 2004; analyzed in 2006) from all Lifestyle Interventions and Independence for Elders pilot (LIFE-P) study participants. The study has been described in detail elsewhere.^{17, 18}

Clinical Trial

The LIFE-P study is a multisite, randomized controlled trial (RCT) in which older adults (aged 70–89) at risk for disability were assigned to either a physical activity or a successful aging intervention, both lasting 12 months. The physical activity intervention consisted of a structured exercise program focused on walking supplemented with behavioral counseling.¹⁹ The successful aging intervention consisted of educational meetings not expected to impact the main study outcomes. The goal of the LIFE-P study was to obtain key design benchmarks in preparation for a larger study of the efficacy of physical activity for preventing disability in this population.

Participants

Participants were 424 older adults considered at risk for disability, which is defined as having a Short Physical Performance Battery (SPPB) score of <10.^{4, 6} Other inclusion criteria were ages 70-89 years, sedentary lifestyle (not actively participating in a formal exercise program within the past three months), and ability to complete a 400 m walk within 15 minutes. Exclusion criteria included history of significant or recent co-morbidity. Comprehensive inclusion and exclusion criteria are given elsewhere.²⁰ Participants were recruited from four communities in geographically diverse areas of the U.S. (Pittsburgh, Winston-Salem, Dallas, and Palo Alto) using a variety of recruitment strategies.²⁰

Measures

Demographic

Participants completed baseline demographic questionnaires.

Comorbidity Index

The index of comorbidity is the sum of yes (1) or no (0) self-report responses for 10 prevalent comorbidities: hypertension, heart attack, heart failure, stroke, cancer, diabetes, broken hip, arthritis, liver disease, and lung disease. To verify reported comorbidities, participants provided evidence of prescribed medications or the exact name of medications. Only 5.4% (39/725) of “yes” responses could not be verified. These responses were coded as “possible” comorbidity and given a value of 0.5. A maximum likelihood (SAS Proc MIXED) approach was used to estimate the comorbidity index from observed responses for cases (18/424 = 1.9%) with missing data.

Mobility/Physical Functioning

Each person completed a timed 400 m self-paced walk without assistance or assistive devices.^{21, 22} Physical functioning was measured using the SPPB⁵ which assesses 3 areas of performance: balance, chair stands, and a 4-m self-paced walk. Trained observers assign a categorical score to each area of function ranging from 0 (inability to complete the test) to 4 (highest performance level). A summary score ranging from 0-12 is calculated by summing the 3 subscale scores.

Hand Grip Strength

Grip strength was measured using an adjustable, hydraulic dynamometer (Jamar Hand Dynamometer, Fred Sammons, Inc.). The best performance of two trials was selected for each side, and the average of the left and right hand were used for analysis. Predictive validity has been shown for both disability²³ and mortality.²⁴

Health-Related Quality of Life

Health-related quality of life was assessed using the Quality of Well-being Scale- Self-administered (QWB-SA).^{13, 25} The QWB-SA is a generic measure of HRQOL that combines preference-weighted values for symptoms and functioning.²⁶ Scores range from 0 (death) to 1.0 (asymptomatic, optimum functioning).²⁷ The measure has been used in multisite NIH clinical trials ^28–30 and with various medical conditions.^31–36

Statistical Analysis

Descriptive statistics reported include means with standard deviations and proportions where appropriate. Linear regression analysis was used to examine correlates of HRQOL. Independent variables were entered into the models in three blocks, with QWB-SA scores as the dependent variable. Initially, age, education, gender, ethnicity, and marital status were entered and retained if p <0.15. Education (no college vs college or more), ethnicity (white vs. non-white) and marital status (married vs. non-married) were converted to binary coding. Next, the comorbidity index was tested and retained (p <0.15). Finally, the 400-m walk time, three SPPB subscales, and average grip strength were entered and variables were retained if p <0.05.

RESULTS

Mean baseline scores are presented in Table 1. There were no missing data for the QWB-SA and other health variables. Table 2 presents QWB-SA scores for a variety of samples from published studies. However, the samples differ on factors often related to HRQOL (age, gender).

Table 1.

Participant demographics and descriptive statistics

Variable	Mean [SD] or %
Age (n=424)	76.77 [4.24]
QWB-SA score (n=424)	0.634 [0.099]
Total SPPB score (n=424)	7.52 [1.42]
SPPB Balance Test (n=424)	2.97 [1.07]
SPPB Chair Stand (n=424)	1.36 [0.83]
SPPB Gait Speed (n=424)	3.19 [0.74]
400 m walk time 9 (n=424)	8.17 [1.89]
Comorbidity index (# of conditions) (n =424)	1.71 [1.14]
Grip strength (n=399)	25.2 [8.8]
Gender (n=424)
Female	68.9%
Education (n=423)
No college	30.0%
College	45.8%
post college	21.2%
Other	3.0%
Ethnicity (n = 423)
Caucasian	74.3%
Black	18.2%
Hispanic	4.7%
Other	2.8%
Income (n =354)
Under $25,000	34.2%
$25,000 – 49,999	26.9%
$50,001 and above	22.4%
Decline to State	16.5%
Marital status (n =423)
Married	39.4%
Widowed	40.8%
Divorced	14.9%
Never married	3.8%
Other	1.1%

Table 2.

A comparison of current study results with mean QWB-SA scores and descriptors for other disease samples.

Sample characteristics	n	Age	% women	QWB-SA mean (SD)
Healthy older adults{Andresen37#502}	301	74.7	59	0.704 (.099)
Adults at risk of developing diabetes{Herman38 #503}	3234	51.2	68	0.681 (.108)
Family medicine outpatients {Frosch31 #284}	562	46.7	57	0.651 (.134)
Older Adults w/ mobility limitations (current study)	424	76.8	69	0.634 (.099)
Migraineurs {Sieber25 #265} (days without headaches)	89	42.2	87	0.628 (.149)
Cancer patients in Germany{Frosch39 #490} (prostate, BPH, colon, rectal)	275	66.3	0	0.619 (.150)
Cataract patients{Rosen36 #497}(directly prior to surgery)	233	72.5	40	0.595 (.134)
Type 1 Diabetes{Coffey34 #495}	784	34.5	55	0.572 (n/a)
Emphysema patients{Kaplan40 #504} (before pulmonary rehabilitation)	1218	67.0	39	0.571 (.114)
Type 2 Diabetes{Coffey34 #495}	1257	57.6	49	0.547 (n/a)
Rheumatology patients {Frosch31 #284}	334	55.1	84	0.516 (.130)
Migraineurs{Sieber25 #265} (days with headache)	89	42.2	87	0.492 (.157)
Major depressive disorder – outpatients {Pyne33 #241}	19	43.6	37	0.479(.112)
Major depressive disorder – inpatients {Pyne33 #241}	39	46.7	15	0.383(.118)

Regression analyses examining correlates of HRQOL are presented in Table 3. Ethnicity was the only demographic variable retained. The co-morbidity index and ethnicity variable tested in the second block were both retained. Of the functional variables entered in the third and final block, the 400 m walk time, SPPB balance subscale, and SPPB chair stand subscale were retained. The correlation between the 400 m walk and the gait speed subscale was substantial (r = −0.55, p <0.0001).

Table 3.

Hierarchical regression analysis results – Variables associated with QWB-SA scores at baseline assessment

Parameter	Estimate	Standard error	f value	p value	Tot R2	R2 change
Step 1-Demographics					0.0109	0.0109
Intercept	0.642	0.044	499.52	<.0001
White	−.028	0.010	6.30	0.0124
Step 2 – Comorbidity					0.0413	0.0304
Comorbidity index	−.011	0.004	8.35	0.0041
Step 3 - Mobility					0.1280	0.0867
400m walk time	−.008	0.003	17.64	<0.0001
SPPB balance	0.010	0.004	5.42	0.0205
SPPB chair stands	0.020	0.006	12.82	0.0004

DISCUSSION

The mean QWB-SA score for a sample of older adults considered at risk for disability was below a score found for healthy older adults³⁷ Although these samples differ slightly, this difference (0.704 − 0.634 = 0.070) is substantial and well beyond the minimally clinically important difference (MCID) of 0.030 estimated for the QWB-SA.^41,42

The decrement of 0.070 is more than the amount attributed to a variety of diseases including colitis, migraine, arthritis, stroke, ulcer, asthma, and anxiety. ¹⁴ Thus declining mobility may have a greater negative impact on HRQOL than many distinct disease states. In this sample, mobility function was related to HRQOL independent of a comorbidity index. This finding highlights the level of impairment in this subpopulation and underscores the need to develop effective interventions for older adults at risk for disability regardless of the diseases they may or may not have.^{2, 6} It is also important to note that the three mobility variables (400-m walk time, the balance SPPB subscale, and the chair stands SPPB subscale) accounted for unique aspects of HRQOL. This finding provides evidence that mobility is multidimensional.

The QWB-SA is only one of many generic HRQOL instruments and includes questions about mobility so a correlation is not surprising. Although the QWB includes mobility items, the QWB-SA assesses 59 symptoms which usually have a larger impact on scores than mobility or other function-related questions.

In contrast to other published studies,^43–45 White participants had lower QWB-SA scores than non-Whites (0.627 vs. 0.652). However, African Americans and other ethnic groups have reported higher satisfaction with physical function than white participants elsewhere.⁴⁶ Although interesting, the difference of 0.025 is below the minimally clinically important difference (MCID) of the QWB-SA^41,42 and differences in QWB-SA scores by race/ethnicity have not been found elsewhere. Unexpectedly, gender and age were not significantly related to QWB-SA scores. Typically, HRQOL scores are lower for women and decrease with older age.^{14,38,47–49} However, the study sample had a restricted range of ages and mobility level

Our results are cross-sectional and subsequently limit causal inference. Also, the inclusion and exclusion criteria used in the LIFE-P trial limit generalizability. Therefore, study results should be interpreted appropriately. Replicating the findings with other measures of HRQOL and mobility is important because the measures used differ from other measures of the same constructs.

In summary, QWB-SA scores for older adults at risk for disability were below those of a sample of healthy older adults, providing evidence that the HRQOL of this segment of older adults may benefit from intervention. Although much of the variance in HRQOL was unexplained, mobility variables were stronger correlates than comorbidity. Taken together with past research, which has demonstrated that loss of mobility predicts loss of independence, mortality, and nursing home admission,⁷ it is clear that interventions that can preserve or improve mobility in older adults could produce increases in both quantity and quality of life.