PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
May 9, 2026Journal of Clinical and Translational Science0 citationsOpen Access

285 Assessing associations between social connectedness and physical activity among US adults using unsupervised machine learning: An analysis of the All of Us Research Program

DVDavid López VenerosIEIpek EnsariYSYashika Sharma

Key Points

  • This analysis aims to explore how different levels of social connectedness affect physical activity among US adults using data from the All of Us Research Program.
  • Conducted secondary analysis of All of Us data from 8,406 participants who provided objective physical activity data via Fitbit.

Structured PICO

Does social connectedness improve objectively measured physical activity and daily steps in US adults?

P
Population
8,406 US adults from the All of Us Research Program who completed the Social Determinants of Health Survey and provided 7 days of objective physical activity data (Fitbit). Mean age 55.6±14.7 years; 70.0% women; 80.1% non-Hispanic White.
I
Intervention
Low social support and/or low neighborhood cohesion clusters (HSS-LSC and LSS-LSC)
C
Comparator
High social support/high neighborhood cohesion (HSS-HSC) cluster
O
Outcome
Meeting moderate-to-vigorous physical activity (MVPA) guidelines and achieving 10,000 steps/day

Lower social connectedness, particularly low neighborhood cohesion, is associated with significantly reduced odds of meeting physical activity guidelines and achieving 10,000 steps per day among US adults.

Abstract

Objectives/Goals: Using k-means clustering, we analyzed data from the All of Us Research Program to 1) identify distinct social connectedness clusters (based on social support and neighborhood social cohesion) and 2) examine their associations with objectively measured physical activity (PA) and daily steps among adults in the USA. Methods/Study Population: PA is an important modifiable factor for preventing noncommunicable diseases, yet <50% of US adults meet moderate-to-vigorous PA (MVPA) guidelines. While social connectedness may influence PA, its relationship with objectively measured PA remains understudied. We conducted a secondary analysis of All of Us data from participants who completed the Social Determinants of Health Survey and provided 7 days of objective PA data (Fitbit). Using k-means clustering, we identified social connectedness profiles based on social support and neighborhood cohesion. We determined the optimal number of clusters using partition criteria and ran logistic regression models (adjusted for demographic, clinical, and neighborhood factors) to examine associations with meeting MVPA guidelines and achieving 10,000 steps/day. Results/Anticipated Results: Among 8,406 participants (mean age 55.6±14.7 years; 70.0% women; 80.1% non-Hispanic White), we identified three clusters: high social support/high neighborhood cohesion (HSS-HSC; 28%), high social support/low neighborhood cohesion (HSS-LSC; 47%), and low social support/low neighborhood cohesion (LSS-LSC; 24%). Compared to HSS-HSC, participants in HSS-LSC and LSS-LSC clusters had lower odds of meeting MVPA guidelines (HSS-LSC: AOR=0.87, 95% CI=0.78–0.98; LSS-LSC: AOR=0.81, 95% CI=0.70–0.93) and achieving 10,000 steps/day (HSS-LSC: AOR=0.82, 95% CI=0.72–0.92; LSS-LSC: AOR=0.84, 95% CI=0.72–0.98). Discussion/Significance of Impact: Social connectedness is associated with greater PA. Nearly 50% of our sample had low neighborhood cohesion despite high social support, highlighting gaps in community-level connectedness. Findings support the need for multilevel interventions to address determinants of PA among adults.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Veneros et al. (2026) studied this question.

synapsesocial.com/papers/69fed03cb9154b0b8287737bhttps://doi.org/10.1017/cts.2026.10479
Ask AI
Helpful
Bookmark
Share
View Full Paper