By Sunmoo Yoon Ph.D., RN

Only 2 out of 10 older adults meet the national guidelines for physical activity in the United States. Little is known about the interrelationships of many socio-ecological factors to improve physical activity behavior among Hispanic older adults. As we move towards a precision medicine approach, we need innovative strategies to discover precisely tailored targets and accurate interventions. Data mining has the potential to offer such insights.

In a recent study published in the Journal of Gerontology Nursing, my colleagues and I investigated the dynamics of the various factors affecting physical activity and precise targets (e.g., precise age groups) among urban Hispanics older adults applying data mining techniques for the future intervention.

Precise targets and intervention to improve physical activity among older Hispanics using data mining techniques

Key Findings

• Physical Activity Level (n = 2,343). Few participants reported moderate (5.0 percent) or vigorous physical activity (3.7 percent) on a weekly basis among urban Hispanic order adults.
• Active Transport (26.3 percent walking to work, school and doing errands ≥30 minutes per week). Immigrants were less likely to engage in active transport than non-immigrants and there were no modifiable factors identified for immigrants. Among non-immigrants who had a place for exercise, those with Patient Reported Outcomes Measurement Information System anxiety scores greater than 65.3 (i.e., higher anxiety levels) were more likely to engage in active transport.
• Screen Time (37.5 percent sitting and watching TV or videos >3 hours daily). Among those unwilling to make time for physical activity, participants less than 77 years of age as compared to those older were more likely to have shorter TV or video time. Among those with willing to make time and having a place for physical activity, participants less than 71 years of age and old with family support to look up information on the Internet were more likely to have shorter TV or video time.

Study investigates factors affecting physical activity and precise targets among urban Hispanics older adults

Implications for Physical Activity Intervention Design

Physical activity goals. Although the study sample was fairly healthy (93.4 percent excellent to fair general heath), the levels of physical activity were low, and more than one third of the participants reported sitting and watching TV three or more hours daily. This finding supports that there is need for a shift in physical activity goals to reduce sedentary time and promote light activities which may be more realistic than intense exercise among older adults.

Autonomy. To reduce screen time, willingness to make time was detected as the strongest predictors as a novel finding for Hispanics. This finding suggests that future interventions focus on manipulating a person’s autonomy which promote willingness to make time.

Having a place for exercise. To improve active transport or to reduce screen time, strategies to figure out accessible a place for exercise are continued to be applied.

Hispanic older immigrants. To promote active transport, culturally sensitive interventions to promote physical activity are needed among immigrants which were detected as the strongest predictors.

Age. The non-modifiable factor of age was a strong predictor for screen time, regardless of whether or not the participant had a place to exercise. This suggests that interventions aimed at decreasing screen time should consider age and that special interventions may be needed for adults above 71 years of age.

Methodological Implications

Data mining techniques. Data mining techniques led by domain experts allowed us to investigate and to represent relations among individual, interpersonal, environmental, and structural level factors simultaneously and efficiently among a large number of variables. Moreover, the findings compare favorably with the findings of physical activity studies that used logistic regression while having the advantage of more clearly defining subgroups for targeted intervention.

N-of-1 design.  The role of psychological factors, such as anxiety, on physical activity behavior has been understudied. In our study, anxiety was depicted as a strong factor for active transport among non-immigrant Hispanic older adults who have places for exercise (n=57). Future studies utilizing N-of-1 design to elucidate physical activity among anxious Hispanic older adults physical activity are encouraged.

Health Policy Implications

Vulnerable population. Culturally sensitive interventions designed to address modifiable factors at the individual and environmental levels—Hispanic immigrants in their 70s and older saying at home—are urgently needed.

 Funding Acknowledgement: This study was supported by Agency for Healthcare Research and Quality grants WICER (R01HS019853, PI: Bakken) and WICER4U (R01HS022961, PI: Bakken).

Read the Article

Yoon, S., Suero-Tejeda, N., & Bakken, S. (2015). A data mining approach for examining predictors of physical activity among urban older adults. Journal of gerontological nursing, 41(7), 14-20.