Causal Attributions for Exercise Among Sedentary Older Latinos Christina Metzler and Esther Jun Mentor: Catherine Sarkisian, MD MSPH
Background Causal attribution theory holds that individuals’ explanations for given events have underlying dimensional structures The Attribution dimensions include –Locus of Causality –Personal Control –External Control –Stability Measuring these four causal attributions increases understanding of individuals’ explanations for events
Background Causal attribution research has been instrumental for behavioral change in education, sports, and health maintenance This theory’s usefulness for increasing exercise level is still unknown
Specific Aims Describe the qualitative reasons older Latinos give for exercising or not exercising. Examine the causal attributions associated with these reasons. Identify characteristics associated with causal attributions for/not exercising, including: –physical activity level –medical comorbidities – expectations regarding aging
Methods Cross sectional analysis of baseline data from RCT (¡Caminemos!) of a behavioral intervention to increase walking Older Latinos recruited from 27 senior centers in the greater Los Angeles region Eligibility: age 60 years or older, not participating in 20 minutes of exercise 3x per week or more
The ¡Caminemos! Study (n=571) Age, mean 73 years Female 77% Married29% 8 th Grade education or less 85% Income < $10,000/year69% BMI > 25 kg/m 2 84%
Do you Exercise? (Defined by doing a physical activity lasting at least 10 min) Measuring Causal Attributions N=571 No=129Yes=442 What is the Main Reason You Do/Don’t Exercise? Causal Attributions (CDSII) Locus of Causality Internal Control External Control Stability Qualitative Analysis Type of Reason
Reasons For/Not Exercising Not ExercisingExercising
Mean Causal Attributions for/not Exercising More Internal Cause More Control More Stable * * * * P <.05
Mean Causal Attributions By Reason for Exercising * * * * P <.05
Mean Causal Attributions By Reason for Not Exercising * * * * * P <.05
Attributions for Exercising (n=442) Variables (High/Low) Locus of Causality Personal Control External Control Stability Mean p- value Mean p- value Mean p-value Mean p-value Physical Activity (hr/week) H < L Physical Activity (kcal/week) H 7.80< < L Physical Performance Score H L Expectations Regarding Aging H L Medical Comorbidities H L
Attributions for Not Exercising (n=129) Variables (High/Low) Locus of Causality Personal Control External ControlStability Mean p- value Mean p-value Mean p- value Mean p- value Physical Activity (hours/week) H L Energy Expended (kcal/week) H L Physical Performance Score H < L Expectations Regarding Aging H L Medical Comorbidities H L
Conclusions Among older Latinos, causal attributions for exercising are more internal, personally controllable, and stable than causal attributions for not exercising. Older Latinos who cite a medical reason for exercising have the highest personal control along the causal attribution scale, while those who are not exercising have the lowest personal control
Conclusions Causal attributions for exercise with greater internal locus of causality, greater personal control, and greater stability are associated with –Higher levels of physical activity –Better physical performance –Higher expectations regarding aging –Lower medical comorbidities
Significance Improving individuals’ feelings of control over health and medical conditions could increase their likelihood of adopting a healthy lifestyle. Furthermore, increasing physical activity levels or expectations regarding aging might increase individuals’ perceived stability, internal causality, and personal control for reasons for exercising
Next Steps Multivariate Model To look at longitudinal data to see if increasing exercise is associated with changes in the causal attributions
Limitations Sample is not reflective of all sedentary older Latinos Relatively small sample size Subjective biases while assigning categories to the free-answer exercise question
Acknowledgements Catherine Sarkisian, MD Jason Wang (The best programmer!) Caminemos Investigators National Institute on Aging Grant MSTAR Program Williams College