This study involved the follow-up of 42 previously hospitalised COVID-19 patients who had received varying doses of vaccination at different intervals following recovery. However, by the end of the one-year follow-up period, only 24 participants completed all scheduled visits due to exclusion and attrition factors. Attrition is a common challenge in longitudinal studies, particularly those requiring repeated blood sampling or extended commitments from participants. Loss to follow-up can introduce selection bias and reduce statistical power, potentially leading to underestimation of prevalence or weakening of observed associations.²⁰ Nevertheless, our findings showed an increasing trend in antibody titre post-vaccination, which is consistent with the one reported by Haq et al.²¹ Importantly, well-conducted longitudinal studies with smaller sample sizes can still yield valid and meaningful insights, particularly when data collection is rigorous and participants are well-characterised.²² Despite the reduced cohort size, our data started to show an increase in antibody trends, reflecting the long-term dynamics of IgG and neutralising antibody responses after vaccination and natural infection.

Although our finding for male gender was not significant, the higher proportion of male participants aligns with earlier reports suggesting that men are more likely to experience severe or critical COVID-19 outcomes.^23,24 This trend may be attributed to a combination of biological, psychosocial, and behavioural factors. Biologically, men exhibit higher expression of ACE2 receptors and possess relatively weaker immune responses than women, possibly due to the immunomodulatory advantage conferred by the second X chromosome in females. Behavioural factors such as greater occupational exposure, higher tobacco and alcohol use, and reduced healthcare-seeking behaviour may further contribute to increased severity among males.^25,26

Our findings showed a significant correlation between BMI and the severity of disease, indicated by the category of infection (P = 0.006). Higher BMI was reported to increase the risk of severe COVID-19 outcomes, which include intensive care unit (ICU) admission, the requirement of mechanical ventilation and death.^27,28 The severe outcomes observed in this group may be partly explained by the overexpression of ACE2 receptors in individuals with obesity, enabling increased infectivity by the virus.²⁹ The elevated rate of viral entry then triggers the overstimulation of inflammatory mediators, thereby inducing chronic low-grade inflammation and possible manifestation of acute respiratory distress syndrome (ARDS).^30,31 Despite this, obesity is not directly correlated with the antibody titre, with several studies suggesting that this condition negatively impacts the persistence of vaccine-induced immunity.^32,33

Reinfections, confirmed through antigen-based rapid test kits (Ag-RTK), were observed in four participants during the study period. Although reinfection can serve as a natural booster, enhancing antibody production, this effect was not consistent across our cohort. Patients reinfected with the Delta variant showed variable antibody responses, whereas those who were infected with the Omicron variant have been associated with higher reinfection rates but possibly lower immunogenicity.^34-36 As in other studies, the reinfections observed in this study did not result in severe outcomes, highlighting the potent protection afforded by hybrid immunity.³⁷ It was assumed that these reinfections occurred possibly due to the Omicron variant, due to its domination during the timing of reinfection, although there was no evidence to support this.

Our findings demonstrated the persistence of anti-SARS-CoV-2 IgG antibodies up to 390 days post-infection and post-vaccination, reaffirming the relevance of serological assays in monitoring convalescent-phase immunity. Long-lasting seropositivity has been previously reported for other coronaviruses, including MERS-CoV, with IgG detectable for up to four years.³⁸ Among participants who received two vaccine doses, the data analysis demonstrated a negative correlation between antibody titre and time since vaccination, indicating a gradual decline in IgG levels over time. This observation is consistent with global studies reporting waning antibody levels following the primary series of vaccination.^21,39,40 Conversely, participants who received a third (booster) dose exhibited more stable and higher antibody titres over time. This trend supports findings from studies in Japan, which demonstrated prolonged antibody persistence following booster vaccination.^13,41 It is plausible that some participants sampled early post-booster had not yet reached peak antibody levels, while those sampled later were already in the plateau phase of antibody persistence.

The severity of the initial infection also appeared to influence antibody levels. Participants with Category 5 infection, representing the most critical cases, recorded a higher mean antibody titre compared to those with Category 4. This finding is consistent with prior studies showing that patients with severe disease tend to develop stronger and longer-lasting antibody responses, possibly due to higher viral loads and more pronounced immune activation during acute illness.^42-44 Moreover, individual variability in immune fitness may explain differing levels of antibody production among patients with similar disease categories.^45,46

Neutralising antibodies (NAbs), which play a critical role in blocking viral entry into host cells, were assessed in this study using a surrogate virus neutralisation test (sVNT). The sVNT-based assay serves as a rapid and effective tool for anti-SARS-CoV-2 NAbs detection without the requirement of high-level biosafety facilities for conventional cell-based neutralisation tests.^16,47 The high specificity and sensitivity of the platform in correlation with the traditional assay augment its robustness to be implemented in assessing the NAbs presence, which reflects the protective level following hybrid immunity.^48,49 These antibodies were detectable in nearly all participants, except for one, who recorded the lowest IgG titre at the final visit (79.60 BAU/mL). Factors which may affect the production of this antibody include prolonged interval, coupled with the use of an inactivated virus vaccine, and co-morbidities which are known to impair immune responses.^50-52

The type of vaccine administered also influenced antibody responses. Heterologous COVID-19 vaccination regimens, such as combining an inactivated vaccine like CoronaVac with an mRNA vaccine booster (e.g., BNT162b2), have been shown to elicit a stronger and broader immune response than homologous regimens.^53-56 In our study, however, nearly all of the 24 participants received homologous vaccinations consisting entirely of CoronaVac, with only one participant receiving a heterologous regimen (CoronaVac followed by BNT162b2). Despite this, we observed a consistent trend of increasing IgG antibody titres following each booster dose over a one-year follow-up period, highlighting that even homologous regimens may provide sustained immune protection, particularly when multiple booster doses are administered. When comparing types of vaccinations, the inactivated vaccine was reported to induce a less robust humoral immune response than the other strategies due to the non-replicative nature of the attenuated pathogens, which limits the magnitude and duration.^57,58 In addition, the presentation of a wide array of antigens on the attenuated viral surface triggers a broad-spectrum antibody response or polyclonal antibody production, not mainly targeting the spike. This reduces the production of neutralising anti-spike antibodies, which is being assessed as the indication of protection and neutralising capacity in this study.^59,60 With the low number of participants, this study is unable to assess the correlation between the type of vaccination and the dynamics and longevity of the anti-SARS-CoV-2 antibody titres.

Although 266.7 ± 30.9 BAU/mL was the highest titre measured in our cohort, this value is still lower than antibody levels typically reported or expected elsewhere. For instance, Lau et al. and Nanda et al. reported median titres exceeding 1,000 BAU/mL and 285 BAU/mL following mRNA-based and viral vector platforms, respectively.^60,61 In another study, SinoVac or inactivated virus-based vaccine recipients recorded a median IgG titre of 214.⁵⁰ BAU/mL at week 16 following vaccinations.⁶² Despite this relatively low mean antibody titre in our cohort, nearly all participants except one demonstrated the presence of neutralising antibodies (NAbs), suggesting functional immune protection. This finding underscores that IgG titres alone, particularly those targeting only the RBD, may not fully reflect the protective immunity against SARS-CoV-2. Previous research has shown that neutralising and protective antibodies can also target other spike protein domains, such as the N-terminal domain (NTD), and even non-spike viral antigens.^7,63,64 Hence, the inclusion of additional antigenic targets in antibody assessments may provide a more accurate picture of immune protection, beyond the measured RBD-specific IgG levels.

This study had several limitations. First, we lacked a control group comprising individuals with purely vaccine-induced or natural infection-induced immunity, which limited our ability to isolate the effects of each exposure type. Second, the lack of comprehensive data on the exact vaccine types and dates for each participant hindered detailed vaccine-specific analysis. Third, the interval between vaccination and antibody measurement varied considerably among participants, which may have influenced the observed antibody titres, as antibody levels decline over time following COVID-19 vaccination.^61,62

Next, our small sample size and focus on patients attending follow-up at a respiratory clinic reduce the generalisability of the findings to the broader population. Additionally, the variable timing of blood sample collection post-vaccination or infection may have introduced heterogeneity in observed antibody levels. While this study focused on assessing the longitudinal antibody dynamics following hybrid immunity, the small number of participants receiving heterologous vaccinations limits our ability to draw accurate conclusions on the effectiveness of heterologous vaccinations, despite supportive clinical evidence in various studies.

Nonetheless, this study provides valuable insights into the durability and variability of humoral immune responses following hybrid immunity in a real-world clinical cohort. The hybrid immunity, conferred through both infection and vaccination, appeared to result in more sustained IgG and NAbs responses. However, responses varied depending on age, sex, comorbidities, timing of vaccination, and vaccine type.

Future research should include larger and more diverse cohorts to better understand the long-term durability and breadth of hybrid immunity, especially against emerging SARS-CoV-2 variants. Investigating the therapeutic potential of hybrid-induced antibodies and their role in post-COVID-19 recovery syndromes would further enhance our understanding of protective immunity. Such findings may inform future vaccination policies, optimise the timing and composition of booster programmes, and contribute to more effective pandemic preparedness and response strategies.