JSJiao Song

Jiao Song

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Jiao Song has a total of 80 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Towards a Social Media-based Disease Surveillance System for Early Detection of Influenza-like Illnesses: A Twitter Case Study in Wales

    by Mark Drakesmith, Dimosthenis Antypas, Clare Brown, Jose Camacho-Collados, Jiao Song

    Mark Drakesmith, Dimosthenis Antypas, Clare Brown, Jose Camacho-Collados, Jiao Song. Proceedings of the Tenth Workshop on Noisy and User-generated Text. 2025.

  • Open Access

    Towards a Social Media-based Disease Surveillance System for Early Detection of Influenza-like Illnesses: A Twitter Case Study inWales

    by Mark Drakesmith, Dimosthenis Antypas, Claire Brown, Jose Camacho-Collados, Jiao Song

    Social media offers the potential to provide detection of outbreaks or public health incidents faster than traditional reporting mechanisms. In this paper, we developed and tested a pipeline to produce alerts of influenza-like illness (ILI) using Twitter data. Data was collected from the Twitter API, querying keywords referring to ILI symptoms and geolocated to Wales. Tweets that described first-hand descriptions of symptoms (as opposed to non-personal descriptions) were classified using transformer-based language models specialised on social media (BERTweet and TimeLMs), which were trained on a manually labelled dataset matching the above criteria. After gathering this data, weekly tweet counts were applied to the regression-based Noufaily algorithm to identify exceedances throughout 2022. The algorithm was also applied to counts of ILI-related GP consultations for comparison. Exceedance detection applied to the classified tweet counts produced alerts starting four weeks earlier than by using GP consultation data. These results demonstrate the potential to facilitate advanced preparedness for unexpected increases in healthcare burdens.

  • Open Access

    Health of unpaid carers in Wales, UK: a population data linkage study

    by Fangzhou Huang, Jiao Song, Alisha R Davies

    Background: The population of unpaid carers in Wales increased to record. There is no systematic approach to record unpaid caring status, resulting in limited quantitative evidence on unpaid carers’ health. The aim of this study is to: (i) create an e-cohort of unpaid carers by linking routinely collected health and administrative datasets in Wales, UK. (ii) investigate whether long-term health conditions and multimorbidity are more prevalent amongst unpaid carers than non-carers. Methods: Unpaid carers were identified by linking primary care dataset, National Survey for Wales data with demographic characteristics in the Secure Anonymise Information Linkage Databank. The clinical codes identified in Cambridge Multimorbidity Score were used to explore the prevalence of long-term health conditions. Results: A total of 91 220 unpaid carers in Wales were identified between 1 January 2010 and 1 March 2022. Unpaid carers were found at higher risk of managing 35 of 37 long-term health conditions and multimorbidity than non-carers, exacerbated amongst younger age groups and deprived communities. Conclusions: The creation of the first e-cohort of unpaid carers in Wales provides opportunities to perform rapid analysis to systematically understand health needs and evaluate initiatives in future. To better support unpaid carers, flexible approaches focusing on early identification and prevention is crucial.

  • Open Access

    Green–blue space exposure changes and impact on individual-level well-being and mental health: a population-wide dynamic longitudinal panel study with linked survey data

    by Rebecca S Geary, Daniel A Thompson, Joanne K Garrett, Amy Mizen, Francis M Rowney, Jiao Song, Mathew P White, Rebecca Lovell, Alan Watkins, Ronan A Lyons, et al.

    Background: Cross-sectional evidence suggests that living near green and blue spaces benefits mental health; longitudinal evidence is limited. Objectives: To quantify the impact of changes in green and blue spaces on common mental health disorders, well-being and health service use. Design: A retrospective, dynamic longitudinal panel study. Setting: Wales, UK. Participants: An e-cohort comprising 99,682,902 observations of 2,801,483 adults (≥ 16 years) registered with a general practice in Wales (2008–2019). A 5312-strong ‘National Survey for Wales (NSW) subgroup’ was surveyed on well-being and visits to green and blue spaces. Main outcome measures: Common mental health disorders, general practice records; subjective well-being, Warwick–Edinburgh Mental Well-being Scale. Data sources: Common mental health disorder and use of general practice services were extracted quarterly from the Welsh Longitudinal General Practice Dataset. Annual ambient greenness exposure, enhanced vegetation index and access to green and blue spaces (2018) from planning and satellite data. Data were linked within the Secure Anonymised Information Linkage Databank. Methods: Multilevel regression models examined associations between exposure to green and blue spaces and common mental health disorders and use of general practice. For the National Survey for Wales subgroup, generalised linear models examined associations between exposure to green and blue spaces and subjective well-being and common mental health disorders. Results and conclusions: Our longitudinal analyses found no evidence that changes in green and blue spaces through time impacted on common mental health disorders. However, time-aggregated exposure to green and blue spaces contrasting differences between people were associated with subsequent common mental health disorders. Similarly, our cross-sectional findings add to growing evidence that residential green and blue spaces and visits are associated with well-being benefits: Greater ambient greenness (+ 1 enhanced vegetation index) was associated with lower likelihood of subsequently seeking care for a common mental health disorder [adjusted odds ratio (AOR) 0.80, 95% confidence interval, (CI) 0.80 to 0.81] and with well-being with a U-shaped relationship [Warwick–Edinburgh Mental Well-being Scale; enhanced vegetation index beta (adjusted) –10.15, 95% CI –17.13 to −3.17; EVI2 beta (quadratic term; adj.) 12.49, 95% CI 3.02 to 21.97]. Those who used green and blue spaces for leisure reported better well-being, with diminishing extra benefit with increasing time (Warwick–Edinburgh Mental Well-being Scale: time outdoors (hours) beta 0.88, 95% CI 0.53 to 1.24, time outdoors2 beta −0.06, 95% CI −0.11 to −0.01) and had 4% lower odds of seeking help for common mental health disorders (AOR 0.96, 95% CI 0.93 to 0.99). Those in urban areas benefited most from greater access to green and blue spaces (AOR 0.89, 95% CI 0.89 to 0.89). Those in material deprivation benefited most from leisure time outdoors (until approximately four hours per week; Warwick–Edinburgh Mental Well-being Scale: time outdoors × in material deprivation: 1.41, 95% CI 0.39 to 2.43; time outdoors2 × in material deprivation −0.18, 95% CI −0.33 to −0.04) although well-being remained generally lower. Limitations: Longitudinal analyses were restricted by high baseline levels and limited temporal variation in ambient greenness in Wales. Changes in access to green and blue spaces could not be captured annually due to technical issues with national-level planning datasets. Future work: Further analyses could investigate mental health impacts in population subgroups potentially most sensitive to local changes in access to specific types of green and blue spaces. Deriving green and blue spaces changes from planning data is needed to overcome temporal uncertainties. Funding: This project was funded by the National Institute for Health and Care Research (NIHR) Public Health Research programme (Project number 16/07/07) and will be published in full in Public Health Research; Vol. 11, No. 10. Sarah Rodgers is part-funded by the NIHR Applied Research Collaboration North West Coast.


  • Open Access

    Exploring the Health Impacts and Inequalities of the New Way of Working

    by Melda Lois Msc Griffiths, Benjamin J. Gray, Richard G. Kyle, Jiao Song, Alisha R. Davies

    Objective: The aim of the study is to provide insights into the working Welsh adult population’s perceptions of the health impacts of working from home (WFH), their ability to WFH, and their WFH preferences. Methods: Data were collected from 615 working adults in Wales between November 2020 and January 2021 in a household survey. Results: More than 45% of those able to WFH reported worsened mental well-being and loneliness. Working from home worsened the diets, physical activity, smoking, and alcohol use of those in poorer health. Approximately 50% were able to WFH, although individuals living in more deprived areas, in atypical employment or with precarious income, were less able to WFH. Nearly 60% wanted to WFH to some capacity. Conclusions: The new way of working introduces new challenges to preserving workforce mental well-being, regulating health behaviors, and tackling inequalities. Hybrid models and targeted health support could make WFH healthier and more equitable.


  • Open Access

    Ang-(1–7) protects skeletal muscle function in aged mice

    by Ying Li, Jiao Song, Yangyang Jiang, Xue Yang, Li Cao, Chun Xiao, Suli Li, Birong Dong, Xiaoli Huang

    Background: The angiotensin-converting enzyme 2 (ACE2)/angiotensin 1–7 (Ang-(1–7)) axis has been shown to protect against the age-associated decline in skeletal muscle function. Here, we investigated the protective effects of ACE2 in mitigating the age-associated decline of skeletal muscle function and to identify the potential underlying molecular mechanisms.Methods: We measured the expression levels of Ang-(1–7) in C57BL/6J mice of different ages and correlated these levels with measures of skeletal muscle function. We also investigated the expression of myocyte enhancer factor 2 A (MEF2A) in ACE2 knockout (ACE2KO) mice and its relationship with muscle function. We then treated aged ACE2KO mice for four weeks with Ang-(1–7) and characterized the levels of MEF2A and skeletal muscle function before and after treatment. We assessed the impact of Ang-(1–7) on the growth and differentiation of C2C12 cells in vitro and assessed changes in expression of the glucose transporter type 4 (Glut4).Results: Aged mice showed reduced skeletal muscle function and levels of Ang-(1–7) expression in comparison to young and middle-aged mice. In ACE2KO mice, skeletal muscle function and MEF2A protein expression were significantly lower than in age-matched wild-type (WT) mice. After one month of Ang-(1–7) treatment, skeletal muscle function in the aged ACE2KO mice improved, while MEF2A protein expression was similar to that in the untreated group. In C2C12 cells, Ang-(1–7) was shown to promote along with the upregulated expression of Glut4.Conclusions: The ACE2/ Ang-(1–7) axis has a protective function in skeletal muscle and administration of exogenous Ang-(1–7) can delay the age-related decline in the function of skeletal muscle.


  • Open Access

    Characteristics of those most vulnerable to employment changes during the COVID-19 pandemic: a nationally representative cross-sectional study in Wales

    by Benjamin J Gray, Richard G Kyle, Jiao Song, Alisha R Davies

    Background: The public health response to the SARS-CoV-2 (COVID-19) pandemic has had a detrimental impact on employment and there are concerns the impact may be greatest among the most vulnerable. We examined the characteristics of those who experienced changes in employment status during the early months of the pandemic. Methods: Data were collected from a cross-sectional, nationally representative household survey of the working age population (18–64 years) in Wales in May/June 2020 (n=1379). We looked at changes in employment and being placed on furlough since February 2020 across demographics, contract type, job skill level, health status and household factors. χ2 or Fisher’s exact test and multinomial logistic regression models examined associations between demographics, subgroups and employment outcomes. Results: Of our respondents, 91.0% remained in the same job in May/June 2020 as they were in February 2020, 5.7% were now in a new job and 3.3% experienced unemployment. In addition, 24% of our respondents reported being placed on furlough. Non-permanent contract types, individuals who reported low mental well-being and household financial difficulties were all significant factors in experiencing unemployment. Being placed on ‘furlough’ was more likely in younger (18–29 years) and older (60–64 years) workers, those in lower skilled jobs and from households with less financial security. Conclusion: A number of vulnerable population groups were observed to experience detrimental employment outcomes during the initial stage of the COVID-19 pandemic. Targeted support is needed to mitigate against both the direct impacts on employment, and indirect impacts on financial insecurity and health.


  • Open Access

    Creating an e-cohort of individuals with lived experience of homelessness and subsequent mortality in Wales, UK

    by Jiao Song, Charlotte N B Grey, Alisha R Davies

    Background: Homelessness is an extreme form of social exclusion, with homeless people experiencing considerable social and health inequities. Estimates of morbidity and mortality amongst homeless populations is limited due to the lack of recording of housing status across health datasets. The aim of this study is to: (i) identify a homelessness e-cohort by linking routine health data in Wales, and (ii) explore whether a period of reported past homelessness, places this population at greater risk of morbidity and mortality. Methods: Homelessness identified through linkage across primary, secondary care and substance misuse datasets in the Secure Anonymised Information Linkage (SAIL) Databank. Mortality was examined through linkage to the Office for National Statistics mortality data. Results: E-cohort of 15 472 individuals with lived experience of homelessness identified. Of those, 21 individuals died between February and July 2020 involving coronavirus disease of 2019 (COVID-19). Those with lived experience of homelessness had increased mortality from many causes including accidents, liver diseases and suicides. Conclusion: Linking multiple routine datasets provides a more comprehensive dataset of a marginalized population, including individuals who are not included in government homeless statistics. Application of the cohort demonstrated that individuals with lived experience of homelessness have increased mortality involving COVID-19 and other causes. The underlying reasons, health needs and causes of death warrant further exploration.