COVID-19, Genetics, and Risk: Content Analysis of Facebook Posts Early in the Coronavirus Pandemic

Abstract

The COVID-19 pandemic represented a unique information seeking and sharing context for billions of Internet users across the globe. Although previous research has established that people often seek health information on the Internet, including through social media platforms, there was a political element to pandemic responses that is not typical of health issues such as seasonal influenza or heart conditions. This unique context, in conjunction with the public posting of the novel coronavirus DNA by Chinese scientists in January 2020, begs for analysis of how people used social media in the early phase of the pandemic to make sense of the role of genetics in COVID-19. This study represents such an analysis as a qualitative content analysis of Facebook posts concerning genetics and COVID-19. Data were collected from March through August of 2020 to identify how genetics issues were being shared on Facebook and the types of accounts that were sharing that information. Through analysis, four themes emerged representing Facebook posts about genetics and COVID-19: disease risk, testing, vaccines, and virus characteristics. These posts appeared on eight types of accounts, with five of those representing 88% of the data: education, health, lifestyle, news, and political. Results are interpreted with constructs from media dependency theory and implications for future research are presented.

Research has recently demonstrated the significance of social media for sharing (dis/mis)information about health topics (e.g., Jang & Baek, 2019; Oren et al., 2020). As the coronavirus pandemic of 2020–2021 spread across the world, social media posts about the virus spread just as quickly. As of 2021, there are approximately 4.2 billion social media users across the globe (Kemp, 2021). Of these, Facebook users represent a range of generations and multiple segments of the world population. These users joined the information sharing free-for-all in response to a Science magazine article published online in March 2020 that described the role of genetics in COVID-19 severity (Kaiser, 2020). March 2020 was also the first month that much of the world went into lockdown mode in an attempt to slow the virus spread.

As health scientists and the lay public attempted to make sense of the new virus that was killing thousands of people each day, the role of genetics emerged as a topic of interest. The past decade has seen a large increase in research about genetics in health, ranging from carrier screening technologies to cancer prevention and treatment (e.g., Addie et al., 2016; Canary et al., 2015; Canary et al., 2018; Hay et al., 2018; Holton et al., 2017). Accordingly, it is not surprising that genetic research would gain public attention in the fight against the virus. Chinese scientists made the DNA sequence of the SARS-CoV-2 novel coronavirus publicly available in January 2020. This information sharing led to international efforts to use genetic science to fight the virus.

Despite public attention devoted to genetics and COVID-19 early in the pandemic, very little is known regarding how Facebook users responded to the original Science magazine article and other subsequent news articles that discussed possible connections between the virus and genetics. This issue warrants scholarly attention for several reasons. First, a growing body of research suggests that genetic information can shape perceptions of disease risk (e.g., Shiloh et al., 2015; Lebowitz et al., 2021), and beliefs regarding genetic risk of disease can affect health-related behavior (e.g., Nguyen et al., 2015; Diseati et al., 2016). Second, we know that misunderstandings about genetics and disease risk have previously emerged with the potential to adversely affect public health (e.g., Hogarth et al., 2008; Leighton et al., 2012; McGuire & Burke, 2008). Third, a lack of awareness regarding how accurately and appropriately the public understands genetics and disease risk may present an important barrier to effective communication between public health leaders and members of the public.

The purpose of the current study is to expand the research on public understanding of genetics and disease by exploring how people engaged with news and information about genetics and COVID-19 on Facebook. Our analysis uses media dependency theory (MDT; Ball-Rokeach, 1985; Ball-Rokeach & deFleur, 1976) to frame findings about posted content and accounts where the posts appeared. We begin with a brief review of MDT and relevant literature before describing our study method. After presenting results of the analysis, we discuss theoretical and practical implications.

Media Dependency Theory

Although the general public has long been dependent on the media system for learning about events and information that impact everyday life, the pandemic presented a new media environment of extreme uncertainty and social upheaval. Ball-Rokeach and DeFleur (1976) noted in their elaboration of MDT that dependency “is defined as a relationship in which the satisfaction of needs or the attainment of goals by one party is contingent upon the resources of another party” (p. 6). According to the theory, dependency arises from a need to understand and act meaningfully in the world, and increases when social change and conflict increase. The spring and summer of 2020 represented change and conflict as the pandemic raged on due to the lack of both knowledge about the novel coronavirus and consensus about its seriousness and control, and as racial tensions erupted across the United States and the world. Even though MDT was developed prior to the existence of social media, the theoretical emphases on dependency, ambiguity, and social structure provide a useful framework for analyzing online information dissemination about genetic links to COVID-19 via social media. The precipitating online news article about a potential genetic link to susceptibility to the virus and severity of symptoms was sure to create an information context of ambiguity, anxiety, and attempts to reduce uncertainty.

According to MDT, when people encounter information from a media source, there is often an element of ambiguity about how to interpret the information, whether it is correct or complete, and what it means to them. Ball-Rokeach (1985) asserted that people often depend on their interpersonal networks to interpret incomplete or conflicting information they access through the media. Ball-Rokeach (1998) extended the theoretical application of MDT by asserting that the Internet is not likely to remove dependency relations between media producers and consumers, but rather to change them in scope. Tai and Sun (2007) built on this concept in their analysis of the role of social media during the SARS outbreak in China, finding that Chinese media users also became information producers and disseminators with SMS text messaging and the Internet.

Social media platforms such as Facebook serve as an additional source of interpretation, adding a new layer of semi-media/semi-interpersonal networks that converge around shared interests, values, or experiences. In a recent social media study using MDT, Silberman and Record (2021) found interconnectedness between media (Reddit), audience, and society through information sharing and seeking about smoke-free policies. Media producers leverage these networks to retain dependency on the media system while individuals and groups likewise use social media platforms to change the nature of the dependence relations. This perspective has received elevated relevance in light of recent research that has focused on the role of social media in dissemination of problematic types of information, including the spread of disinformation and misinformation (e.g., Jang & Baek, 2019; Liu et al., 2016; Oren et al., 2020). In the context of COVID-19 and uncertainty surrounding the virus, heightened dependency on media by audiences may lead to adoption of misperceptions if the media consulted is laden with misinformation and/or unverified health rumors. We know that within the current information ecosystem, COVID-19 misinformation exists and circulates on social media sites such as Facebook (Yang et al., 2021) with the potential to shape the trajectory of the pandemic in unpredictable ways (Himelein-Wachowiak et al., 2021).

MDT also accounts for how dependencies differ based on a variety of factors such as the sociocultural context and group membership (Ball-Rokeach & DeFleur, 1976). Ball-Rokeach (1985) noted that the relationship the mass media system has with other social systems creates structural dependency, which also fluctuates over time. Thus, this study also analyzes how group membership, operationalized as Facebook accounts, may account for differences in the dissemination of mass media information about genetics and COVID-19. Ball-Rokeach (1985) noted that when people are in a state of ambiguity about our social world, “individuals will seek information from whatever sources are useful and available” (p. 500). In the great coronavirus pandemic of 2020–2021, Facebook was a useful and available source for a large swath of society.

News, Social Media, and Health Science

Previous research has demonstrated the increased dependence on social media for information dissemination and gathering, as well as uncertainty/ambiguity reduction, during health and other public crises (e.g., Jang & Baek, 2019; Liu et al., 2016; Tai & Sun, 2007).

Interplay of News and Social Media

Social media is actively changing our media landscape by allowing individuals to select news from a range of sources and choose content based on social recommendations (Messing & Westwood, 2014). One of the most prominent social media sites is Facebook, which boasts an average of 2.74 billion monthly active users, remaining a prominent news source in the public sphere (“Global social media stats,” n.d.). Shearer and Mitchell (2021) reported that 53% of U.S. adults responded that they often or sometimes get their news from social media. However, news on social media can be dangerous because sites cannot effectively filter content as it is shared (Allcott & Gentzkow, 2017; Zubiaga et al., 2018). Although fake news sharing is less common than generally believed (Guess et al., 2019), Facebook remains a key site for disseminating fake news (Hopp et al., 2020). Defined as “fabricated information that mimics news media content in form but not in organizational process or intent” seen in both political and health contexts (Lazer et al., 2018, p. 1904), fake news is closely related to misinformation (incorrect or misleading information) and disinformation (purposefully spread false information) and spreads faster and farther than true news (Lazer et al., 2018; Vosoughi et al., 2018).

The 2016 election provided an important look into how Facebook can become a battleground for issues, leading to several studies focusing on what type of information was shared and who shared it (e.g., Allcott & Gentzkow, 2017; Grinberg et al., 2019; Guess et al., 2019; Guess et al., 2020). Although fake news sharing occurs regardless of political party, in general, people are most likely to engage with false or misleading news if they are over 65 years old, ideologically conservative, a Trump supporter, or have a high exposure to fake news with a low exposure to factual news (Balmas, 2014; Guess et al., 2019; Guess et al., 2020). This dissemination of false or misleading news on Facebook ultimately may put individuals at risk because of the difficulty moderating these news outlets.

With the outbreak of COVID-19 occurring during the social media age, news sharing about the virus has varied in credibility. Many studies have focused on how and why false or misleading information is shared about COVID-19 on social media to provide ways to fight against the issue (e.g., Apuke & Omar, 2021; Pennycook et al., 2020; Pulido et al., 2020; Su, 2021). The lack of credible information about COVID-19 due to new technologies and social media caused the World Health Organization (WHO, 2020) to call on everyone to fight the infodemic in the same way they fought the pandemic. The information a person receives about disease from others is associated with their risk and prevention behaviors (Bauch & Galvani, 2013); yet 74% of COVID-19 Facebook posts were linked to news organizations, with only 1% of COVID-19 Facebook posts linked to healthcare or science sites (Stocking et al., 2020).

At the same time, leaders within public health organizations were not completely immune from criticisms of credibility. For example, leadership from the WHO initially did not recommend the use of masks among healthy members of the general public; however, as the science around masks evolved, so did the recommendations about masks in public settings (BBC, 2020; Chu et al., 2020). Unfortunately, this change in recommendation created confusion, was inconsistent with recommendations from some elected officials (including the President of the U.S.), and coincided with a rise in the politicization of public health recommendations including mask-wearing in public (Noar & Austin, 2020; Oreske, 2020). In sum, public trust is not a given, and fostering continued public trust in experts and scientific guidance for policymaking remains vital, especially during crises such as COVID-19 marked by high levels of uncertainty (Bennett, 2020; Cairney & Wellstead, 2021).

Interplay of Health Science and Social Media

Considering the pervasiveness of social media in contemporary society, it is not surprising that social media also plays a large role in health information seeking and sharing (e.g., Glowacki et al., 2016; Robillard et al., 2013). The Pew Research Center reported in 2013 that seven in ten adult Internet users searched for health information online that year (Fox & Duggan, 2013). Although Internet sources include a variety of sources that are not social media, many people include social media sites as they search the web for health information. Hwang (2020) investigated differences in information sources and flu vaccine perceptions and found that while there were positive correlations among various sources of health information with social media sources, social media sources were unique in their negative indirect effects on vaccine uptake through perceived vaccine efficacy when compared to positive indirect effects from other sources such as medical journals and newspaper articles. One implication of Hwang’s (2020) research for the current study is that while social media users tend to also rely on other sources for health information, social media may have a unique impact on attitudes and health choices.

The different effect on vaccine perceptions found in Hwang’s (2020) study comports with other research on the unique role that social media plays in shaping people’s health attitudes. For example, Jang and Baek (2019) found that when participants perceived that public health officials were less credible, they tended to obtain public health crisis information (i.e., a MERS outbreak in South Korea) from online news, interpersonal networks, and social media. Jang and Baek (2019) used media dependency theory to frame their analysis, noting that people turn to alternative sources of information when primary information sources seem less credible or increase ambiguity. Similarly, Khamis and Geng (2020) found in their qualitative study of communication and health professionals that a majority of participants reported relying on social media to both seek and share health information regarding COVID-19. Consistent with Hwang’s (2020) findings, however, participants in the Khamis and Geng (2020) study noted the danger of misinformation on social media. Such misinformation could lead to misperceptions about risk, preventative measures, and outcomes in a public health crisis such as the coronavirus pandemic.

Clearly, online news and social media play a large role in contemporary health information sharing and seeking. With the parallel increased interest in the role of genetics in personal health, the following research questions guided this analysis:

RQ1:

How did Facebook posts present associations between COVID-19 and genetics following an original news article regarding such associations?

RQ2a:

What types of Facebook accounts included posts about genetics and COVID-19?

RQ2b:

How do posts differ by account type?

Method

This qualitative content analysis is part of a larger study analyzing Facebook posts about COVID-19 and genetics during spring and summer 2020.

Data Collection

The research team identified March 27, 2020 as the start date for collecting posts, which corresponded with the date of the first popular press article found referencing COVID-19 and genetics (Kaiser, 2020). We then selected June 15, 2020 as the end date for identifying relevant published news articles, as the majority of U.S. states with initial stay-at-home orders expired by this date (Mervosh et al., 2020). However, in order to account for posts referencing news articles published from March through June that Facebook users continued to share through July and August, the final dataset also includes July and August 2020.

Only the top five U.S. news publications were used to collect data, as the number of relevant news articles regarding COVID-19 and genetics appeared to decrease as circulation numbers decreased. The top five news publications based on circulation were: 1) USA Today, 2) The Wall Street Journal, 3) The New York Times, 4) New York Post, and 5) Los Angeles Times (Cision Media Research, 2019). Key word combinations were used to search each of the publications: coronavirus and genetics, coronavirus and genetic risk, COVID-19 and genetics, and COVID-19 and genetic risk. In order to be included in the study, each news article was required to contain at least one of these key word combinations. All publications were searched four times using each respective key word combination to ensure all relevant articles would be included in data collection. A total of 137 relevant articles met our criteria and were then used to download link sharing data from Facebook using CrowdTangle, a public insights tool owned and operated by Facebook (CrowdTangle Team, 2020). The resulting dataset of Facebook posts (N = 2234) was then sorted by month according to the date of each post, and included only posts written in the English language.

Data Analysis

Text files for each month were uploaded to NVivo version 12, and then we ran a word query for the word “genetic” with the query setting included similar words (i.e., the term “genetic” included “genetics”). Two additional word queries were then conducted for each month specifically for “RNA” and “DNA” to capture text that used those more specific terms. These queries resulted in 2,171 instances of the search terms. Queried terms were used as a starting point for coding, which focused on the meaning of the sentence in which each term appeared. For example, “To identify the virus, the C.D.C. test used three small genetic sequences to match up with portions of a virus’s genome extracted from a swab” was coded as “Genetics in Testing/Development.” Coding units were full sentences unless a surrounding sentence was also needed to provide full meaning. Sentences were also coded for Facebook account type where the post appeared, such as “Political Group” or “Health Page.” The research team collaboratively coded March posts and half of April posts using the “genetic” query results to build the code structure and develop coding decision rules. This constituted 20% of the data using the “genetic” search term. The last half of April was then coded independently to test inter-rater reliability, representing 16% of the “genetic” search term data. The two coders coded almost identically, with a mean coding agreement of 99% across all codes.¹ The remaining data were then divided among the coders to complete. The final code structure includes 1,664 instances coded into four content categories and eight account categories, plus one labeled “unidentified” due to the use of special letters or characters. Table 1 presents the final coding structure with number of instances for each category and constitutive code.

Table 1.

Code Structure

Theme	Constitutive Code	Instances
Account Type		1664
	Political	409
	News	309
	Lifestyle	302
	Health	290
	Education	112
	Individual	67
	Business	66
	Religious	52
	Unidentified	50
	Environmental	7
Virus Characteristics		632
	Background Information	276
	Variants	213
	Virus Spread	92
	Genetic Changes	38
	Genetic Characteristics	13
Vaccines		447
	Background Information	268
	Development	179
Disease Risk		435
	Recent Studies	270
	General Information	107
	Sex Differences	54
	Treatment	4
Testing		144
	Development	110
	Implementation and Results	34

Results

Answers to our research questions are discussed below, elaborating on information presented in Tables 1 and 2 and Figure 1.

Table 2.

Content Themes, Sub-themes, and Examples

Theme	Description	Example
Virus Characteristics*	Describes the behavior, elements, or other characteristics of COVID-19
Background Information	General information about coronaviruses or viruses more broadly	“Although the coronavirus RNA—the genetic blueprint of the virus—was present in the aerosols, scientists do not know yet whether the viruses remain infections…”
Variants	Information on different strains or variants of COVID-19	“Some are genetically identical to each other, while others carry distinctive mutations.”
Virus Spread	References to the virus spreading across populations or regions	“The virus came from the New York area mainly from Europe, not Asia, genetic analysis shows, arriving weeks before the first confirmed case.”
Genetic Changes	References to genetic mutations or changes in COVID-19	“More than 2,000 genetic sequences of the virus have been submitted from labs to the open database NextStrain, which shows it mutating on maps in real time, according to the site.”
Genetic Characteristics	Identifying genetic characteristics about the coronavirus	“The SARS-CoV-2 (COVID-19) was identified and RNA genetically sequenced by Chinese scientists on Jan. 7th.”
Vaccines	References to the role of genetics, genetic technologies, or genetic science in vaccine development or delivery
Background Information	General descriptions of how genetics work, or might work, in vaccines	“Another new approach, mRNA vaccines, works by exploiting messenger RNA—a kind of courier that communicates the genetic instructions for making proteins—to drive immune response.”
Development	Information on the role of genetics, or genetic technologies, in developing vaccines for the virus	“Invovio’s candidate, called INO4900, is a DNA vaccine, which packages a piece of the coronavirus genetic code inside synthetic DNA.”
Disease Risk	References to the role of genetics in increased or decreased risk for infection or symptom severity for COVID-19 or viruses more broadly
Recent Studies	Specific studies, findings, or methods related to genetic variants in individuals or genetics in disease risk broadly	“Meanwhile, new research is starting to solidify that there are genetic differences in how seriously ill people get if they become infected.”
General Information	General information about genetics in immune systems or immune responses	“Genetics may be a factor.”
Sex Differences	Differences in genetics between males and females that impact health	“What lies behind this female genetic superiority?”
Treatment	How treatments may reduce severity of symptoms or risk of dying from the virus	“Discovering any gene variant underlying severe COVID-19 infections would have many potential uses, Dr. Cassanova said…”
Testing	Describes the role of genetics in COVID-19 testing or other references to testing programs and development
Development	References to the role of genetics in developing tests for the virus	“To identify the virus, the C.D.C. test used three small genetic sequences to match up with portions of a virus’s genome extracted from a swab.”
Implementation and Results	Implementation of test protocols involving genetics, or outcomes of testing that references genetics	“Testing so far has relied on detecting the nucleic acids of the virus’s genetic material.”

Figure 1.

Facebook Account Comparisons by Theme

Research Question 1

Research Question 1 asked, “How did Facebook posts present associations between COVID-19 and genetics following an original news article regarding such associations?” Most themes were found in several months; however, each theme tended to have one or two months when the theme was more prevalent than other months. Table 1 highlights the themes with frequency counts, presented in descending order of number of instances. Table 2 presents descriptions and examples for each theme and constitutive code, summarizing results discussed below.

Virus Characteristics

Posts in this category describe behavior, elements, or other characteristics of the novel coronavirus. It highlights how COVID-19 and other viruses function in the population and at the genetic level. This theme was present in every month and was the most common theme by far; however, it was most prominent in April and May, with these two months comprising nearly all of the data in this theme. This theme includes five sub-themes, described below.

Background information.

Background information was the most common of the four sub-themes and was found most often in April. Posts in this category included general information about either the coronavirus or viruses more broadly. This sub-theme illuminated how Facebook posts explained the interactions between viruses and the population. For example, an article noted:

…even if those virus particles are no longer active or capable of infecting humans, they may still carry genetic material that can be detected using an approach called PCR (polymerase chain reaction), which amplifies the genetic signal many orders of magnitude, creating billions of copies of the genome for each starting virus.

This type of general overview was also seen when referencing how information about the coronavirus was collected, such as stating “an international guild of viral historians ferrets out the history of outbreaks by poring over clues embedded in genetic material of viruses taken from thousands of patients.” Thus, this theme represents how posts drew attention to associations between COVID-19 and genetics as experts and lay people alike were learning about this new lethal virus.

Variants.

This sub-theme included any information about different strains or variants of the novel coronavirus; however, it did not include references to genetic mutations or changes, which were placed in the Genetic Changes category. In addition, most posts within this theme considered how genetic data was used to identify different strains of COVID-19 throughout the world or used genetics to describe the virus itself. Mentions of variants appeared most often in May. Sentences in this category may refer to the number of mutations (e.g., “the novel coronavirus has mutated into at least 30 different genetic variations, according to a new study in China.”) or where the strains came from (e.g., “Heguy and the team have pinpointed strains originating from European countries, including France, Austria, and the Netherlands, in two-thirds of the patients they’ve conducted genetic sequencing of the virus.”).

Virus Spread.

This sub-theme was most common in April. Focusing on how COVID-19 can be traced and understood genetically, this theme highlights how the virus moved throughout the United States and how it generally spreads. Facebook posts that discussed virus spread often specifically referenced how the genetics of COVID-19 allows individuals to trace the virus. For example, an article explained:

…two separate teams of scientists studying the genetics of the SARS-CoV-2 virus in the region came to similar conclusions: People were spreading the virus in New York as early as late January, before more widespread testing began, and it came mostly from Europe, not Asia.

This sub-theme also included specific references to how the novel coronavirus could spread from person to person (e.g., “Adding to the growing evidence that the novel coronavirus can spread through air, scientists have identified genetic markers of the virus in airborne droplets, many with diameters smaller than one-ten-thousandth of an inch.”). Considering the level of ambiguity around the virus and the changing messages coming from public health officials, it makes sense that the lay public was turning to Facebook posts to increase understanding about virus spread and act meaningfully in the face of the life-threatening disease.

Genetic Changes.

Occurring most often in April, this sub-theme referenced the specific ways genetic components of the novel coronavirus occur and can be used, with posts attempting to explain the genetic changes seen within the COVID-19 virus. In one instance, an article noted,

“Scientists around the world are also racing to use small genetic changes in the virus — biological markers that act as something like fingerprints for disease detectives — to help map how the pathogen swept across the country and around the world.”

However, this category also included references to how the coronavirus can change genetically, with one article explaining, “As the coronavirus passes from one person to another, it can mutate, or change its genetic information slightly, Heguy said.”

Genetic Characteristics.

This sub-theme occurred exclusively in April and June. The theme included references to identifying the genetic characteristics of the coronavirus. Articles referenced in these posts explained the genetic characteristics, specifically focusing on the benefits of tracing genetic information of viruses. For example, one article stated, “Sequencing the virus’ genetic information now will also help in case a second wave were to occur, she said.” Posts also described how genetic characteristics of the novel coronavirus were initially shared, such as with one article explaining, “The SARS-CoV-2 (COVID-19) was identified and RNA genetically sequenced by Chinese scientists on Jan. 7th.” These last two sub-themes underscore the increasing role that genetics is playing in the public understanding of health and health science.

Vaccines

Genetics in vaccines included sentences that discussed the role of genetics, genetic technologies, or genetic sequences present in vaccine development or delivery. Each month, excluding July and August, included at least some mention of genetics in reference to vaccines; however, references were most common in May and, to a lesser extent, June. Two sub-themes constitute this theme, described below.

Background Information.

Background information appeared most frequently in May and June. This subcategory includes posts about the current and future landscape of vaccinations. A significant portion of the information focused on broad descriptions of how genetics have worked in vaccines (e.g., “As a whole, DNA vaccines are known to be very safe, Dr. Klinman has written”). In addition, the background information included how genetics could work in specific types of vaccines (e.g., “Gene-based vaccines, such as DNA vaccines and RNA vaccines do not consist entirely of the virus particle” and “The virus delivers the spike-protein DNA into the cells, but it cannot replicate in cells, so it’s a safe delivery system”).

Development.

Text that specifically referenced the role of genetics, or genetic technologies, in developing vaccines for the virus were included in the development category. Posts about development were far more common in May than June. Although these references were less common than the background information, the development category highlighted the complexities associated with the COVID-19 vaccine. Posts in this category engaged with genetics in vaccines through highlighting specific ways the coronavirus vaccine was being developed (e.g., “Moderna uses genetic material — messenger RNA — to make vaccines, and the company has nine others in various stages of development, including several for viruses that cause respiratory illnesses”).

Disease Risk

Disease risk highlights the role of genetics in increased or decreased risk of infection or symptom severity of COVID-19 or viruses in general. This theme includes four sub-themes. Each sub-theme highlights how genetics plays a role in our current conceptualization of COVID-19 risk. Genetics in disease risk posts were most prevalent in June.

Recent Studies.

This theme included references to specific studies (e.g., “Now, a study by European scientist is the first to document a strong statistical link between genetic variations and Covid-19…”), findings (e.g., “…that initiative’s collected data were beginning to point to a single spot on Chromosome 3 as a potentially important player.”), or methods (e.g., “The researchers did not sequence all three billion genetic letters in the same genome of each patient.”). Thus, articles often presented new and emerging information for readers, which potentially allowed them to better understand how genetics plays a role in COVID-19 severity.

General Information.

Articles within the Facebook posts often included general information related to disease risk by providing information about genetics related to the immune system or the immune response; however, they did not reference specific cases of immune system response. Examples include, “genetic factors explain the risk, at least in some cytokine storms” and “Is it something about your biology or some genetic predisposition?” In addition, some articles explained that “it is theoretically possible that genetic variations influence” immune response. These sentences often made broad statements regarding genetics and COVID-19 and informed readers about the impact of genetics on how a person’s body response to illness.

Sex Differences.

Sex differences included any references to genetic differences between males and females that impact their health and was only present in the month of April. These differences were primarily found in one article shared across multiple Facebook groups that focused on why biological females were often able to recover better and faster than biological males. For instance, one post quoted, “the bottom line is when it comes to dealing with the trauma and stressors of life — whether it’s avoiding a serious congenital malformation, a developmental disability, or fighting off an infection — females have genetic options.” Another post of the same article stated, “for the most part, the medical establishment has largely overlooked the profound chromosomal, hormonal, and anatomical uniqueness of genetic females.”

Treatment.

This theme was by far the least common and only occurred in April, but it includes interesting references to how treatments may reduce the severity of coronavirus symptoms or reduce the risk of dying from the virus. This theme includes sentences that specifically mentioned potential treatments (e.g., “Remdesivir and favipiravir, for instance, each mimics a key building block in a virus’s RNA, which, when inserted, keeps the virus from replicating.”) and how understanding COVID-19 may provide important future uses (e.g., “…it could pave the way for prevention or treatment of disease, make it possible to apply existing drugs that help address a genetic deficiency, or simply inform future genetic counseling of family members.”). Thus, even though it was much less common than the other themes within the Disease Risk category, it provides important information regarding COVID-19 and potential treatments using and impacted by genetics.

Testing

This theme involves the role genetics played in COVID-19 testing and references to testing programs. This theme mainly occurred in March and April, with only a few references in May. By June, there was no further mention of the relation between genetics and testing. This theme includes two sub-themes, described below.

Development.

These references occurred mostly in March and included sentences that explained the role genetics played in developing tests COVID-19. Many articles focused on how Chinese scientists were able to provide the genetic sequence of the virus, which significantly informed test development. For example, one article stated, “they note that a Chinese laboratory posted the genetic sequence of the virus in early January, making it possible for laboratories across the world to start working on diagnostic tests.” Also, “by Jan. 20, just two weeks after Chinese scientists shared the genetic sequence of the virus, the C.D.C. had developed its own test…”

Implementation and Results.

This sub-theme includes any references to the outcome of a test or testing protocol that related to genetics. These posts appeared most often in April and included information about how the tests were implemented (e.g., “The rush for widespread testing has also created unprecedented global demand for essential testing components…”) as well as how the genetics of the test work (e.g., “it converts cells’ RNA and DNA, and then, using polymerase enzymes, duplicates the DNA again, so that there’s enough of the virus that it can be detected, if it is present at all.”).

Overall, content themes demonstrate that people were turning to social media, specifically Facebook, to satisfy their informational needs in a highly ambiguous and threatening public health context. The socio-cultural positionality of these individuals is explored below, addressing how media dependency relations differ based on sociocultural context and group membership.

Research Question Two

Research Question 2a concerns types of Facebook accounts that included posts about genetics and COVID-19, whereas Research Question 2b asks if there are differences in terms of content posted among those account types. Refer to Table 1 for frequencies of each account type.

Account Types

We present account types with genetics-COVID-19 posts in descending order of frequencies, from account types with the most posts to those with the least.

Political.

Political Facebook accounts include pages for official political parties, for individual politicians posting as a politician, and pages or groups that identify their account as being about politics. Examples of political accounts include “Deny Barisan Nasional Majority in Next Election!” and “Forward 2020 USA.” Political accounts constitute 25% of the coded data.

News.

News Facebook accounts include pages for mass media sources, such as television stations, newspapers, podcasts, and radio stations, as well as regional news pages, and groups devoted to sharing general news about a region or locality. Examples of news accounts include “The Tom Joyner Morning Show” and “Ventura LGBT News.” News accounts constitute 19% of the coded data.

Lifestyle.

Lifestyle Facebook accounts include pages about living in a particular neighborhood or city, groups devoted to hobbies or interests, entertainment pages, and other pages or groups with the purpose of posting information about lifestyle or cultural issues. Examples of lifestyle accounts include “Animal Kingdom” and “Old is the New Black.” Lifestyle accounts also constitute 19% of the coded data.

Health.

Health Facebook accounts include pages dedicated to disseminating COVID-19 information, healthcare organizations, healthcare professional networks, and pages devoted to people with specific health conditions. Examples of health accounts include “COVID-19 St. Louis & St. Charles Support Group” and “Flatten the Curve Texas.” Health accounts constitute 18% of coded data.

Education.

Education Facebook accounts include pages from formal education institutions as well as pages focused on educating members about a variety of issues. Examples of education accounts include “Evolutionary Psychology News” and “National Academy of Young Scientists.” Education accounts constitute 7% of coded data.

Business.

Business Facebook accounts include pages representing companies as well as groups devoted to business issues or professional networking. Examples of business accounts include “KineticoAZ” and “Millennium Protective Services Inc.” Business accounts constitute only 4% of coded data.

Individual.

Individual Facebook accounts include only private individuals posting on behalf of themselves. Politicians were coded as Political and individual healthcare providers identifying themselves by their health profession were coded as Health or as Business if they were promoting their health-related business. Individual accounts also constitute only 4% of the coded data.

Religious.

Religious Facebook accounts include pages by organized religious entities, such as a church or denomination, as well as religious figures and pages or groups concerned with general religious issues. Accounts focusing more on spirituality than a specific religion were coded as Lifestyle. Examples of religious accounts include “Group-Biblical and Prophetic News-Noticias Biblicas y Profeticas” and “End Time Evidence & Prophecy Updates.” Religious accounts constitute only 3% of the coded data.

Environmental.

Environmental Facebook accounts include pages focused on environmental issues, but not businesses promoting environmentally related products. Examples of environmental accounts include “Proyecto Planeta – Iniciativa Talento + Consciencia” and “Real Coastal Warriors.” Environmental accounts constitute only 0.04% of the coded data.

Account Differences

The vast majority of posts included in this analysis (88%) appeared in political, news, lifestyle, health, and education accounts. Figure 1 presents a comparison of how thematic categories differed across account type. Specifically, the politicization of the pandemic is clearly demonstrated with the 25% of data excerpts being posted on political Facebook accounts, with political accounts representing the largest number of posts about testing (N = 73), vaccines (N = 121), and virus characteristics (N = 165). While news, lifestyle, and health accounts were roughly similar in the raw number of posts included in this analysis, themes emerge differently across these account types. For example, lifestyle accounts had more references to disease risk (N = 122) than other account types and the third most references to virus characteristics (N = 119). News accounts, on the other hand, are fairly well represented across all four content themes (see Figure 1). Surprisingly, health accounts included very few references to testing (N = 8) and relatively few references to vaccines (N = 30). Health accounts were well represented in posts including information about disease risk (N = 118) and virus characteristics (N = 132). Differences across remaining account types are minor due to the small number of posts in those account types (Figure 1).

Discussion

The goal of this study was to identify how Facebook users were circulating information about the role of genetics in various aspects of the COVID-19 pandemic during the early months of the pandemic, following the online posting of a news article about genetics and disease severity. This socio-historical context represented heightened ambiguity and conflict in societies across the globe, leading to our use of MDT to frame our interpretation. Results of our analysis not only revealed themes of content being posted on Facebook regarding genetics and COVID-19, but also compared those themes across types of Facebook accounts. Results lead to interesting conclusions about the interplay of health and politics in the early months of the pandemic with both theoretical and practical implications.

Theoretical Implications

It is commonly accepted that one factor leading to the rapid and wide spread of the virus was the lack of knowledge about virus characteristics that would provide information about prevention and treatment. Indeed, the most predominant theme in posts included in this analysis was “Virus Characteristics” as Facebook users posted information and articles about what was known, being discovered, and yet to be known about the virus. Many of these articles and posts included basic genetics information to which most lay people have not had exposure. Such background information was also found in discussions about disease risk and vaccines, indicating that Facebook users were reading articles about the virus in mass media and then turning to their social media communities to help make sense of what was being disseminated through the media system by official sources such as governments and public health officials.

This finding is consistent with previous research about media dependencies and social media when public sources are not trusted or when there is considerable uncertainty about information provided (e.g., Hwang, 2020; Jang & Baek, 2019; Khamis & Geng, 2020; Tai & Sun, 2007). Ball-Rokeach (1985) asserted that people use their interpersonal networks to understand and act meaningfully in the face of incomplete or conflicting information. Tai and Sun (2007) extended this theoretical principle in their study of SMS and social media use in China during the 2003 SARS outbreak. Results of the current analysis further demonstrate this MDT explanatory mechanism as the informational ecosystem early in the COVID-19 pandemic was rife with contradictory messaging about mask wearing, the origins of the novel coronavirus, and how COVID-19 was spreading. People across the globe turned to their Facebook social networks to help them interpret messages about the virus. Furthermore, posts across account types indicate that, much like users in the Tai and Sun (2007) study, people were taking it upon themselves to become information producers and disseminators, not just consumers of information from official sources.

A basic assumption of MDT is that dependencies increase when people perceive the need for information and conversely decrease when media information stops being useful to media consumers (Ball-Rokeach & DeFleur, 1976). Results of our analysis demonstrate this waxing and waning of media dependency about COVID-19 topics as most themes were predominant in one or two months then decreased significantly in other months. For example, testing was a topic that was posted about mostly in March and April, then it all but disappeared in later months. However, virus characteristics, representing emerging knowledge across the data collection timeframe, continued to be posted across months. This finding builds on results from the Silberman and Record (2021) study in which a Reddit post was created to investigate interactivity regarding new smoke-free policies. Whereas that study demonstrates dependencies on social media, the current study also demonstrates how such dependencies change over time as the information context changes.

The comparative analysis also sheds light on how dependencies changed in the early months of the pandemic. The news account type, which includes mass media outlets as well as more local or individualized news pages (such as podcasts), had fairly consistent posting frequencies across themes (Figure 1). However, every other account type had certain themes that were either under-represented or over-represented. Accordingly, we can conclude from this comparison that members of certain types of Facebook accounts were more or less dependent on media information about different themes. Most notable is the difference in content themes between health accounts and political accounts. Postings with references to vaccines and testing were much more prevalent on political accounts than on health accounts. This result points to the highly charged political climate of 2020 as well as early concerns among health communities about mitigating risks of infection and severe symptoms or death. Disease risk is less “political” than topics such as testing programs and vaccination protocols, especially considering the reliance of such programs on governmental health systems and resources. Furthermore, filter bubbles, which filter content using algorithms based on user preferences, and echo chambers, which reinforce pre-existing ideological beliefs of users, could explain why more content about testing programs and vaccination protocols was magnified on political accounts (Kitchens et al., 2020). These science topics and their implications for economic and public policy were politicized as the pandemic crisis unfolded.

Practical Implications

This project was first conceived due to the unique media discussion about genetics and COVID-19. Such public and mass media attention to genetics in virology had not emerged in prior public health crises such as influenza outbreaks. This emergence of genetics as an explanation for disease risk, a key to testing, treatment, and vaccines, and as part of the description of a new virus points to the emerging relevance of genetics in the larger health conversation. As more people are exposed to the role of genetic science in cancer treatment and screening for a variety of heritable disorders, it is unsurprising that the media system would seize on the topic for disseminating information about this new public health crisis.

Research regarding access to and uptake of genetic and genomic testing for a variety of medical conditions has shown that barriers remain to reach certain populations, such as non-white patients and those with lower education levels (e.g., Addie et al., 2016; Hay et al., 2020). Results of this study indicate the importance of social media and social networks for reaching various audiences with health information that resonates with them. Furthermore, such outlets represent a type of interpersonal network that people trust more than political, governmental, or public health officials, providing an important link in the audience-media-society interrelationship that makes up the media system and sociocultural context.

Limitations and Future Directions

Although findings of this study are useful for understanding how Facebook users used their social media communities to share and interpret genetic information in mass media articles about COVID-19, there are inevitably limitations to the study design and findings. First, we originally conceived the study to include Instagram and Twitter data. There were not enough Instagram posts meeting inclusion criteria to justify keeping that platform in the dataset. However, Twitter data would have been useful for seeing a broader spread of genetics information about COVID-19. Unfortunately, the tool used to collect the data, CrowdTangle, does not have the capacity to also gather Twitter data at this time. Our data were also limited to posts written in English, lacking the ability to generalize findings of this study to posts written in other languages. Future research with appropriate data gathering capabilities can build on these study results by examining other social media platforms as well as posts written in languages other than English.

A second limitation of this study is that we limited our analysis to sentences including the key terms of interest, “genetic,” “RNA,” and “DNA.” Our qualitative approach prevented us from coding entire articles or posts. Other themes would likely emerge regarding the articles and posts if the analytic scope were widened. Similarly, we did not code for political bias in the political account type. It is possible that more conservative political Facebook accounts had different emphases in posts than did more liberal accounts. This limitation could also be addressed in future research that codes for political bias in Facebook accounts.

Finally, future research should continue to attend to the role of genetics in the mass media and social media conversations. Genetics is quickly becoming a central feature of health sciences research. As researchers continue to increase understandings of the role of genetics in healthcare and health outcomes, it will be important to leverage the unique power of social media to keep the public informed and engaged around these issues.

Additionally, the findings of our study also have implications that extend beyond health to other science-related topics, especially in areas that intersect science and public policy. It is possible that for other complex and evolving science topics, we might see similar patterns of media dependency and bifurcation of political versus nonpolitical aspects in public deliberations surrounding these topics. For example, science topics related to climate change, renewable energy, environmental conservation, and genetically modified organisms that can have socio-economic impacts on a regional or global level may represent areas in which such patterns are consistent with what we found in the current study, especially if they occur concurrently with a large-scale crisis.

Conclusion

The coronavirus pandemic is the first global public health crisis that emerged in the age of social media. Likewise, it is the first such health crisis to emerge in the age of broad genetic research advances in healthcare. The convergence of these two factors, combined with the social and political upheaval of the first months of the pandemic in 2020, created a unique social media context for studying media dependencies across Facebook groups in the early stage of the pandemic. Findings reveal the role that social media plays in how people enact their media dependencies for health information. Genetics in healthcare is here to stay. The inextricable relationship between health and politics is also likely here to stay. Future research can build on the findings of this descriptive study to further understand the role of social media in public understandings of genetics in healthcare and health outcomes.