Professional Documents
Culture Documents
Durability and Determinants of Anti SARS CoV 2 Spi
Durability and Determinants of Anti SARS CoV 2 Spi
Durability and Determinants of Anti SARS CoV 2 Spi
e5
Research Note
a r t i c l e i n f o a b s t r a c t
Article history: Objectives: To examine the differences in durability and its determinants of humoral immunity following
Received 16 November 2022 2- and 3-dose COVID-19 vaccination.
Received in revised form Methods: Throughout the pandemic, we evaluated the anti-spike IgG antibody titers of 2- and 3-dose
10 April 2023
mRNA vaccine recipients over time among the staff of a medical and research center in Tokyo. Linear
Accepted 16 May 2023
Available online 24 May 2023
mixed models were used to estimate trajectories of antibody titers from 14 to 180 days after the last
immune-conferred event (vaccination or infection) and compare antibody waning rates across prior
Handling Editor: M. Cevik infection and vaccination status, and across background factors in infection-naïve participants.
Results: A total of 6901 measurements from 2964 participants (median age, 35 years; 30% male) were
Keywords: analyzed. Antibody waning rate (percentage per 30 days [95% CI]) was slower after 3 doses (25% [23e26])
Durability than 2 doses (36% [35e37]). Participants with hybrid immunity (vaccination and infection) had further
Prior infection slower waning rates: 2-dose plus infection (16% [9e22]); 3-dose plus infection (21% [17e25]). Older age,
SARS-CoV-2 male sex, obesity, coexisting diseases, immunosuppressant use, smoking, and alcohol drinking were
Spike antibody
associated with lower antibody titers, whereas these associations disappeared after 3 doses, except for
Vaccination
sex (lower in female participants) and immunosuppressant use. Antibody waned slightly faster in older
participants, females, and alcohol drinkers after 2 doses, whereas it did not differ after 3 doses across
except sex.
Discussion: The 3-dose mRNA vaccine conferred higher durable antibody titers, and previous infection
modestly enhanced its durability. The antibody levels at a given time point and waning speed after 2
doses differed across background factors; however, these differences mostly diminished after 3 doses.
Shohei Yamamoto, Clin Microbiol Infect 2023;29:1201.e1e1201.e5
© 2023 The Author(s). Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology
and Infectious Diseases. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction
https://doi.org/10.1016/j.cmi.2023.05.020
1198-743X/© 2023 The Author(s). Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1201.e2 S. Yamamoto et al. / Clinical Microbiology and Infection 29 (2023) 1201.e1e1201.e5
Shown are the estimated GMT of anti-SARS-CoV-2 spike protein antibody titers on 14, 30, 90, and 180 d since the last immune-conferring event (vaccination or infection) and the estimated RoM of antibody waning slopes, with
adjustment for age (<40 or 40 y), sex (male or female), body mass index (<27.5 or 27.5 kg/m2), coexisting diseases (yes or no), immunosuppressant use (yes or no), smoking status (current or non-smoker), and frequency of
1.13 (1.05e1.21)
1.05 (1.01e1.11)
including the recommendation regarding the timing and target
reference population of an additional booster dose.
Epidemiological data are scarce regarding the waning pattern of
humoral immunity following 2- and 3-dose COVID-19 vaccination
and its related factors. Studies among 2-dose recipients showed
Comparison of slopes
1.17 (1.16e1.19)
1.32 (1.23e1.43)
1.24 (1.17e1.31)
0.64
0.75
0.84
0.79
Methods
the National Center for Global Health and Medicine, Japan [6,7]. We
4699 (4540e4864)a
50,000
Estimated spike antibody titers (AU/mL)
(33 590e41 961)a,b
(19 450e21 146)a
(19 323e33 388)a
20,000
(9893e10 690)
5,000
2,000
Day 30
284
280
400
543
10
20
25
37
1,000
500
032)a,b
783) a
942)a
199)
0 60 120 180
GMT, geometric mean titer; RoM, ratio of means.
426
389
904
131
Fig. 1. Waning of anti-SARS-CoV-2 spike protein antibody titers from the last exposure
to SARS-CoV-2 (infection or vaccination).
14
24
25
35
Curves shown are the estimated trajectories adjusted for age (<40 or 40 years), sex
(male or female), body mass index (<27.5 or 27.5 kg/m2), coexisting diseases (yes or
no), immunosuppressant use (yes or no), smoking status (current or non-smoker), and
2-dose þ infection
3-dose þ infection
frequency of alcohol drinking (<1 or 1 time/wk). A blue line shows participants who
received 2-dose with infection-naïve (N ¼ 2994), and a red line shows those who
received 2-dose with a history of SARS-CoV-2 infection (N ¼ 53). A green line shows
those who received 3-dose with infection-naïve (N ¼ 3582), and an orange line shows
2-dose
3-dose
those who received 3-dose with a history of the infection (N ¼ 272). Solid lines are the
Table 1
Table 2
Characteristics of antibody waning after the second and third vaccination across the background factors among infection-naïve participants
Characteristics Estimated spike antibody titers (AU/mL) after 2-dose, GMT (95% CI) Slope of antibody waning, RoM (95% CI)
Age (y)
<40 16 076 (15 103e17 113) 11 618 (11 074e12 189) 3967 (3803e4137) 1210 (1140e1285) 0.64 (0.63e0.65) reference
40 13 405 (12 187e14 744)a 9137 (8538e9778)a 2651 (2502e2809)a 749 (677e829)a 0.61 (0.60e0.63) 0.96 (0.93e0.99)a
Sex
Male 11 947 (10 886e13 112) 8685 (8098e9314) 3041 (2865e3228) 973 (885e1070) 0.65 (0.64e0.67) reference
Female 15 632 (14 706e16 617)a 11 078 (10 588e11 591)a 3585 (3446e3729)a 1070 (1008e1135) 0.63 (0.62e0.64) 0.97 (0.94e0.99)a
Obesity
<27.5 (kg/m2) 14 508 (13 765e15 291) 10 376 (9977e10 791) 3450 (3335e3569) 1050 (997e1106) 0.64 (0.63e0.65) reference
27.5 (kg/m2) 15 147 (11 122e20 627) 9956 (8081e12 266) 2690 (2257e3206)a 827 (622e1101) 0.62 (0.57e0.67) 0.97 (0.89e1.05)
Coexisting diseases
No 14 786 (14 077e11 086) 10 583 (10 167e11 017) 3525 (3404e3650) 1071 (1016e1129) 0.64 (0.63e0.65) reference
Yes 14 077 (11 086e17 875) 9350 (7977e109 59) 2572 (2259e2928)a 764 (621e941)a 0.62 (0.58e0.65) 0.96 (0.91e1.02)
Immunosuppressant use
No 14 564 (13 825e15 343) 10 388 (9993e10,798) 3432 (3319e3549) 1046 (993e1101) 0.64 (0.63e0.65) reference
Yes 8074 (5221e12 486)a 6315 (4588e8694)a 2681 (2108e3409)a 897 (595e1352) 0.68 (0.61e0.76) 1.06 (0.95e1.20)
Smoking
Non-smokers 14 663 (13 881e15 490) 10 463 (10 048e10 894) 3459 (3340e3582) 1052 (998e1110) 0.64 (0.63e0.65) reference
Current smokers 11 911 (10 084e14 069)a 8606 (7563e9793)a 2957 (2642e3309)a 930 (780e1108) 0.65 (0.62e0.68) 1.01 (0.97e1.06)
Alcohol drinking
Non or occasional 14 818 (13 926e15 768) 10 712 (10 220e11 228) 3671 (3522e3826) 1134 (1065e1207) 0.64 (0.63e0.65) reference
Weekly or daily 14 269 (12 962e15 706) 9868 (9212e10 572) 2998 (2830e3176)a 884 (807e968)a 0.62 (0.61e0.64) 0.97 (0.94e0.99)a
Characteristics Estimated spike antibody titers (AU/mL) after 3-dose, GMT (95% CI) Slope of antibody waning, RoM (95% CI)
Day 14 Day 30 Day 90 Day 180 Per 30 days Comparison of slopes
Age (y)
<40 25 788 (24 167e27 517) 21 110 (20 077e22 196) 10 696 (10 242e11 171) 4756 (4543e4980) 0.75 (0.74e0.75) reference
40 23 286 (20 993e25 829) 19 566 (18 132e21 114) 10 517 (9992e11 069) 4552 (4323e4794) 0.75 (0.74e0.76) 1.00 (0.99e1.02)
Sex
Male 25 563 (22 422e29 144) 22 274 (20 255e24 494) 13 168 (12 360e14 029) 5827 (5472e6204) 0.76 (0.75e0.78) reference
Female 24 200 (22 803e25 683) 19 770 (18 889e20 692)a 9926 (9558e10 309)a 4336 (4166e4513)a 0.74 (0.73e0.75) 0.97 (0.95e0.99)a
Obesity (kg/m2)
<27.5 24 455 (23 117e25 870) 20 284 (19 442e21 161) 10 601 (10 251e10 963) 4678 (4517e4845) 0.75 (0.74e0.76) reference
27.5 27 046 (20 590e35 526) 23 099 (18 859e28 291) 12 810 (11 073e14 820)a 5323 (4630e6120) 0.75 (0.72e0.78) 1.00 (0.96e1.04)
Coexisting diseases
No 24 632 (23 250e26 096) 20 342 (19 472e21 250) 10 528 (10 168e10 901) 4668 (4503e4839) 0.75 (0.74e0.75) reference
Yes 24 091 (20 053e28 942) 20 814 (18 166e23 850) 11 966 (10 833e13 216) 5133 (4619e5704) 0.76 (0.74e0.78) 1.01 (0.98e1.04)
Immunosuppressant use
No 24 737 (23 391e26 160) 20 554 (19 708e21 437) 10 788 (10 436e11 151) 4757 (4596e4923) 0.75 (0.74e0.76) reference
Yes 17 168 (12 090e24 379)a 14 067 (10 717e18 465)a 7065 (5580e8945)a 2987 (2331e3828)a 0.74 (0.69e0.78) 0.98 (0.93e1.04)
Smoking
Non-smokers 24 621 (23 232e26 092) 20 479 (19 606e21 392) 10 777 (10 416e11 150) 4750 (4584e4922) 0.75 (0.74e0.76) Reference
Current smokers 23 542 (19 672e28 175) 19 415 (16 893e22 313) 9959 (8824e11 239) 4309 (3829e4848) 0.74 (0.72e0.77) 0.99 (0.96e1.02)
Alcohol drinking
Non or occasional 25 174 (23 477e26 994) 20 807 (19 741e21 931) 10 814 (10 372e11 275) 4839 (4632e5056) 0.75 (0.74e0.76) Reference
Weekly or daily 23 513 (21 483e25 735) 19 708 (18 411e21 097) 10 509 (9961e11 087) 4513 (4272e4769) 0.75 (0.74e0.76) 1.00 (0.98e1.01)
Shown are the estimated GMT of anti-SARS-CoV-2 spike protein antibody titers on 14, 30, 90, and 180 d since the third mRNA vaccination and the estimated RoM of antibody
waning slopes, with adjustment for age (<40 or 40 y), sex (male or female), body mass index (<27.5 or 27.5 kg/m2), coexisting diseases (yes or no), immunosuppressant use
(yes or no), smoking status (current or non-smoker), and frequency of alcohol drinking (<1 or 1 time/wk).
GMT, geometric mean titer; RoM, ratio of means.
a
p <0.05.
to 180 days after the last immune-conferred event (vaccination or Among the infection-naïve participants, the waning rate of
infection) and compare antibody waning rates across prior infec- antibody titers was 17% slower per 30 days after the 3 doses (25%
tion and vaccination status, and across background factors in per 30 days) compared with 2 doses (36% per 30 days) (Table 1 and
infection-naïve participants. Written informed consent was ob- Fig. 1). The 2- and 3-dose recipients with a history of infection had
tained from all participants, and the study procedure was approved slower antibody waning rates than their infection-naïve counter-
by the NCGM ethics committee (approved number: NCGM-G- parts. Among the 2-dose recipients, those with a history of infec-
003598). Details of the study setting, participants, variables infor- tion had consistently higher antibody titers during days 14e180
mation, and statistical approach are described in Text S1. and a 32% slower waning rate than those without infection. Among
the 3-dose recipients, those with a history of infection had higher
antibody titers during days 14e180 and a 5% slower decline rate
Results than those without infection. The 2-dose recipients with infection
had higher antibody titers and a 13% slower waning rate than 3-
Among the 2964 participants, the median age was 35 (inter- dose recipients without infection.
quartile range: 27e47) and 30.2% were male (Tables S1 and S2). Table 2 and Fig. 2 show the patterns of antibody waning
Those with obesity, coexisting diseases, and immunosuppression following 2- and 3-dose vaccines, respectively, according to back-
were 5.5%, 11.1%, and 1.8%, respectively, and tobacco product users ground factors among infection-naïve participants. Older age, male
and regular drinkers were 7.7% and 37.6%, respectively. sex, obesity, coexisting diseases, immunosuppressant use, smoking,
1201.e4 S. Yamamoto et al. / Clinical Microbiology and Infection 29 (2023) 1201.e1e1201.e5
A 50,000
Estimated spike antibody titers (AU/mL) B 50,000
5,000 5,000
2,000 2,000
1,000 1,000
500 500
0 60 120 180 0 60 120 180
Days since the vaccination Days since the vaccination
2-dose, Age <40 2-dose, Age ≥40 2-dose, Male 2-dose, Female
3-dose, Age <40 3-dose, Age ≥40 3-dose, Male 3-dose, Female
C 50,000
Estimated spike antibody titers (AU/mL)
20,000
5,000
2,000
1,000
500
0 60 120 180
Days since the vaccination
Fig. 2. Trajectories of anti-SARS-CoV-2 spike protein antibody titers following 2- and 3-dose vaccination by age (A), sex (B), and alcohol drinking status (C) among infection-naïve
participants. Curves shown are the estimated trajectories for individuals who had received 2-dose (blue and red lines) and those who had received 3-dose (green and orange lines)
adjusted for age (<40 or 40 years), sex (male or female), body mass index (<27.5 or 27.5 kg/m2), coexisting diseases (yes or no), immunosuppressant use (yes or no), smoking
status (current or non-smoker), and frequency of alcohol drinking (<1 or 1 time/wk). Solid lines are the estimated geometric mean of anti-SARS-CoV-2 spike protein antibody
titers, and shaded areas are the corresponding 95% CIs.
and alcohol drinking were each associated with lower antibody with vaccination conferred more durable immunity than vaccine
titers after 2 doses, whereas these associations disappeared after 3 alone, which is in line with the results of meta-analysis [13,14].
doses, except for sex (females became lower) and immunosup- Among the infection-naïve participants, we observed that older
pressant use (still lower). Antibody waning rates were slightly but age, male sex, obesity, coexisting diseases, immunosuppressant
significantly faster in those who are older (4%), female (3%), and use use, smoking, and regular alcohol use were each associated with
alcohol (3%) after 2 doses, and somewhat faster in females (3%) lower antibody titers throughout 14e180 days following the second
after 3 doses. dose. These results were consistent with previous cross-sectional
studies [6e10]. After the 3-dose vaccine, these associations dis-
appeared except for sex and immunosuppressant use, suggesting
Discussion that a booster shot can enhance immunity and minimize the dif-
ference between groups. Still, those with immunosuppressive sta-
Among recipients of COVID-19 mRNA vaccines, we found that tus had lower antibody titers after 3 doses, a finding compatible
the durability of antibody titers was higher after 3 doses than 2 with the VE study among 2- and 3-dose vaccine recipients [2]. Our
doses, consistent with a previous study of Israeli healthcare results highlight the need for careful monitoring of infection risk
workers [3] and compatible with previous studies showing a and consideration of additional boosters among individuals
slower waning of VE for infection over time in recipients of 3-dose receiving immunosuppressants.
vaccine relative to those who received 2-dose vaccine [12,13]. In Regarding antibody waning, we observed a slightly faster
addition, we found that hybrid immunity (vaccination and infec- decline (3e4%) associated with older age, female sex, and regular
tion) has added durability over the immunity induced by vaccina- alcohol use after 2 doses. After 3 doses, the differences by age and
tion alone. The novelty of this study is that 3-dose plus infection has alcohol used disappeared, whereas the female sex remained a
led to more durable antibodies than 3-dose alone, albeit the dif- slight but significant driver of antibody decline (3%). This result
ference was modest, which is compatible with a study reporting a suggests that a booster dose plays a role in eliminating gaps in
higher VE against infection for 3-dose plus prior infection versus 3- antibody durability by background factors. We have no plausible
dose alone [13]. Interestingly, we observed that those with 2-dose explanation for the faster waning in females after both 2 and 3
plus infection had more durable antibody titers than those with doses. Interestingly, a meta-analysis of clinical trials on primary
3-dose alone. This result suggests that natural infection combined COVID-19 vaccines showed that females have a significantly lower
S. Yamamoto et al. / Clinical Microbiology and Infection 29 (2023) 1201.e1e1201.e5 1201.e5