"Twitter Bots and Russian Trolls Amplify the Vaccine Debate": Blame content pollutors first and foremost

"Nyet Comrade"

I'm slightly outside of my comfort zone when talking about the findings reported by David Broniatowski and colleagues [1] which has been reported in the lay media with headlines such as: "Russian trolls 'spreading discord' over vaccine safety online".

Outside of my comfort zone not just because this topic is one on the periphery of the core material of this blog (see here and see here for examples) but also because I've long held the belief that this blog should be apolitical and areligious. There are plenty of other outlets online to satisfy such needs these days. To reiterate, the typical currency on this blog is peer-reviewed science and science really should be free from politics, religion and 'allegiance'.

But here I am talking about a paper titled: "Weaponized Health Communication: Twitter Bots and Russian Trolls Amplify the Vaccine Debate" and a study which did mix science with politics. Indeed, a paper which seemingly complements a whole debate on how social media in particular, really is the mouthpiece of the masses in a global sense, but at the same time, is a media ripe for 'promoting discord'.

Set within the reality that people say pretty much anything and everything on social media and that you don't really know who you're 'socialising with' when using such resources (ever heard of 'stranger danger'), Broniatowski et al set about looking to "understand how Twitter bots and trolls (“bots”) promote online health content." The specific online health content they were interested in was "vaccine-relevant messages" as part of a scheme of work previously undertaken by some of the authors. On this occasion, they looked at well over a million tweets over a 3 year period (2014-2017) and "estimated the likelihood that users were bots, comparing proportions of polarized and antivaccine tweets across user types." They also "conducted a content analysis of a Twitter hashtag associated with Russian troll activity."

Results: "Compared with average users, Russian trolls..., sophisticated bots..., and “content polluters”... tweeted about vaccination at higher rates." Researchers observed that 'content pollutors' - "bot accounts that distribute malware, unsolicited commercial content and disruptive materials" tended to post more 'anti-vaccine content'. One would assume this was because their aim is to 'distribute malware and related materials' and by posting such provocative material, it is more likely to get people to click 'their way'.

Also: "Whereas content polluters posted more antivaccine content..., Russian trolls amplified both sides." Bearing in mind that the overall numbers of those tweets from accounts linked to that Twitter hashtag with Russian activity was small, very small, as a percentage of the whole dataset - "the researchers found 253 tweets containing the #VaccinateUS hashtag among their sample" - their content was more mixed - "43% were pro-vaccine, 38% were anti-vaccine, and the remaining 19% were neutral" - if not actually more likely to be pro-vaccine or neutral. The authors interpret this by saying "its messages were more political and divisive."

I have to say that I'm left slightly unimpressed with the interpretation being made from the Broniatowski data. I'm not quibbling with the idea that: "Accounts masquerading as legitimate users create false equivalency, eroding public consensus on vaccination" or that "More research is needed to determine how best to combat bot-driven content." Yes, social media typically does not typically have many filters ('evidence-based' or otherwise) and that is a downside to the content that users are exposed to.

But the whole 'Russian trolls' promoting 'discord' and being 'divisive' thing seems to be built on fairly flimsy evidence at best. I say this on the basis of the small number of tweets from those Russian troll accounts and the majority direction of their stance with regards to vaccines (neutral if not pro-vaccine in the majority). The bigger story here is the continued effects of 'content pollutors' and how, through their attempts to spread their 'mal-wares', so vaccination policy is being potentially undermined as a result. I say all that wearing the objective blinkers of science and recognising that I am not affiliated to any particular political group, red or otherwise.

I know there's something of an obsession these days about how 'social media is being used to subvert democracy', and how the population at large are somehow unable to form their own opinions as a function of 'what social media says'. Personally, I've always felt such sentiments belittle people, and indeed, say more about those who subscribe to such 'control' ideas than those who are actually exposed to them.

It's also interesting that, outside of the continuing search - often a desperate search - for examples of how Russia is somehow trying to 'take over the world' (😏), mention of the word 'vaccine' in the paper from Broniatowski and colleagues, taps into another topic which can ruffle feathers. I shouldn't really need to emphasise how important vaccines are as part of the modern-day medical arsenal (read Roald Dahl's account if you need more evidence of this). But there does seem to be a general feeling that if one even mentions something that distracts from the life-saving message about vaccines, one becomes persona non grata in certain quarters. Y'know, like the idea that for some groups - "immunodeficient individuals are more likely to develop complications resulting from live-virus vaccines" - vaccination might not be potentially advantageous, or how side-effects, rare as they are, can occur: "Within 48 hours after immunizations... the patient developed a fever to 38.9°C, inconsolable crying, irritability, and lethargy and refused to walk." By linking Russian trolls to social media posts about vaccination, it's not difficult to further 'demonise' various groups of people; and with it, 'pull in' (mainly) parents, some of whom are not 'anti-vaccine' but rather, have some genuine concerns about the potential effects of vaccination on their children (see here)...

And just in case anyone needs any advice on vaccination, there are lots of helpful resources around (see here and see here).

Update: 25 August 2018. The Broniatowski paper is open-access (see here).

----------

[1] Broniatowski DA. et al. Weaponized Health Communication: Twitter Bots and Russian Trolls Amplify the Vaccine Debate. AJPH. 2018. Aug 23.

----------

Heavy metals, heavy conflicts and autism?

Two papers are presented for your reading delight today, both based on the often contentious issue of heavy metals and autism.

The first paper is from Farida El Baz Mohamed and colleagues [1] (open-access available here) and further substantiates the claim that for whatever reason(s) the levels of various heavy metals seem to be increased or raised in some children diagnosed with an autism spectrum disorder (ASD). The second paper by Janet Kern and colleagues [2] (open-access available here) provides results based on a systematic review of the literature on one specific heavy metal, mercury (Hg), with autism in mind and concludes that: "of the studies with public health and/or industry affiliation, 86 % reported no relationship between Hg and ASD" whilst "among studies without public health and/or industry affiliation, only 19 % find no relationship between Hg and ASD."

The Mohamed paper continues a theme that for some on the autism spectrum (remember: the autisms) there may be quite a bit more to see when it comes to the burden of heavy metals (see here). Based on hair analysis for various heavy metals - "mercury, lead, and aluminum" - researchers looked at samples from 100 children diagnosed with ASD compared with 100 age- and gender-matched controls. "Comparison between cases and controls as regards heavy metal levels shows that the mean levels of the three toxic heavy metals in hair were significantly higher among the studied ASD cases than the controls" was the primary finding. The authors further report that said heavy metal findings correlated with "maternal fish consumptions, living nearby gasoline stations, and the usage of aluminum pans" and that: "Environmental exposure to these toxic heavy metals, at key times in development, may play a causal role in autism."

I know that to talk about issues with mercury for example, can be a contentious topic with autism in mind but as uncomfortable as it might be to accept, there is some evidence of a 'connection' between elevated levels of the stuff and at least some autism (see here). The link is not, I might add, seemingly generalisable to all autism (see here) but for those children/adults presenting with elevated levels of this heavy metal, further inspection is warranted save any further health inequalities appearing when the label of autism is discussed. As to the suggestion of a connection between some autism and lead (Pb) exposure, well, again this is not the first time that this has been mentioned in the peer-reviewed literature following a more general assertion that there is nothing very good to come from when lead and children in particular meet (see here).

The paper by Kern and colleagues pours further fuel on to something of an on-going debate in some autism circles about how much factors such as conflicts of interest and research transparency may impact on autism research. Focused specifically on the issue of mercury and autism (again), the authors argue that public health or pharmaceutical industry affiliation seems to show something of a potential relationship with study outcomes when it comes to mercury and autism research. So: "over 80 % of the studies without public health or industry affiliation found evidence a relationship between Hg exposure and ASD." This contrasted with something quite a bit less when it came to studies with a public health or industry affiliation slant. Further: "The dramatic discrepancy in these results... provides evidence of biased outcomes, indicative of a conflict of interest."

The authorship group involved in the Kern paper have something of a peer-reviewed research history when it comes to this topic [3] including independent findings that hair levels of mercury might show an association with autism [4]. Their mention of mercury related compounds found in certain vaccines [5] as potentially showing a correlation with risk of autism makes for controversy in some quarters in view of the 'hot potato' that such as association has become down the years.

Appreciating that whilst the Kern findings report something of a connection between affiliations and research outcomes, I'm hard-pressed to suggest that this is hard evidence of some sort of deliberate scheme to exonerate mercury of any potential role in the very wide autism spectrum. Taking for example one paper that is cited in the Kern analysis - the paper by Barry Wright and colleagues [6] - which I've discussed before on this blog (see here) who found no overall connection between urinary mercury levels and autism, they did also call for further research in this area in light of 'outliers' being reported in their autism and their learning disability group. To quote: "further research is warranted to better understand whether a subgroup with autism or learning disabilities have mercury poisoning or excretion difficulties." In these days of plural autism (see here) one can see the logic in such a suggestion.

Music: Radiohead - Lucky.

----------

[1] Mohamed Fel B. et al. Assessment of Hair Aluminum, Lead, and Mercury in a Sample of Autistic Egyptian Children: Environmental Risk Factors of Heavy Metals in Autism. Behav Neurol. 2015;2015:545674.

[2] Kern JK. et al. Systematic Assessment of Research on Autism Spectrum Disorder and Mercury Reveals Conflicts of Interest and the Need for Transparency in Autism Research. Sci Eng Ethics. 2015 Oct 27.

[3] Geier DA. et al. Thimerosal: Clinical, epidemiologic and biochemical studies. Clinica Chimica Acta. 2015; 444: 212-220.

[4] Geier DA. et al. Hair toxic metal concentrations and autism spectrum disorder severity in young children. Int J Environ Res Public Health. 2012 Dec 6;9(12):4486-97.

[5] Geier DA. et al. A two-phase study evaluating the relationship between Thimerosal-containing vaccine administration and the risk for an autism spectrum disorder diagnosis in the United States. Translational Neurodegeneration. 2013;2:25.

[6] Wright B. et al. A Comparison of Urinary Mercury between Children with Autism Spectrum Disorders and Control Children. PLoS ONE. 2012;7(2):e29547.

----------

Mohamed, F., Zaky, E., El-Sayed, A., Elhossieny, R., Zahra, S., Salah Eldin, W., Youssef, W., Khaled, R., & Youssef, A. (2015). Assessment of Hair Aluminum, Lead, and Mercury in a Sample of Autistic Egyptian Children: Environmental Risk Factors of Heavy Metals in Autism Behavioural Neurology, 2015, 1-9 DOI: 10.1155/2015/545674




Kern JK, Geier DA, Deth RC, Sykes LK, Hooker BS, Love JM, Bjørklund G, Chaigneau CG, Haley BE, & Geier MR (2015). Systematic Assessment of Research on Autism Spectrum Disorder and Mercury Reveals Conflicts of Interest and the Need for Transparency in Autism Research. Science and engineering ethics PMID: 26507205