Among the > 6,000 young adults initially assessed in our study, only 3% had received two doses of the varicella vaccine prior to study enrollment. Since 2003, the National Immunization Plan has supported varicella vaccination for adolescents who are not naturally immunized. Thus, for the population born before 1998, as was the case in the majority of our study participants, varicella coverage in Italy is very low.
Despite two doses of varicella vaccine, almost 34% (95%CI = 27–41) of the enrollees did not have circulating anti-VZV IgG. However, among those who then received a booster dose, an immunity response was elicited such that the seroconversion rate was 100%. Furthermore, the increase in the GMT (from 67 to 1,608) after the booster dose resulted in a significantly higher titer than in the group that was seroprotected at enrollment (1,608 vs. 657; p < 0.0001), indicating a decrease over time in the antibody titer among the vaccinated population.
Our study showed a better immunological response and duration of circulating antibodies in females than in males, based on comparisons of seroprevalence at enrollment, GMT, and PAS, the IRR, and the multivariate regression models. Previous studies of the sex-based differences in the response to vaccines or infection [22, 23] have consistently shown that females have a more effective immune response to immunization, and thus presumably varicella vaccination (immunogenicity and, probably, effectiveness), and against infection.
Our antibody survival model showed that antibody levels tend to decline as early as 1 year after completion of the basal routine, such that after 9 years half of the fully vaccinated population will have lost circulating antibodies. In another study, albeit of low scientific evidence, a period of 10 years was determined . For other live-attenuated virus vaccines, such as the MMR (measles, mumps, rubella) vaccine, the duration of circulating antibodies is more than double that of the varicella vaccine [21, 24, 25]. A loss of circulating antibodies over time was also shown in this study by the multivariate regression models, in which the interval between the last vaccine dose and the serological assessment was identified as a determinant of serosensitivity.
The strengths of this study are that it provides further knowledge on a topic that has been poorly studied thus far, including evidence of sex-based difference in VZV immunity. However, a major limitation was the sample size, although this was not unexpected because, in Italy, varicella vaccination for adolescents who are not naturally immunized became routine only in 2003 whereas the study participants were for the most part born before 1998. Another limitation was our inability to analyze the immune status as a function of vaccine formulation (monovalent vs. MMRV). A determination of the differences between these vaccines is crucial to understanding their immunogenicity performance. In addition, whether the study enrollees had ever come into contact with the wild virus (without developing disease) was unknown. Future research should assess the management of non-responders, which will require a larger sample size and a longer follow-up after routine vaccination, perhaps also stratifying the study population per vaccine type (monovalent vs. MMRV). This approach will provide a fuller picture of immunogenicity over time.
The efficacy of at least one (and potentially a second) booster dose in fully vaccinated but non-seroprotected healthcare workers (HCWs), such as our study population, has not been well studied. Behrman et al.  evaluated the humoral immune responses of 101 HCWs (median age 30 years) who had previously received a two-dose regimen of varicella vaccine. including 12 (11.9%) who were seronegative at enrollment but received a third vaccine dose during the study. Seven of those 12 HCWs (58.3%) seroconverted after the third dose (safety data were not reported). Our study population comprised a much larger number of serosusceptibles and the seroconversion rate after the booster dose was nearly twice as high. In a 2020 Italian study, Trevisan A et al.  tested 234 full vaccinated medical school students and determined a serosusceptibility rate of 21.4%. The authors found a decrease in the antibody titer with an increasing interval from vaccination to the antibody titer evaluation. Within the group of serosusceptible students who received a third vaccine dose, 91.4% seroconverted, without any major symptoms following vaccination. The results of that study were in agreement with our own, especially considering the similar study populations, but Trevisan A et al.  found no evidence of a difference between sexes.
The risk of a loss of antibodies in vaccinated individuals over time is an important finding, as between now and the next 10–15 years the loss of circulating antibodies against VZV in the vaccinated population will increase their susceptibility to the virus. Since it is highly unlikely that VZV will be eliminated in the immediate future, the loss of immunity in a substantial portion of the population implies a risk of varicella outbreaks in the coming years. Indeed, many cases of chickenpox among fully vaccinated adults have been reported in the literature, although the disease in this group seems to be less severe [28–30]. However, it should also be noted that the role of cell-mediated immunity in the protection of the non-seroprotected is a matter of debate within the scientific community [26, 31], such that whether serum IgG titers alone accurately reflect vaccine protection is unclear. A 2020 study  compared humoral and T cell immunity in women of childbearing age who had been immunized against VZV either by vaccination or naturally. The authors found no significant differences in the antibody titers between the two groups (p > 0.050), although the naturally immunized group had significantly higher levels of VZV antigen-specific CD4 T cells (p = 0.004).