Skip to main content
Tessa NewsLink
Paste a health news link, or browse
Latest checks

Source study found

Story checked

What you did last week may be the strongest clue to how much you'll exercise this week (opens in a new tab)

news-medical.net · 2026-10-18

Short answerEvidenceSource

Short answer

Mixed

Mixed.

The claims we could check match the study, but some claims were not covered by the evidence reviewed.

  • 3 supported
  • 4 not covered

Checked against the study summary. The full text wasn't available, so some details couldn't be settled either way.

Share this check

Follow the evidence trail
1
2

NewsLink checks it

Mixed

Every claim we could check holds up. Three of seven claims match the study. This overall rating is based only on the claims we could check. Four claims the study doesn't address.

  • 3 supported
  • 4 not covered
Open claim evidence
3
Then inspect each claim

Evidence layer

Claim by claim

Each claim gets a verdict. Expand it to see the evidence directly below.

7 claims in this story

Showing all 7 claimsChoose a verdict to focus the list.

Claim 1 of 7Not coveredThe study analyzed Strava exercise records from January 1, 2019, through June 30, 2021, and included 5.17 million users across 1,441 administrative regions in 65 countries.View evidenceHide evidence

As stated5.17 million users across 65 countries

Why this verdict

The abstract profile supports the core dataset size, broad timeframe, and 65-country scope: 5.17 million app users, Jan 2019–Jun 2021. However, at abstract depth it does not verify the app name as Strava, the exact start/end dates as Jan 1, 2019 through Jun 30, 2021, or the stated 1,441 administrative regions.

1

Study evidence

A dataset of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) was used to predict weekly PA adherence (≥150 min/week).

“We analysed PA data from 5.17 million individuals across 65 countries, collected via a smartphone PA tracking app between January 2019 and June 2021.”
2

Study evidence

Cities in lower- and upper-middle-income countries experienced greater reductions in weekly physical activity adherence during the COVID-19 pandemic compared with cities in high-income countries.

“Marked global disparities were observed: cities in lower- and upper-middle-income countries experienced greater reductions in PA during the pandemic compared with cities in high-income countries.”
Claim 2 of 7Not coveredPandemic-related changes in activity differed by country income group, with users in high-income countries showing an immediate increase and users in lower- and upper-middle-income countries experiencing greater disruption.View evidenceHide evidence

Why this verdict

The profile supports the general point that pandemic-related PA changes differed by income grouping, with lower- and upper-middle-income cities experiencing greater reductions than high-income cities. But the specific statement that high-income-country users showed an immediate increase is not available in the abstract-level evidence.

1

Study evidence

Cities in lower- and upper-middle-income countries experienced greater reductions in weekly physical activity adherence during the COVID-19 pandemic compared with cities in high-income countries.

“Marked global disparities were observed: cities in lower- and upper-middle-income countries experienced greater reductions in PA during the pandemic compared with cities in high-income countries.”
2

Study evidence

Female sex was associated with a higher likelihood of maintaining regular physical activity during the COVID-19 pandemic.

“We also applied surrogate models and econometric analyses to assess the associations of individual and urban-level attributes with PA resilience globally.”
Claim 3 of 7Not coveredMore economically developed cities experienced larger activity reductions during lockdowns but also greater long-term recovery, and higher sports-amenity density was positively associated with longer recorded weekly exercise duration.View evidenceHide evidence

Why this verdict

The abstract profile supports a related but less specific association between better access to outdoor sports infrastructure and maintaining PA, and it reports income-group disparities. It does not verify the more detailed claims about economically developed cities having larger lockdown reductions and greater long-term recovery, nor the specific outcome of sports-amenity density being associated with longer recorded weekly exercise duration.

1

Study evidence

Cities in lower- and upper-middle-income countries experienced greater reductions in weekly physical activity adherence during the COVID-19 pandemic compared with cities in high-income countries.

“Marked global disparities were observed: cities in lower- and upper-middle-income countries experienced greater reductions in PA during the pandemic compared with cities in high-income countries.”
2

Study evidence

Female sex was associated with a higher likelihood of maintaining regular physical activity during the COVID-19 pandemic.

“We also applied surrogate models and econometric analyses to assess the associations of individual and urban-level attributes with PA resilience globally.”
Claim 4 of 7Not coveredParks and outdoor pitches were consistently associated with more recorded exercise, while indoor facilities such as fitness centers and swimming pools had negative or nonsignificant associations.View evidenceHide evidence

Why this verdict

The profile supports only the broader finding that better access to outdoor sports infrastructure was associated with greater PA resilience. It does not provide abstract-level evidence separating parks and outdoor pitches from indoor facilities, nor does it verify negative or nonsignificant associations for fitness centers and swimming pools.

1

Study evidence

Female sex was associated with a higher likelihood of maintaining regular physical activity during the COVID-19 pandemic.

“We also applied surrogate models and econometric analyses to assess the associations of individual and urban-level attributes with PA resilience globally.”
Claim 5 of 7SupportedResearchers evaluated whether digital exercise records could predict adherence to physical activity and identified individual and urban factors associated with resilience in physical activity during the COVID-19 pandemic.View evidenceHide evidence

Why this verdict

The abstract-level profile supports that the study used longitudinal exercise-app data to predict weekly physical-activity adherence and applied surrogate/econometric analyses to identify individual- and urban-level attributes associated with PA resilience. The claim is framed associationally and does not imply causation.

1

Study evidence

A dataset of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) was used to predict weekly PA adherence (≥150 min/week).

“We analysed PA data from 5.17 million individuals across 65 countries, collected via a smartphone PA tracking app between January 2019 and June 2021.”
2

Study evidence

Female sex was associated with a higher likelihood of maintaining regular physical activity during the COVID-19 pandemic.

“We also applied surrogate models and econometric analyses to assess the associations of individual and urban-level attributes with PA resilience globally.”
Claim 6 of 7SupportedPersonal exercise history was the strongest predictor of weekly adherence, and adding activity duration from the preceding one to three weeks improved prediction accuracy by nearly 20 percentage points in the Middlesex County analysis.View evidenceHide evidence

As statednearly 20 percentage points

Why this verdict

The profile states that personal exercise history was the single strongest predictor and improved model accuracy by about 20 percentage points. It also supports that models used the preceding 1–3 weeks of exercise history and that Middlesex County, MA was the subsample where model classes were compared.

1

Study evidence

A dataset of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) was used to predict weekly PA adherence (≥150 min/week).

“We analysed PA data from 5.17 million individuals across 65 countries, collected via a smartphone PA tracking app between January 2019 and June 2021.”
Claim 7 of 7SupportedThe article says these findings could inform timely, targeted support and urban planning for public health emergencies, while noting that Strava users may not represent the wider population.View evidenceHide evidence

Why this verdict

The profile discussion supports potential use of the findings for targeted interventions and urban policy, and the limitations identify non-representativeness of commercial exercise-app users as a concern. The claim is appropriately hedged as 'could inform' rather than asserting proven policy effects.

1

Study evidence

A dataset of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) was used to predict weekly PA adherence (≥150 min/week).

“We analysed PA data from 5.17 million individuals across 65 countries, collected via a smartphone PA tracking app between January 2019 and June 2021.”
2

Study evidence

Cities in lower- and upper-middle-income countries experienced greater reductions in weekly physical activity adherence during the COVID-19 pandemic compared with cities in high-income countries.

“Marked global disparities were observed: cities in lower- and upper-middle-income countries experienced greater reductions in PA during the pandemic compared with cities in high-income countries.”
Then look for missing context

Context layer

What the story left out

Important study details the story did not include.

  • Model performance metrics were a material result: Middlesex County GNN accuracy was 89.9% pre-pandemic and 82.1% during the pandemic; global NN accuracy was 87.6% pre-pandemic and 82.8% during the pandemic.

    The story mentions the approximately 20 percentage-point contribution of prior exercise history but does not reflect the headline model accuracy results or the pre-pandemic versus pandemic performance decline.

    From retrospective secondary-data supervised prediction study

  • The association analyses are observational and do not establish causal effects of individual traits, social connectedness, or infrastructure on PA resilience.

    The story’s claims are mostly framed associationally, but its caveats as presented do not explicitly state that causal inference is not established or that confounding may affect the association analyses.

    From retrospective longitudinal observational analysis; retrospective observational association analysis

  • The model-class comparison was reported for a single US county subsample, limiting how confidently comparative model performance can be generalized across all settings.

    The story mentions the Middlesex County analysis but does not note that the head-to-head model-class comparison was limited to that county subsample.

    From retrospective secondary-data supervised prediction study

6 things the story did carry across
  • The paper’s central contribution is predictive modeling of weekly PA adherence using app-based exercise histories, including linear regression, neural-network, and graph-neural-network models.
  • The dataset comprised 5.17 million exercise-app users across 65 countries from Jan 2019 to Jun 2021, with weekly adherence defined as at least 150 minutes/week of exercise.
  • Personal exercise history was the strongest predictor, improving predictive accuracy by about 20 percentage points, while added model complexity and social-network features gave modest incremental gains.
  • The paper reports global disparities in PA reductions during the pandemic, with lower- and upper-middle-income cities experiencing greater reductions than high-income cities.
  • Individual- and urban-level attributes associated with PA resilience included female sex, social connectedness, and better access to outdoor sports infrastructure.
  • The app-user cohort may not represent the general population, creating selection-bias and external-validity concerns.
Then read the study layer

Study layer

Study at a glance

Scan the study first. Expand only the parts you want to inspect.

Pieces of work

3

Evidence read

study summary

Lead result

secondary data

1Lead resultsecondary dataDevelop and evaluate statistical and machine-learning models (linear regression, neural network, graph neural network) to predict weekly physical activity adherence from longitudinal exercise-app data before and during the COVID-19 pandemic.retrospective secondary-data supervised prediction studyExpand

In plain English

Retrospective observational analysis of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) that developed and evaluated supervised prediction models (linear regression, neural networks, graph neural networks) to classify weekly physical activity (PA) adherence (≥150 min/week). Models used 1–3 weeks of prior individual exercise history plus sociodemographic, weather, and pandemic-related covariates, with optional social-network features; performance was reported separately for pre-pandemic and during-pandemic periods and compared in a US county subsample.

Key findings

  • A dataset of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) was used to predict weekly PA adherence (≥150 min/week).
  • In a US county subsample (Middlesex County, MA), a graph neural network (GNN) achieved the highest classification accuracy: 89.9% pre-pandemic and 82.1% during the pandemic.Accuracy: 89.9% (pre-pandemic); 82.1% (during pandemic)
“We analysed PA data from 5.17 million individuals across 65 countries, collected via a smartphone PA tracking app between January 2019 and June 2021.”
What this piece can’t prove
  • Population comprises exercise-app users; abstract does not report representativeness of the sample nor potential selection biases that could affect external validity.
  • Abstract does not provide uncertainty quantification (confidence intervals) for reported accuracy metrics or statistical tests for reported associations.

2 further details could not be confirmed from the summary.

2secondary dataQuantify global changes/disparities in physical activity resilience during the pandemic across countries/cities using the exercise-app cohort.retrospective longitudinal observational analysisExpand

In plain English

Retrospective longitudinal analysis of 5.17 million exercise-app users from 65 countries (Jan 2019–Jun 2021) describing global changes in weekly physical activity (PA) adherence (>=150 min/week) before vs during the COVID-19 pandemic, with stratified comparisons across cities/countries by income group. Main finding: marked disparities in PA reductions during the pandemic, with greater declines in cities in lower- and upper-middle-income countries compared with high-income countries; certain individual- and city-level attributes (female sex, social connectedness, access to outdoor sports infrastructure) were associated with greater PA resilience.

Key findings

  • Cities in lower- and upper-middle-income countries experienced greater reductions in weekly physical activity adherence during the COVID-19 pandemic compared with cities in high-income countries.
  • Females, socially connected individuals, and residents in cities with better access to outdoor sports infrastructure were more likely to maintain regular physical activity during the pandemic.
“Marked global disparities were observed: cities in lower- and upper-middle-income countries experienced greater reductions in PA during the pandemic compared with cities in high-income countries.”
What this piece can’t prove
  • Analysis is observational and descriptive (pre-pandemic vs during-pandemic comparisons) using an app-based cohort; causal inference is not established by these comparisons.
  • Abstract does not report quantitative effect sizes for the reported cross-country/city disparities.
3secondary dataIdentify individual-level and urban-level factors associated with physical activity resilience during the pandemic using surrogate-model interpretation and econometric association analyses.retrospective observational association analysisExpand

In plain English

Using surrogate-model interpretation and econometric association analyses on a global smartphone exercise app dataset (5.17 million users across 65 countries, Jan 2019–Jun 2021), the study reports associations between individual- and city-level attributes and maintenance of regular physical activity (weekly adherence, ≥150 min) during the COVID-19 pandemic. Reported associations: female sex, greater social connectedness, and better access to outdoor sports infrastructure were each associated with a higher likelihood of maintaining regular PA during the pandemic; cities in lower- and upper-middle-income countries experienced larger reductions in PA than cities in high-income countries.

Key findings

  • Female sex was associated with a higher likelihood of maintaining regular physical activity during the COVID-19 pandemic.
  • Greater social connectedness was associated with a higher likelihood of maintaining regular physical activity during the pandemic.
“We also applied surrogate models and econometric analyses to assess the associations of individual and urban-level attributes with PA resilience globally.”
What this piece can’t prove
  • Abstract does not report econometric model specifications, covariates included, or adjustment strategies—limiting assessment of confounding control.
  • Surrogate-model interpretation can indicate feature importance but does not by itself establish causal relationships.

4 further details could not be confirmed from the summary.

Finally, the search trail

Method layer

NewsLink found the paper. Tessa takes you deeper.

NewsLink checks the story. Tessa is where you inspect the paper, authors, evidence, and research context.

Papers considered

The selected paper, plus nearby candidates.

PubMed, Europe PMC, Crossref · 39 candidate papers

And 33 more candidates considered.