COVID-19 vaccination and involuntary movements: A longitudinal panel study

Objective: Despite high COVID-19 vaccination rates in many populations, concerns persist about potential adverse events, including concerns about involuntary movements. While case studies have shown occurrences of involuntary movements following COVID-19 vaccination, no systematic studies have explored this association. Our study aims to investigate the relationship between COVID-19 vaccination and involuntary movements. Methods: This study employs a longitudinal panel design. The study population consists of 165,834 responses from a total of 97,537 unique individuals sourced from the BiCoVac cohort, which is a randomly sampled cohort of Danish individuals aged 16 to 65. Data were collected through a combination of questionnaires and national registers, and analyses were conducted using mixed effects logistic regression. Results: Vaccinated individuals had lower odds of reporting involuntary movements compared to non-vaccinated individuals. Although adjustments attenuated the results, a consistent pattern of lower odds was observed among the vaccinated individuals. The strongest association for the first dose was observed in individuals who received the vaccine within the last 4 weeks before reporting symptoms (OR = 0.72 (0.60; 0.85)). For the second dose, the strongest association was found in individuals who received the second vaccine dose more than 4 weeks before reporting symptoms (OR = 0.77 (0.65; 0.91)). Conclusion: The results of the study do not indicate involuntary movements as an adverse reaction to the COVID-19 vaccine. These findings support the safety profile of the COVID-19 vaccine concerning involuntary movements and contribute to enhancing public trust in vaccination programs.


Introduction
In January 2020, the World Health Organization reported a potential pneumonia outbreak in Wuhan, China, which escalated into one of five recorded pandemics in history, now recognized as the coronavirus disease 2019 (COVID-19) pandemic [1].In Denmark, the first COVID-19 case was reported in February 2020, and vaccination commenced on December 27, 2020 [2,3].The rapid rollout of the vaccine has played a pivotal role in the global effort to control the pandemic, with over 13.5 billion vaccine doses administered worldwide [4].However, the rapid rollout of the vaccine has raised public concern about potential adverse events [5].One area of concern pertains to involuntary movements.In 2021, a video circulated on social media showing a middle-aged woman allegedly undergoing involuntary movements, which she claims were developed after receiving a COVID-19 vaccine [6].Despite efforts by news outlets and health authorities to combat claims about adverse effects, the video reappeared on social media in 2023 and garnered over 72 million views, reigniting concerns over the safety of COVID-19 vaccination [7].Furthermore, a growing number of case studies describes occurrences of hyperkinetic movement disorders following COVID-19 vaccination, involving vaccines developed by Pfizer/Bio-NTech, AstraZeneca, Sinopharm, and Moderna [8][9][10][11][12][13][14][15][16][17][18].These disorders are characterized by abnormal involuntary movements, distinct from non-pathological involuntary movements like yawning or blinking [19].The case studies encompass a diverse field of individuals, including both young and elderly individuals, as well as men and women.To our knowledge, there have been no comprehensive investigations exploring the association between the COVID-19 vaccine and involuntary movements, making this study the first of its kind.

Study design
The study is designed as a longitudinal panel study, involving repeated self-reported information from the participants in the BiCoVac cohort.During the study period, spanning 8 months (May 2021 to December 2021), the cohort was assessed on three occasions through three questionnaires.The self-reported questionnaire data was subsequently linked to register data.

The BiCoVac cohort
BiCoVac is a Danish population-based cohort initiated in April 2021 [20].The cohort was established to gather information on immediate and long-term symptoms following COVID-19 vaccination.It comprises a sample of 911,613 Danish citizens aged 16-65 (representing 25% of the target population), randomly selected from the Danish Civil Registration System.Baseline questionnaires were distributed in May 2021 via the national digital mail system, e-Boks, before widespread national vaccination.Follow-up questionnaires were distributed stepwise by age in accordance with the national vaccination plan [21].The first followup questionnaire was distributed 2-4 weeks after the scheduled age groups were set to receive their first vaccine dose (primarily June-July 2021).The second follow-up questionnaire was distributed approximately three months after the various age groups had been offered the second vaccine dose (October 2021).Invitation to follow-ups required prior questionnaire completion.
In the present study, only participants who had completed at least the first follow-up questionnaire and had not been vaccinated against COVID-19 when they completed the baseline questionnaire were included.Also, individuals were excluded if they did not have information on all included variables.Questionnaires were available in Danish, English, and Arabic.The vaccines available in the national COVID-19 vaccination program during the BiCoVac study period were primarily from Pfizer/BioNTech and Moderna.

Register data
As the BiCoVac cohort is derived from the Danish Civil Registration System (all Danish citizens receive a unique 10-digit civil registration number), it facilitated the linking of the personal identification number of all cohort members with register data.Statistics Denmark, the primary authority for Danish statistical information, provided data on highest completed education, prior ICD-10 diagnoses (as reported by the National Patient Register), ATC codes of prescribed medicines (as reported by the Register of Pharmaceutical Sales), COVID-19 test results, and COVID-19 vaccination dates for the entire cohort.The COVID-19 test results and COVID-19 vaccination dates were monitored by Statens Serum Institut, who operates under the Danish Ministry of Health.

Exposure and outcome
The occurrence of involuntary movements, which is the outcome of this study, was self-reported and collected through the questionnaires.The outcome was assessed by asking: "During the past four weeks, how much have you been bothered by?" followed by a list of symptoms that included involuntary movements, without specifying what constitutes an involuntary movement.Responses were collected using a 5-point Likert scale with the options: 1 "Not at all", 2 "A bit", 3 "Somewhat", 4 "Quite a bit", and 5 "A lot".The outcome variable was dichotomized into 1-2 = no and 3-5 = yes.
Information on vaccination dates and number of COVID-19 vaccine doses received was derived from the register data.The exposure variable was constructed to account for the time interval between the most recent administered vaccine dose and the completion of the questionnaire.This allowed differentiation between potential short-term and long-term adverse events.Hence, 5 exposure groups were used: 0 dose, 1 dose ≤4 weeks, 1 dose >4 weeks, 2 doses ≤4 weeks, and 2 doses >4 weeks.For individuals who had received both vaccine doses by the time of completing a questionnaire, the time interval indicates the period between the date of completing the questionnaire and the date of receiving the second vaccine dose.In situations where the questionnaire was filled out on the same day as receiving a vaccine dose, the vaccine dose was considered as having been administered before completing the questionnaire.Ad hoc analyses were performed under the assumption of the opposite scenario: that those vaccinated on the same day had answered the questionnaire prior to receiving the vaccine.These analyses increased the study population by 1814 responses from 1028 unique individuals, with a minuscule impact on the results.

Adjustment variables
To reduce potential confounding, we adjusted for several covariates.These were selected a priori.The following variables, collected from the baseline questionnaire, were adjusted for: stress based on Cohen's 10-item Perceived Stress Scale [22] (low (scores between 0 and 13), moderate (14-26), high (27-40)), vaccine hesitancy (no, skeptical, intending to remain unvaccinated, not sure/prefer not to answer), physical activity (sedentary, active at least four hours a week, active at least three times a week, competitive athlete), smoking (never, former smoker, occasionally, daily smoker), and baseline involuntary movement symptom score (categorized on the original 5-point Likert scale).Adjustment variables collected from the register data included: sex (male, female), highest level of education completed (primary education, secondary education (e.g.high school or vocational education), higher education (e.g.undergraduate education), postgraduate education, missing), prior diagnosis involving symptoms of involuntary movement (no, yes), prior redemption of prescribed psychotropic medication (no, yes), age (16-25, 26-35, 36-45, 46-55, 56-65), and prior COVID-19 infection (no, yes).Prior redemption of prescribed psychotropic medication, age, and prior COVID-19 infection were time-varying variables, while the other adjustment variables were time-invariant.Prior diagnosis involving symptoms of involuntary movement included the following ICD-10 codes, whether recorded as the primary or secondary diagnosis: G10, G11, G20-G26, G40, G41, G51, G80-G83, F44.5, F84.2, and F95.Prior redemption of prescribed psychotropic medication, collected from primo 2020, included the following ATC codes: N05A, N05B, N05C, N06A, and N06B.In cases where the date of completing a questionnaire coincided with the date of a positive COVID-19 test result, the test result was considered an infection prior to the response.The test date used was the first positive test, whether it was a PCR test or an antigen test.A positive antigen test followed by a negative PCR test within 7 days was considered a false positive test result.

Statistical analysis
Characteristics of the study population were described stratified by C.M. Nielsen et al. exposure group.Unadjusted and adjusted mixed-effects logistic regressions were used to estimate odds ratios (OR) for the association between COVID-19 vaccination status and involuntary movements.To account for the same respondents answering surveys over time, survey results were nested within each respondent.The adjusted analyses included adjustment for age, sex, education, prior diagnosis involving symptoms of involuntary movement, prior redemption of prescribed psychotropic medication, stress, vaccine hesitancy, physical activity, smoking, prior COVID-19 infection, and baseline involuntary movement symptom score.
To address potential attrition bias that could be associated with unbalanced panel datasets, the dataset was balanced in a sensitivity analysis, meaning that the analysis exclusively included participants who completed all three questionnaires.In a final sensitivity analysis, where healthy vaccinee bias was addressed, individuals who had not received any vaccines during either of the follow-ups were excluded.The rationale for addressing this bias is that if individuals more prone to experiencing involuntary movements were also more inclined to abstain from vaccination, it could introduce a bias favoring a protective effect of the vaccine.
Statistical analyses were performed using statistical software (Stata/ MP 17).

Ethics
The BiCoVac study was approved by the Danish Data Protection Agency under the Aarhus University comment agreement (journal number 2015-57-0002) and Aarhus University journal number 2016-051-000001, sequential number 2272 (25/3-2021).According to Danish legislation, ethical approval of survey studies based on questionnaires is not required.

Data statement
The dataset is not publicly available due to national data security legislation on sensitive personal data.License to the data is subject to conditions stipulated by the Danish Data Protection Agency and Aarhus University.vaccines when completing the baseline questionnaire and were thus excluded.Of those having e-Boks at the time of follow-up questionnaire distribution, 97,537 completed the first follow-up questionnaire and 68,297 completed the second follow-up questionnaire (Fig. 1).The final study population consisted of individuals who completed the first follow-up questionnaire and those who completed both follow-up questionnaires.A single person could contribute to the study population twice by completing both follow-ups, which led to a study population of 165,834 responses from a total of 97,537 unique individuals.Non-vaccinated individuals (those within the 0-dose exposure group) generally differed from the other exposure groups based on the covariates (Table 1).Non-vaccinated individuals were generally younger and had the highest proportion of females.Also, non-vaccinated individuals more often; had a prior diagnosis involving symptoms of involuntary movements, had moderate to high stress, reported vaccine hesitancy, were smokers, and more often had a prior COVID-19 infection.

Of
The odds of involuntary movement were overall lower among vaccinated exposure groups compared with those not vaccinated (Table 2).Adjustment generally attenuated the results, but a tendency towards lower odds was seen in all exposed groups, although only statistically significant for those having received the first vaccine dose within the last 4 weeks before reporting symptoms (OR = 0.72 (0.60; 0.85)) and for those who received the second vaccine dose more than 4 weeks before reporting symptoms (OR = 0.77 (0.65; 0.91)).
In the balanced sensitivity analysis, 29,240 unique individuals were excluded as they did not complete all three questionnaires.The balanced models still indicated that COVID-19 vaccination did not increase the odds of involuntary movements (Table 3).However, the protective association attenuated in the balanced dataset.
In the sensitivity analysis that accounted for healthy vaccinee bias, 6012 individuals were excluded.In this analysis, we also found odds of involuntary movement to be lower in all vaccinated exposure groups compared with those not vaccinated (Table 4).

Discussion
The results indicate that the COVID-19 vaccine is not associated with an increased risk of involuntary movements.In fact, those who received the vaccine had lower odds of reporting involuntary movements.
The study benefits from its population-based design and the significant size of the BiCoVac cohort.Also, information on several relevant confounders, including vaccine hesitancy was included.Additionally, the timing of data collection, which coincided with the rollout of the COVID-19 vaccine schedule in the general Danish population, is another notable strength.Nevertheless, the relatively large attrition rate is a limitation.
The nonresponse at baseline reduces the size of the study but is unlikely to have biased the results, since participation at baseline most likely was unrelated to future experience of involuntary movement after having adjusted for involuntary movements at baseline.It can, however, not be ruled out that the loss to the first and second follow-up may have led to selection bias.It is possible that experiencing involuntary movements following vaccination could encourage further participation among vaccinated individuals, yet this does not explain the observed protective effect of the vaccine.Conversely, severe involuntary movements might have led to study dropouts due to physical difficulties in completing the questionnaires.While any potential selection bias cannot be entirely dismissed, it does not seem particularly plausible, and as such, its impact on the results is expected to be minimal.
The quality of the retrospective reporting of involuntary movements is challenging to confirm.However, given that participants were asked about the symptoms in a relatively short time span, it is anticipated that they should be able to recall these symptoms.People who supported the vaccine and chose to get vaccinated may have potentially aligned their responses to favor the vaccine's safety profile.On the other hand, individuals who were skeptical but still opted for vaccination might have been more attentive and prone to overreport symptoms.The resulting bias direction is challenging to ascertain but would, to some degree, cancel out each other.
While we may not fully comprehend what respondents considered as involuntary movements during self-reporting, it seems reasonable to assume that the self-reports indicate pathological involuntary movements rather than non-pathological ones.This assumption is grounded in the understanding that non-pathological movements occur automatically without conscious awareness and are thus less likely to be taken into consideration when self-reporting.Hence, our investigation yields insights into the relationship between the COVID-19 vaccine and abnormal involuntary movements.
The existing literature on the relationship between the COVID-19 vaccine and involuntary movements is largely constrained.It is mainly composed of case studies detailing instances of various hyperkinetic movement disorders occurring following vaccination [8][9][10][11][12][13][14][15][16][17][18], raising concerns about a possible relationship between COVID-19 vaccination and involuntary movements.In contrast to these concerns, this study's findings challenge the notion of such an association, indicating a protective association with the vaccine instead.There is currently no biological explanation for this.Instead, it could be ascribed to potential confounding factors that were not thoroughly addressed in our analyses.
Several relevant confounders were included in the analyses, however residual confounding cannot be ruled out.Furthermore, adjustment for prior COVID-19 infection did not take self-testing into account.It is anticipated though that most individuals receiving a positive self-test result would subsequently undergo a PCR test, aligning with the guidance from national health authorities in the study period [23].
Exercising caution when extending the study's findings to other countries is advisable, given that variations in vaccine types may exist depending on the specific vaccination program of a country.This study did not investigate whether the impact on involuntary movements was influenced by the specific vaccine type.However, it is reasonable to generalize the results of this study to individuals aged 16-65 in countries primarily using Pfizer/BioNTech and Moderna vaccines.

Conclusion
The study results show no evidence of involuntary movements as an adverse reaction to the COVID-19 vaccine.Overall, the study supports the safety profile of the COVID-19 vaccine concerning involuntary movements and contributes to building trust in vaccination programs among the population.
C.M. Nielsen et al. 889,503 invited individuals, 1271 were missing vaccination data, and 17,422 had incomplete responses, primarily consisting of individuals initiating the baseline questionnaire without completing it.The incomplete responses at baseline also involved 25 individuals missing covariate information due to technical difficulties encountered during the administration of such a large-scale digital questionnaire.Of the 234,882 individuals successfully completing the baseline questionnaire, 76,406 did not meet the inclusion criteria of not having received any

Table 1
Characteristics of the study population.
Note: The participants' appearance in the study population is equivalent to the number of completed follow-up questionnaires, allowing for a maximum of two appearances.The study population consists of 165,834 responses from a total of 97,537 unique individuals.C.M.Nielsen et al.

Table 2
Odds ratios of involuntary movements by COVID-19 vaccination status.
* Adjusted for age, sex, education, prior diagnosis involving symptoms of involuntary movement, prior redemption of prescribed psychotropic medication, stress, vaccine hesitancy, physical activity, smoking, prior COVID-19 infection, and baseline involuntary movement symptom score.

Table 3
Odds ratios of involuntary movements after *balancing the data.Only participants who completed all three questionnaires were included.** Adjusted for age, sex, education, prior diagnosis involving symptoms of involuntary movement, prior redemption of prescribed psychotropic medication, stress, vaccine hesitancy, physical activity, smoking, prior COVID-19 infection, and baseline involuntary movement symptom score. *

Table 4
Odds ratios of involuntary movements after exclusion of *non-vaccine recipients.