Health education and promotion of #PhysicalActivity on Twitter before and during COVID-19

Abstract

Objective:

Twitter, a microblogging service, allows users to post short messages (tweets) and link posts through hashtags, creating online communities to enable dissemination of health education. The objective of this descriptive study was to examine Twitter’s #PhysicalActivity health education and promotion efforts, including differences in response before (March 2019–February 2020) and during (March 2020–February 2021) the COVID-19 pandemic.

Design, setting and methods:

A codebook was developed to conduct a quantitative content analysis of #PhysicalActivity tweets before and during the COVID-19 pandemic. Twitter’s Advanced Search parameters included minimum 1 reply, 25 likes and 5 retweets. Tweets were double-coded for user characteristics, community response, tweet elements (video, photo, link, questions, events, original text, survey/response, journal article, infographic) and intended audience. T-tests examined differences in researcher-generated NET response (replies + retweets + likes) between COVID-19 and comparison years.

Results:

Data were collected from 400 tweets. Eighty percent of tweets were from personal accounts, with over half indicating the possession of a health professional degree. Twenty-nine percent of tweets gave behavioural recommendations of which 56.8% provided a rationale. Nearly all used at least two components, text and photos most common. The general public was the most intended audience with health professionals second. While the NET response was greater during COVID-19 (M = 139.6, SD = 156.8) than prior to it (M = 116.8, SD = 105.1), the difference was not statistically significant (p = .091).

Conclusion:

Twitter platform provided an opportunity to disseminate health education, specifically for the promotion of physical activity, while adhering to public health recommendations to #StaySafeStayHome. Our findings provide insights into engaging this online community to inform future physical education and promotion dissemination.

Keywords

Coding instrument COVID-19 pandemic health education health promotion physical activity social media

Introduction

The significant role physical activity plays in disease prevention, health education and promotion, and overall well-being is well-established in the literature (Biswas et al., 2015; Blair, 2009) and is associated with a reduction in all-cause mortality, morbidity and associated healthcare costs (Kokkinos, 2012; Towne et al., 2018). Despite the frequency and intensity of these findings, rates of physical activity among the population continue to fall short of recommended guidelines (World Health Organization [WHO], 2022a). One in four adults aged 18 years and older do not meet the globally recommended levels of physical activity, which include 150 minutes per week (WHO, 2022a). Due to this continued challenge, and the rise in popularity of digital communication, this has prompted discussions about using social marketing theory to measure physical activity education and promotion communicated to the broader public, calls for improvement in these strategies and further research about communication and promotion of physical activity (Bergeron et al., 2019).

Social marketing theory argues that the aim of marketing should be to positively encourage the community to voluntarily accept, reject, modify or abandon a behaviour for the resulting benefit. This benefit can be for individuals, groups or society as a whole. The four constructs of marketing (product, price, place and promotion) are then translated to behaviour change, the cost of behaviour change, strategies that support and ease adopting the behaviour, and messaging strategies including platforms, promotion and reinforcement (Kotler et al., 2002). Considering the mass use of the Internet (Pew Research Center, 2014) and rising popularity of social media platforms (Brenner and Smith, 2013), digital communication is identified as an effective and cost-efficient means to disseminate physical activity information and other forms of health education (Korda and Itani, 2013; Lee et al., 2014; Neiger et al., 2012).

Researchers studying social media and health promotion have found that this form of communication aided in the availability of diverse health information to a broad audience (Moorhead et al., 2013; Raggatt et al., 2018) and a reported increase in the popularity of social media as a primary source for health information among adults (Lupton, 2018; Raggatt et al., 2018). The importance of exploring these issues was reinforced by the COVID-19 global pandemic, which signalled the need for communication of public health and safety education. There are unique challenges in participating in physical activity during mandated quarantines including how to adhere to social distancing recommendations while engaging in physical activity (WHO, 2022b). At the same time, evidence suggests that not only does physical activity reduce the risk of acquiring SARS-CoV-2 infection by increasing the immune system response (Sallis et al., 2020), it may reduce the severity of COVID-19 disease (Sallis et al., 2021).

Studies have found that during the COVID-19 pandemic, social media facilitated personal management of physical activity, diet and quality of life (QoL), and the ability of social media consumers to be critical users of a platform to support their own health and well-being (Goodyear et al., 2021). However, there has been limited application of social marketing theory to social media and physical activity education and promotion – specifically within a distinct Twitter community (i.e. #PhysicalActivity). Twitter is a microblogging service that permits users to generate short message content (i.e. tweets) and use hashtags (#) to link content, creating online communities. Applying social marketing theory to physical activity using this form of communication, including types of media and their reach, can help build an understanding of impact and inform health education and promotion strategies (Bergeron et al., 2019).

The aim of this descriptive study was to examine Twitter’s #PhysicalActivity community’s health education and promotion efforts, including differences in response before (March 2019–February 2020) and during (March 2020–February 2021) the COVID-19 pandemic. Twitter was selected due to its rise in popularity among health professionals, researchers and the public as well as the ability to search for physical activity content across users (Korda and Itani, 2013; Lee et al., 2014), free of charge. The study was descriptive in nature, but researchers hypothesised that health education and promotion efforts on Twitter would vary before and during the pandemic.

Findings from the study have implications for future health education and promotion interventions using communication-focused physical activity studies and recommendations for dissemination of physical activity education and promotion via social media. Of note, the authors (Z.S.F., K.M., E.A.R. and C.R.S.) represent a diverse range of academic and professional disciplines brought together by serving in leadership positions for the Physical Activity Section of the American Public Health Association.

Methods

Design

A quantitative descriptive study design was used to systematically describe Twitter’s #PhysicalActivity health education and health promotion efforts using a social marketing theory lens. In addition, health education efforts before and during the COVID-19 pandemic were explored and compared.

Setting

Using Twitter’s Advanced Search Tool to manually extract all tweets meeting inclusion criteria, data collection occurred online in March 2021 focusing on three of the four social marketing theory constructs used to educate and promote safe physical activity: behaviour change, strategies that support and ease adopting the behaviour, and messaging strategies including platforms, promotion and reinforcement. The fourth construct, cost of that behaviour change, can then be calculated by identifying any resulting increase in physical activity.

A flow chart describing the inclusion process for tweets is included in Figure 1. The search parameters were set to return only English-language tweets including ‘#PhysicalActivity’ within the text of the tweet. Parameters were further specified to those receiving a minimum of 1 reply, 25 likes and 5 retweets to ascertain #PhysicalActivity tweets that have garnered online reach posted 1 year before (March 2019–February 2020; n = 205 tweets) and during the first year of the COVID-19 pandemic (March 2020–February 2021; n = 195 tweets). All duplicate tweets, those consisting of only hashtags and those in non-English language were removed to account for spam and bot accounts and enhance the validity of sample (Kim et al., 2013). All original tweets, excluding responses within a thread, that met the inclusion criteria (N = 400) were reviewed for analysis. The Twitter feeds for each month were screen-recorded to ensure consistency and order of tweets for the final coding process. The community response (e.g. likes, retweets and replies) for each individual tweet was summed to generate a researcher-created metric referred to as NET-response score. This NET-response score provides an indication of the relative impact of each tweet in the #PhysicalActivity community.

Figure 1.

Flow chart describing the inclusion process.

Content analysis

A content analysis framework (Neuendorf, 2017; White and Marsh, 2006) was used to develop coding strategies for social marketing theory content and tweet element areas of interest – video, photo, link, question, event notice, text, survey/response, journal article and infographic. In addition, the User Manual for Coding Behaviour Change Techniques (BCTs) in Social Media–Based Health Interventions (Simeon et al., 2020), relevant content from the US Centers for Disease Control and Prevention’s (2021) Clear Communication Index (CDC) and prior coding adapted from Julian et al. (2020) guided the development of coding scheme. These content and element areas of interest are included in Table 1.

Table 1.

Definitions and descriptive statistics of the coded items (n = 400).

Indicator	Definition	N	% M	Minimum	Maximum
Year	Before COVID-19 (March 2019–February 2020)	205	51.30%
Year	During COVID-19 (March 2020–February 2021)	195	48.70%
Time of year of post	Winter	99	24.80%
	Spring	137	34.30%
	Summer	79	19.80%
	Autumn	85	21.30%
Retweet of post	Times the post was retweeted	13,891	34.8	5	314
Post likes	Likes the post received	34,953	87.5	23	953
Total hashtags used	Hashtags used in the post	1,264	3.4	1	47
Replies to the post	Replies the post received	2,120	5.3	1	89
Tags in the post	Tags used in the post	641	1.6	0	9
NET response	Likes + Retweets + Replies	51,110	26.6	0	1,022
COVID-19 mention	Reference to COVID-19
	No	344	86%
	Yes	56	14%
Post source	Content falls under:
	Personal Tweet	320	80.00%
	News Tweet	1	0.30%
	Organisational	60	15.00%
	Government agency	5	1.30%
	Journal account	9	2.30%
	University	5	1.30%
Source professional	Health professional
	No	160	44.80%
	Yes	197	55.20%
Negative framing	Negative	46	12.90%
Negative framing	Non-negative (or neutral)	311	87.10%
Behavioural recommendation	Behavioural recommendation
	No	282	70.70%
	Yes	117	29.30%
Behavioural recommendation reason	Behavioural reason
	No	51	43.20%
	Yes	67	56.80%
Components of the post	Video	17	4.30%
	Photo	50	12.60%
	Link	29	7.30%
	Questions	2	0.50%
	Events	10	2.50%
	Text (original)	203	51.10%
	Survey/Response	2	0.50%
	Journal article	41	10.30%
	Infographic	43	10.80%
Intended audience	General public	203	50.70%
	Health professional	187	46.80%
	Students (or prospective)	10	2.50%
Post messaging	New/Improved	15	3.80%
	Aesthetic claims	1	0.30%
	Barriers	9	2.30%
	Narrowing	1	0.30%
	Health promotion	180	45.20%
	Barriers COVID-19	33	8.30%
	Professional communication	160	40.20%
Referral to a location	No	274	68.50%
	Park/trail	16	4.00%
	Gym/studio	7	1.80%
	Home	30	7.50%
	Neighbourhood	30	7.50%
	School/work	36	9.00%
	Clinical	7	1.80%
Retweet	No	372	92.80%
Retweet	Yes	29	7.20%
Graphic used	No	45	11.30%
Graphic used	Yes	355	88.80%

The codebook of 25 items included measuring social marketing theory constructs of behaviour change, the cost of behaviour change and strategies that support and ease adopting the behaviour. Within these constructs, measurements incorporated messaging strategies including platforms, promotion, education and reinforcement from a preliminary content analysis of #PhysicalActivity tweets posted from before and during the COVID-19 pandemic. Guidance provided by MacQueen et al. (1998), which included code titles, a code definition and examples was also used. The research team followed best practice for codebook development, which included frequent meetings and communication to discuss and negotiate arising themes and coding strategies (MacQueen et al., 1998). Author Z.S.F. identified codes related to the timing and reach of the post, engagement with the post and post topic. Authors K.M., E.A.R. and C.R.S. added physical activity–specific codes covering the intention of the post, message framing, unintended consequences and disparities. Discrepancies were resolved before test coding began using the final codebook. Some items from the codebook were excluded in the full analysis due to heterogeneity of coding among the research team. Whether or not the post contained jargon terms, the perceived clarity of the tweet’s main message, and the categorisation of the post within levels of the Socio-Ecological Model were removed from the codebook due to these discrepancies, resulting in a codebook of 22 items (Table 1).

A random selection of 50 tweets from each of the two timeframes (March 2019–February 2020 and March 2020–February 2021) was used to create a preliminary sample for development of a coding instrument. For this preliminary work, the authors each conducted content analyses of the preliminary sample using the drafted coding strategies. The coding strategies were then revised, and the codebook was finalised for the analysis. Final coding was conducted manually by trained coders – the ‘gold standard’ for content analysis (Kim et al., 2013) – in research pairs. The research pairs each independently examined one of the two time-periods noted. Discrepancies in coding selections among pairing groups were resolved by consensus on a case-by-case basis.

Statistical analysis

Descriptive statistics were used to tabulate the frequency of social marketing theory content, tweet element characteristics employed and characteristics of the user. T-tests examined differences in total NET response and retweets, including likes, hashtags, replies and tags individually between COVID-19 and comparison years.

All the data included in this study are publicly available. Findings are reported in the form of aggregate and anonymous results.

Results

User characteristics (e.g. user type, health professional status) were ascertained by examining components of the users’ profiles (e.g. Twitter handle, name given and profile picture) and are presented in Table 1. There was an even representation of tweets (N = 400) posted 1 year before COVID-19 pandemic (March 2019–February 2020; n = 205 tweets) and in the first year of the COVID-19 pandemic (March 2020–February 2021; n = 195 tweets) as well as at different times of the year (Table 1).

Comparison of the most significant characteristics is included in Table 2. Post sources included tweets from personal accounts (80%), followed by organisational accounts (15%). Over half (55.2%) of the posts were from account handles indicating a professional degree (e.g. PhD, MD, Professor, RN). Tweets (29.3%) gave behavioural recommendations such as ‘keep distance while exercising in a park’, of which 56.8% provided requisite rationale, such as ‘to avoid spreading COVID-19’. Tweets (95.3%) used at least two tweet components (e.g. text, photo and links), with text (51.1%) and photos (12.6%) most common. The intended audience was determined by content of the tweet containing information that mostly all the population (i.e. the general public) could apply versus only those that research or contribute to healthcare and clinical care (health professionals). The general public was the intended audience (50.7%), followed by health professionals (46.8%). Of the tweets posted during the pandemic year, 27.7% contained text or hashtag references to COVID-19.

Table 2.

T-test results of NET response.

Logistic parameter	COVID		Compare		t(397)	p
Logistic parameter	M	SD	M	SD	t(397)	p
NET response	139.6	156.8	116.7	105.1	1.7	.87
Retweets	36.9	45.4	32.8	30.7	1.1	.26
Likes	96.5	111.9	78.9	83.2	1.8	.07
Hashtags	3.3	2.2	3.5	3.5	–0.6	.56
Replies	5.8	10.3	4.8	8.1	1.1	.27
Tags	1.7	2.1	1.5	2.1	0.8	.41

Descriptive statistics indicated the NET response was greater during COVID-19 (M = 139.6, SD = 156.8) compared to before the pandemic (M = 116.8, SD = 105.1) (Figure 2); however, a t-test revealed this difference was not statistically significant, t(334) = 1.7, p = .091. The same was true with retweets, likes, hashtags, replies and tags (Table 2).

Figure 2.

Descriptive statistics of NET response.

Discussion

The COVID-19 pandemic created additional barriers to safe physical activity. There exists an alignment between social marketing theory and Twitter with respect to the dissemination of physical activity evidence and ideas. This includes the importance of adhering to public health recommendations such as #StaySafeStayHome. Findings from this study show the alignment between the Twitter community’s response and key tenets of social marketing theory (i.e. by enhancing the sharing of strategies that support behaviour change adoption) in ways that provide insight into how future forms of physical activity promotion can be undertaken.

While most of the tweets evaluated in this study came from personal Twitter accounts, over half of these accounts were associated with health professionals, highlighting the role of Twitter in health education messaging. In addition, it is possible that some health professionals did not provide their credentials in their Twitter handle, and because of this they may be under-counted. Social media sites such as Twitter have been documented as providing a platform for providing and sharing health information on a variety of topics such as mental health (Shepherd et al., 2015), antibiotic use (Scanfeld et al., 2010), dental needs (Al-Khalifi et al., 2021), and asthma (Leroy et al., 2016). It is important, therefore, to include opportunities to teach health professional students about the use of social media for health education and health promotion (Mather, 2016).

Our findings illustrate the relevance of selected social marketing theory constructs of behaviour change, including the cost(s) of behaviour change, strategies that support and ease adopting a behaviour, messaging strategies and message reinforcement through platforms such as Twitter. These findings can inform the development of public health tools to educate, reach and engage diverse audiences in physical activity education. The need for validated measures of social media impact on health behaviours such as physical activity is clear. The codebook created for this study suggests measures that researchers and practitioners can use to assess the application of social marketing theory to physical activity and other behaviour-specific messaging for intended audience and audience reach.

Implications

The Twitter platform provided an opportunity to disseminate physical activity education in line with public health recommendations concerning COVID-19 control measures such as social distancing and staying at home when possible. Our findings provide an overview of engaging this online community that may inform future physical education and promotion.

Physical activity and health education and promotion researchers and practitioners can work to improve health behaviours including physical activity through promotional messages with wide reach (Neiger et al., 2013). Social marketing theory constructs including behaviour change, the cost of behaviour change, strategies that support and ease adopting the behaviour, and messaging strategies including platforms, promotion and reinforcement can be used by applying social media platforms such as Twitter.

To increase the reach of interventions within the specialist communities represented on Twitter, the use of a variety of hashtags, such as #exercise and #physical activity and post elements (i.e. images and behavioural recommendations), should be considered. Instruction and resources to assist in completing the physical activity(ies) promoted by a tweet and their alignment with theoretical constructs should also be considered. Monitoring posts utilising the codebook developed and tested during this study, and especially the NET-response score, may help practitioners and researchers better understand which items in the codebook result in an improved community response with respect to the health behaviours being promoted.

Limitations

The limitations of this study include the deliberate focus solely on the physical activity hashtag. Future research should consider using a mixed-methods design and include deeper examination of the comments and visuals included with tweets and their alignment with social marketing theory constructs. Widening the scope of the work to include an analysis of related hashtags such as #Exercise will further advance understanding of the relevance of social marketing theory to social media messaging. Future studies should also examine similar processes of communication and effects on other social media platforms beyond Twitter.

Footnotes

Acknowledgements

The authors represent a diverse range of academic and professional disciplines brought together through service in leadership positions of the Physical Activity Section of the American Public Health Association.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Carissa R. Smock

References

Al-Khalifa

Bakhurji

Halawany

, et al. (2021) Pattern of dental needs and advice on Twitter during the COVID-19 pandemic in Saudi Arabia. BMC Oral Health 21(1): 456.

Bergeron

Tanner

Friedman

, et al. (2019) Physical activity communication: A scoping review of the literature. Health Promotion Practice 20(3): 344–353.

Biswas

Faulkner

, et al. (2015) Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: A systematic review and meta-analysis. Annals of Internal Medicine 162: 123–132.

Blair

(2009) Physical inactivity: The biggest public health problem of the 21st century. British Journal of Sports Medicine 43(1): 1–2.

Brenner

Smith

(2013) Pew Research Center Internet & American Life Project Tracking Survey. 72% of Online Adults are Social Networking Site Users. Available at: https://www.pewresearch.org/internet/2013/08/05/72-of-online-adults-are-social-networking-site-users/

Goodyear

Boardley

Chiou

, et al. (2021) Social media use informing behaviours related to physical activity, diet and quality of life during COVID-19: A mixed methods study. BMC Public Health 21: 1333.

Julian

Farley

Beach

, et al. (2020) Do-it-yourself sunscreen tutorials on YouTube. Cancer Epidemiology and Prevention Biomarkers 29(3): 693–693.

Kim

Hansen

Murphy

, et al. (2013) Methodological considerations in analyzing Twitter data. Journal of the National Cancer Institute. Monographs 2013(47): 140–146.

Kokkinos

(2012) Physical activity, health benefits, and mortality risk. ISRN Cardiology 2012: 718789.

10.

Korda

Itani

(2013) Harnessing social media for health promotion and behavior change. Health Promotion Practice 14(1): 15–23.

11.

Kotler

Roberto

Lee

(2002) Social Marketing: Improving the Quality of Life. 2nd ed. Thousand Oaks, CA: SAGE.

12.

Lee

DeCamp

Dredze

, et al. (2014) What are health-related users tweeting? A qualitative content analysis of health-related users and their messages on Twitter. Journal of Medical Internet Research 16(10): e237.

13.

Leroy

Koolippurackal

Swami

, et al. (2021) Reviewing Asthma-related Grey Literature and Personal Opinions on Twitter using LDA and CTM Clustering. Available at: https://nlp.lab.arizona.edu/sites/nlp.lab.arizona.edu/files/AMIA2016-Poster-revision-final-asthma-Twitter.pdf

14.

Lupton

(2018) ‘Better understanding about what’s going on’: Young Australians’ use of digital technologies for health and fitness. Sport Education and Society 25(1): 1–13.

15.

Mather

Cummings

Nichols

(2016) Social media training for professional identity development in undergraduate nurses. Studies in Health Technology and Informatics 225: 344–348.

16.

MacQueen

McLellan

Milstein

(1998) Codebook development for team-based qualitative analysis. Cultural Anthropology Methods 10(2): 31–36.

17.

Moorhead

Hazlett

Harrison

, et al. (2013) A new dimension of health care: Systematic review of the uses, benefits, and limitations of social media for health communication. Journal of Medical Internet Research 15(4): e85.

18.

Neiger

Thackeray

Burton

, et al. (2013) Evaluating social media’s capacity to develop engaged audiences in health promotion settings: Use of twitter metrics as a case study. Health Promotion Practice 14(2): 157–162.

19.

Neiger

Thackeray

Van Wagenen

, et al. (2012) Use of social media in health promotion: Purposes, key performance indicators, and evaluation metrics. Health Promotion Practice 13(2): 159–164.

20.

Neuendorf

(2017) The Content Analysis Guidebook. 2nd ed. Los Angeles, CA: SAGE.

21.

Pew Research Center (2014) Pew Research Center’s Internet and American Life Project. Available at: http://www.pewinternet.org/data-trend/internet-use/latest-stats/

22.

Raggatt

Wright

CJC

Carrotte

, et al. (2018) ‘I aspire to look and feel healthy like the posts convey’: Engagement with fitness inspiration on social media and perceptions of its influence on health and wellbeing. BMC Public Health 18(1): 1002.

23.

Sallis

Adlakha

Oyeyemi

, et al. (2020). An international physical activity and public health research agenda to inform coronavirus disease-2019 policies and practices. Journal of Sport and Health Science 9(4): 328–334.

24.

Sallis

Young

Tartof

, et al. (2021) Physical inactivity is associated with a higher risk for severe COVID-19 outcomes: A study in 48 440 adult patients. British Journal of Sports Medicine 55(19): 1099–1105.

25.

Scanfeld

Larson

(2010) Dissemination of health information through social networks: Twitter and antibiotics. AJIC: American Journal of Infection Control 38(3): 182–188.

26.

Shepherd

Sanders

Doyle

, et al. (2015) Using social media for support and feedback by mental health service users: thematic analysis of a twitter conversation. BMC Psychiatry 15: 29.

27.

Simeon

Dewidar

Trawin

, et al. (2020) Behavior change techniques included in reports of social media interventions for promoting health behaviors in adults: Content analysis within a systematic review. Journal of Medical Internet Research 22(6): e16002.

28.

Towne

Jr Li

Lee

, et al. (2018) Physical activity and associated medical cost savings among at-risk older adults participating a community-based health & wellness program. PLOS ONE 13(6): e0198239.

29.

White

Marsh

(2006) Content analysis: A flexible methodology. Library Trends 55(1): 22–45.

30.

World Health Organization (2022a) Physical activity. Available at: https://www.who.int/news-room/fact-sheets/detail/physical-activity

31.

World Health Organization (2022b) Stay physically active during self-quarantine. Available at: https://www.euro.who.int/en/health-topics/health-emergencies/coronavirus-cov-19/publications-and-technical-guidance/noncommunicable-diseases/stay-physically-active-during-self-quarantine

32.

US Centers for Disease Control and Prevention (2021) The Clear Communication Index. Available at: https://www.cdc.gov/ccindex/pdf/clear-communication-user-guide.pdf