PURPOSE: The objectives of this study were to investigate the psychometric properties of the SF-36 in a sample of older adults with chronic conditions and to test whether measurement bias exists based on the levels of comorbidity. METHODS: Participants included were 979 cognitively intact older adults with comorbidities who were interviewed at their homes. We examined the psychometric properties of the SF-36 and conducted confirmatory factor analysis (CFA) to investigate the assumption of measurement invariance by the levels of comorbidity. RESULTS: Overall data quality was high and scaling assumptions were generally met with few exceptions. Floor and ceiling effects were present for the role-physical and role-emotional subscales. Using CFA, we found that a three-factor measurement model fits the data well. We identified two violations of measurement invariance. Results showed that participants with high comorbidity level place more emphasis on social functioning (SF) and bodily pain (BP) in relation to physical health-related quality of life (HRQoL) than those with low comorbidity level. CONCLUSIONS: Measurement bias was present for the SF and BP components of the SF-36 physical HRQoL measure. Researchers should be cautious when considering the use of SF-36 in clinical studies among older adults with comorbidities.
PURPOSE: The objectives of this study were to investigate the psychometric properties of the SF-36 in a sample of older adults with chronic conditions and to test whether measurement bias exists based on the levels of comorbidity. METHODS:Participants included were 979 cognitively intact older adults with comorbidities who were interviewed at their homes. We examined the psychometric properties of the SF-36 and conducted confirmatory factor analysis (CFA) to investigate the assumption of measurement invariance by the levels of comorbidity. RESULTS: Overall data quality was high and scaling assumptions were generally met with few exceptions. Floor and ceiling effects were present for the role-physical and role-emotional subscales. Using CFA, we found that a three-factor measurement model fits the data well. We identified two violations of measurement invariance. Results showed that participants with high comorbidity level place more emphasis on social functioning (SF) and bodily pain (BP) in relation to physical health-related quality of life (HRQoL) than those with low comorbidity level. CONCLUSIONS: Measurement bias was present for the SF and BP components of the SF-36 physical HRQoL measure. Researchers should be cautious when considering the use of SF-36 in clinical studies among older adults with comorbidities.
Authors: Jordan N Kohn; Judith D Lobo; Emily A Troyer; Kathleen L Wilson; Gavrila Ang; Amanda L Walker; Christopher Pruitt; Meredith A Pung; Laura S Redwine; Suzi Hong Journal: Aging Ment Health Date: 2022-03-21 Impact factor: 3.514
Authors: Mattia Roppolo; E Saskia Kunnen; Paul L van Geert; Anna Mulasso; Emanuela Rabaglietti Journal: Clin Interv Aging Date: 2015-10-29 Impact factor: 4.458