Advertisement

Impact of Routine Infant BCG Vaccination on COVID-19

Published:August 11, 2020DOI:https://doi.org/10.1016/j.jinf.2020.08.013

      Highlights

      • We investigated impact of BCG vaccination on prevention of local COVID-19 spread.
      • Prefectural data on SARS-CoV-2 prevalence and BCG vaccine coverage were analyzed.
      • The BCG vaccine coverage was higher in uninfected than in prevalent prefectures.
      • Prevalence of the infection significantly correlated to the BCG vaccine coverage.
      • The infant BCG vaccine plays a significant protective role in local COVID-19 spread.

      Summary

      Objectives

      In Japan, the first case of coronavirus disease 2019 (COVID-19) was diagnosed on January 15, 2020 and subsequent infections rapidly increased. The Bacillus Calmette-Guérin (BCG) vaccination program is the principal element of tuberculosis control in Japan. We investigated the impact of routine infant BCG vaccination on prevention of local COVID-19 spread.

      Methods

      Data on the prevalence of SARS-CoV-2 infection, annual routine infant BCG vaccine coverage (represented by the number of BCG vaccinations per live births), and other candidate factors in each prefecture were obtained from the official notifications database in Japan. We analysed the association of vaccine coverage with the prevalence of SARS-CoV-2 infection.

      Results

      The BCG vaccine coverage in 1999–2002, 2004, and 2012 in five prefectures with no COVID-19 infections was significantly higher than that in five prefectures with a high prevalence of infections (Mann-Whitney U test, p<0.05). The prevalence of SARS-CoV-2 infection was significantly negatively correlated with BCG vaccine coverage in 2004 and was significantly positively correlated with age groups 20–34 and 40–54 years (Spearman's rank correlation, p<0.01).

      Conclusions

      Our findings suggest that routine infant BCG vaccination coverage in young generation had a significant impact on prevention of local COVID-19 spread in Japan.

      Key words

      Introduction

      In Japan, the first case of coronavirus disease 2019 (COVID-19) was diagnosed in Kanagawa Prefecture on January 15, 2020.

      Ministry of Health, Labour and Welfare of Japan. About Coronavirus Disease 2019 (COVID-19). 2020. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000164708_00001.html. Accessed 28 May 2020.

      Because of an explosive increase in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections since late March, the Japanese Prime Minister declared a state of emergency for seven prefectures on April 7 and for the whole country on April 16. Japan lifted its nationwide state of emergency on May 25, 2020. The cumulative number of COVID-19 cases in Japan had topped 16,623 (13.1 per 100,000 population) on that date; however, this figure was much lower than that in Western countries.
      Based on the possible effect of the Bacillus Calmette-Guérin (BCG) vaccination to prevent nonspecific infections other than tuberculosis, several clinical trials have been initiated with the aim to protect health care personnel and older people from COVID-19 infection.
      de Vrieze
      Can a century-old TB vaccine steel the immune system against the new coronavirus?.
      ,
      • Redelman-Sidi G
      Could BCG be used to protect against COVID-19.
      Japanese bacteriologist Kiyoshi Shiga introduced BCG to Japan in 1924. The 172nd passage from the first culture is the origin of the Japanese routine infant BCG vaccine (Tokyo-172).
      • Yamamoto S.
      • Yamamoto T.
      Historical review of BCG vaccine in Japan.
      The BCG vaccination program has been the principal element in tuberculosis control in Japan for more than 70 years.
      Here, we investigated the degree of impact of BCG vaccination on local prevalence of patients with COVID-19 infection, SARS-CoV-2 polymerase chain reaction-positive (PCR+) individuals, and deaths owing to COVID-19 in prefectures of Japan. Data on the number of BCG vaccinations, number of live births, and other candidate factors in each prefecture were obtained from the official notifications database of the Japanese government.

      Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

      Materials and methods

      Prevalence of patients with COVID-19, SARS-CoV-2 PCR+ individuals, and deaths in each prefecture

      Anonymous data on the number of patients with COVID-19 from January 15 to March 29, 2020, and thereafter the number of SARS-CoV-2 PCR+ individuals and the number of deaths owing to COVID-19 among Japanese residents in each prefecture were obtained from official notification records of the Ministry of Health, Labour and Welfare of Japan.

      Ministry of Health, Labour and Welfare of Japan. About Coronavirus Disease 2019 (COVID-19). 2020. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000164708_00001.html. Accessed 28 May 2020.

      Data on March 29, April 24, and May 1, 2020 were used. Passengers aboard the Diamond Princess cruise ship anchored in Yokohama were not included. Crude SARS-CoV-2 prevalence in each prefecture was calculated as the number of patients/PCR+ individuals divided by the number of individuals in each prefecture in 2018, estimated using the results of the national census in 2010 and in 2015 taken by the Statistics Bureau, Ministry of Internal Affairs and Communications of Japan.

      Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

      BCG vaccination

      Information on the annual number of routine infant BCG vaccinations in each prefecture from 1998 to 2017 was obtained from the Report on Regional Public Health Services and Health Promotion Services, and the annual number of live births in each prefecture in the same year was obtained from the latest version of the Vital Statistics; their ratio was used as a surrogate for BCG vaccine coverage. These surveys were conducted by the Vital, Health and Social Statistics Office, Ministry of Health, Labour and Welfare of Japan.

      Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

      Other factors

      We investigated other candidate factors related to the spread of SARS-CoV-2. The number of days from diagnosis of the first patient with COVID-19 infection in each prefecture up to March 29, 2020 was obtained from official notification records.

      Ministry of Health, Labour and Welfare of Japan. About Coronavirus Disease 2019 (COVID-19). 2020. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000164708_00001.html. Accessed 28 May 2020.

      The elderly population in 2018 and the population by 5-year age groups in 2019 in each prefecture was calculated based on official population estimates.

      Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

      The populations that migrated to and from each prefecture in 2018 was obtained from the annual report on internal migration in Japan derived from the basic resident registration; inhabitable area was calculated using data of the Statistics Bureau, Ministry of Internal Affairs and Communications.

      Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

      The ratio of day- to night-time population, ratio of households of 5 individuals or more to entire households, and ratio of workers in the primary, secondary and tertiary sectors of industry to entire working population were obtained from results of the national census in 2015.

      Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

      Statistical analyses

      The Kolmogorov–Smirnov test revealed that the variables did not have a normal distribution. Therefore we performed non-parametric analyses to compare annual BCG vaccine coverage among prefectures and to examine the correlation of SARS-CoV-2 prevalence with BCG vaccine coverage and other related factors. The significance level was set at p = 0.05 for group comparisons and p = 0.01 for correlation analyses. All statistical analyses were performed using SPSS 16.0 J (IBM Japan, Tokyo, Japan).

      Results

      Table 1 depicts the prevalence of patients with COVID-19, SARS-CoV-2 PCR+ individuals, and deaths owing to COVID-19 in each prefecture. The prevalence of COVID-19 varied widely among the 47 prefectures. On March 29, no infections were detected in five prefectures (Iwate, Yamagata, Toyama, Tottori, and Shimane), whereas 33.30 individuals/million were diagnosed in Hokkaido. On April 24 and May 1, only Iwate had no infections. PCR+ individuals were the most prevalent in Tokyo on May 1 (298.08 individuals/million). The mortality rate was high in Fukui, Gunma, and Ishikawa prefectures (9.04, 7.68, and 5.25 deaths/million, respectively). The case fatality rate was high in Gunma, Aichi, and Ehime prefectures (10.27, 6.38, and 6.38 deaths/PCR+ individuals (%)).
      Table 1Prevalence of patients with COVID-19, SARS-CoV-2 PCR+ individuals, and deaths in each prefecture.
      Prevalence (per million)Deaths/PCR+ (%)
      Patients with COVID-19 on Mar 29PCR+ on Apr 24Deaths on Apr 24PCR+ on May 1Deaths on May 1On Apr 24On May 1
      All Japan14.3999.571.95108.992.281.962.09
      Hokkaido33.30106.324.73137.344.924.453.58
      Aomori5.5417.42020.59000
      Iwate00000
      Miyagi0.8636.27038.00000
      Akita4.0816.31016.31000
      Yamagata060.55062.39000
      Fukushima1.0735.41038.63000
      Ibaraki5.5654.572.0956.312.093.823.70
      Tochigi6.1727.24027.75000
      Gunma9.2271.724.6174.807.686.4310.27
      Saitama11.32104.642.05115.012.051.961.78
      Chiba25.58122.782.88128.863.682.342.85
      Tokyo31.54271.091.37298.081.370.510.46
      Kanagawa11.6699.382.62109.193.382.633.09
      Niigata13.8028.94033.39000
      Toyama0155.242.86180.952.861.841.58
      Ishikawa7.87194.234.37218.725.252.252.40
      Fukui16.80155.046.46157.629.044.175.74
      Yamanashi4.9062.42063.65000
      Nagano3.8832.48031.99000
      Gifu10.5273.113.5173.113.514.794.79
      Shizuoka1.0916.12018.58000
      Aichi22.1662.893.9864.484.116.336.38
      Mie5.0325.130.5625.130.562.222.22
      Shiga4.2565.860.7167.990.711.081.04
      Kyoto17.75110.382.70121.963.472.452.85
      Osaka23.71164.302.84181.444.201.732.31
      Hyogo24.07110.323.10117.073.102.812.65
      Nara8.2256.760.7561.240.751.321.22
      Wakayama18.1855.611.0764.172.141.923.33
      Tottori05.3605.36000
      Shimane023.53033.82000
      Okayama1.5810.54012.12000
      Hiroshima2.1350.410.3554.310.350.700.65
      Yamaguchi4.3822.63023.36000
      Tokushima1.366.7906.79000
      Kagawa1.0429.11029.11000
      Ehime2.9634.762.2234.762.226.386.38
      Kochi19.83100.572.83104.824.252.824.05
      Fukuoka4.31114.352.94125.123.332.572.66
      Saga1.2239.07046.40000
      Nagasaki1.4911.930.7512.680.756.255.88
      Kumamoto5.6924.470.5726.750.572.332.13
      Oita24.4852.450.8752.450.871.671.67
      Miyazaki2.7815.73015.73000
      Kagoshima0.626.2006.20000
      Okinawa5.5291.852.7697.382.763.012.84
      The annual change in routine infant BCG vaccine coverage is shown in Fig. 1 and Table 2. BCG vaccine coverage was relatively low and variable among prefectures between 1999 and 2002. In eight prefectures coverage was less than 50%, and in Tokyo and Kanagawa less than 30% of the live births. In 2004, in accordance with the national policy on improving tuberculosis prevention and change in the target age of vaccination, the nationwide BCG vaccine coverage was 1.179 and that of Iwate was as high as 1.469. In 2006 and thereafter, the vaccine coverage was more than 50%, except for 42% in Miyagi in 2010. The vaccination coverage gradually stabilized, with all prefectures maintaining a greater than 90% coverage rate since 2013.
      Fig 1
      Fig. 1Annual change in routine infant BCG vaccine coverage. Top: Change in annual number of routine infant BCG vaccines (grey squares) and live births (black diamonds). The number of live births has gradually decreased over the last 20 years. The number of BCG vaccines has fluctuated according to changes in the routine vaccination system. Bottom: Annual change in routine infant BCG vaccine coverage, represented by the number of BCG vaccines per number of live births. *p<0.05, Mann–Whitney U test. aCOVID-19 prevalence on March 29 (patients per million).
      Table 2Routine infant BCG coverage in each prefecture, 1998–2007.
      Year1998199920002001200220032004200520062007
      All Japan0.9651.0050.9480.9690.9740.9691.1790.9360.8951.000
      Hokkaido0.9320.6490.5520.5280.500*0.9531.0990.9530.8280.984
      Aomori0.9731.0080.9290.8810.9671.0141.4620.6130.8300.990
      Iwate0.9701.0360.9660.9891.0240.9971.4690.4720.7220.991
      Miyagi0.9270.5220.4630.4600.5290.7931.1860.8600.9200.978
      Akita1.0100.7520.6820.6560.5951.0411.4231.0020.9770.999
      Yamagata0.9910.9501.0221.0001.0391.0461.2091.0340.9860.995
      Fukushima0.9190.6810.6190.6980.6211.0151.2701.0240.9750.991
      Ibaraki0.8490.9810.9660.9030.9760.9481.2491.0150.8061.004
      Tochigi0.9590.7770.7190.7060.7570.9951.3820.9700.8470.965
      Gunma0.9761.0721.0060.9071.0340.9051.2260.8640.6751.022
      Saitama0.9371.0000.9010.9690.8250.9821.1461.0360.7680.953
      Chiba0.9280.8350.7780.7360.7311.0031.1690.8620.8600.978
      Tokyo0.9450.3360.2890.3420.2850.9230.9520.8980.9260.950
      Kanagawa0.9820.4250.4150.2900.2960.9991.0240.9730.9261.000
      Niigata1.0020.7940.8250.7730.7640.8821.1840.9960.9380.995
      Toyama0.8910.7010.6680.7120.6931.0441.3991.9080.9160.916
      Ishikawa1.0240.6790.5690.5960.5990.9891.0801.0400.9911.058
      Fukui0.9911.0600.9811.0001.0151.0221.4620.9600.6341.009
      Yamanashi0.9790.9990.8860.9480.8440.9561.0791.0060.8240.913
      Nagano0.9900.8530.7790.8020.8471.0111.4830.8920.8570.985
      Gifu0.9000.8320.7040.8000.7980.7191.0250.8090.7510.988
      Shizuoka0.9790.6690.7340.7440.7031.0191.2410.9901.0040.988
      Aichi0.9960.6310.6000.6110.6321.0431.1880.9790.9801.005
      Mie1.0131.0970.9901.0060.9991.0421.2400.9970.7021.003
      Shiga1.0370.9781.0280.9601.0051.0371.3280.8120.9100.990
      Kyoto1.0020.3320.4280.4680.4580.9921.1750.9470.9381.425
      Osaka0.9650.6140.5460.5190.5480.9231.0990.8530.8780.961
      Hyogo1.0060.6810.6500.4930.4691.0161.1461.0000.9021.010
      Nara0.9600.9200.9631.0020.7400.9961.2701.0130.9320.985
      Wakayama0.9700.6550.5510.6250.6270.9711.2010.9701.3010.966
      Tottori1.0141.1490.9300.6501.0090.6571.1951.0270.9800.880
      Shimane0.9561.0370.9311.0230.9821.0711.3861.0071.1040.972
      Okayama0.9020.6360.4630.3660.3661.0151.2200.8950.9500.985
      Hiroshima0.9880.4150.4130.3360.3120.9321.1230.9581.0160.945
      Yamaguchi0.9851.0220.9390.8420.8501.0031.1581.0030.9100.991
      Tokushima0.6780.9980.9561.0380.9761.0381.3890.9170.9430.967
      Kagawa0.9180.6050.5420.6210.6020.8361.3411.0160.9790.919
      Ehime0.9971.1110.6490.6410.6461.0531.2491.0390.9910.987
      Kochi1.0840.5290.5050.5040.5221.0151.2230.9380.8110.988
      Fukuoka0.9850.5280.4460.3720.5150.8541.1950.7510.7721.252
      Saga1.0201.0580.9660.9841.0180.9131.2771.1470.9770.990
      Nagasaki0.9640.8020.7290.5450.5530.9301.2530.9900.9500.990
      Kumamoto0.9930.5760.5450.5430.5821.0261.4010.9940.9450.972
      Oita0.9440.6940.4730.4220.6811.0091.2750.5010.9680.960
      Miyazaki0.9870.7190.6720.6560.6830.9871.3160.8050.8670.950
      Kagoshima1.1480.6840.6550.6850.6631.0791.3870.9720.9510.978
      Okinawa0.9270.8850.9260.9250.9321.0161.4510.8800.9630.967
      Average0.9680.7870.7220.7080.7190.9731.2490.9490.9060.993
      Bold indicate values <0.5, and bold italics indicate <0.3. *0.4997 before rounding.
      Fig. 1 shows a comparison of annual BCG vaccine coverage among the five prefectures with no COVID-19 infections and the top five prefectures with the highest COVID-19 prevalence (Hokkaido, Tokyo, Chiba, Oita, and Hyogo) on March 29, 2020. In 1999, 2000, 2001, 2002, 2004, and 2012, the BCG vaccine coverage of the five prefectures with no infections was significantly higher than that of the prefectures with a high prevalence of COVID-19 infections (p<0.05). Prior to 2005, the target age of vaccination was less than 4 years old; therefore, the vaccine coverage in 1999 was actually relevant to the generation born between 1995 and 1999.
      The prevalence of patients with COVID-19, SARS-CoV-2 PCR+ individuals, and deaths owing to COVID-19 showed a significant negative correlation with BCG vaccine coverage in 2004, ratio of the elderly population (age 65 years or more), and the primary sector of industry, and a significant positive correlation to days from diagnosis of the first patient with COVID-19 to March 29 (p<0.01) (Fig. 2, Table 3). Because the elderly population showed an inverse association with prevalence of SARS-CoV-2 infection, additional correlation analyses by age group were performed. A significant positive correlation was shown between the prevalence of SARS-CoV-2 infection and age groups 20–34 and 40–54 years (Fig. 3). As for occupation, workers in the tertiary sector of industry were significantly correlated with SARS-CoV-2 infection in a positive manner (Fig. 3) (Table 3).
      Fig 2
      Fig. 2Correlation between COVID-19 prevalence and related factors. Scatter plot of COVID-19 prevalence (ordinate) against related factors (abscissa). Prevalence of (a) patients with COVID-19 on March 29, (b) SARS-CoV-2 PCR+ individuals on April 24, and (c) deaths on April 24 showed significant negative correlation with BCG vaccine coverage in 2004, the population age 65 years or more, and workers in the primary sector of industry; and significant positive correlation with days from the first patient diagnosis with COVID-19 up to March 29. Dotted squares indicate significant correlation (Spearman's rank correlation, p<0.01). Regression lines are shown for reference.
      Table 3Correlation coefficient (Spearman's ρ) between SARS-CoV-2 infection and related factors.
      Prevalence (per million)Deaths/PCR+ (%)
      Patients with COVID-19 on Mar 29PCR+ on Apr 24Deaths on Apr 24PCR+ on May 1Deaths on May 1On Apr 24On May 1
      BCG vaccine coverage (ratio)
      19980.028−0.047−0.021−0.0620.0160.0000.006
      1999−0.274−0.306*−0.189−0.303*−0.217−0.064−0.079
      2000−0.271−0.254−0.205−0.254−0.230−0.115−0.133
      2001−0.299*−0.223−0.254−0.219−0.276−0.196−0.208
      2002−0.324*−0.268−0.254−0.268−0.273−0.185−0.188
      2003−0.175−0.162−0.073−0.156−0.083−0.031−0.054
      2004−0.460**−0.489**−0.366**−0.497**−0.381**−0.245−0.264
      2005−0.278−0.025−0.121−0.015−0.140−0.116−0.141
      2006−0.332*−0.261−0.298*−0.230−0.285−0.300*−0.272
      20070.1830.161−0.305*0.1510.303*−0.374*0.361*
      20080.0140.1000.1120.1100.0940.0840.080
      2009−0.100−0.037−0.08−0.036−0.052−0.160−0.112
      20100.033−0.0760.011−0.0720.0090.0450.020
      20110.1320.2100.2660.1990.2430.2200.235
      2012−0.1040.1310.0790.1210.0690.0130.027
      2013−0.154−0.036−0.207−0.048−0.215−0.196−0.215
      2014−0.107−0.074−0.244−0.089−0.259−0.205−0.228
      20150.0890.1370.2170.1270.1950.2010.213
      20160.0350.1100.0460.1110.0240.0250.032
      20170.2070.1210.1640.1240.1500.1570.169
      Inhabitable area (km2)−0.003−0.162−0.098−0.160−0.148−0.039−0.085
      Age (ratio,%)
      65 years or more−0.382**−0.537**−0.436**−0.519**−0.413**−0.340*−0.302*
      0–4 years−0.0480.0680.1360.0660.1020.1290.083
      5–9 years−0.148−0.1090.073−0.1130.0330.1480.102
      10–14 years−0.209−0.1230.037−0.138−0.0160.1090.063
      15–19 years0.0050.2150.2200.1950.1800.1830.164
      20–24 years0.505**0.678**0.551**0.663**0.539**0.383**0.363*
      25–29 years0.430**0.562**0.462**0.549**0.445**0.345*0.307*
      30–34 years0.300*0.471**0.375**0.464**0.356*0.2840.243
      35–39 years0.2100.321*0.2310.314*0.2050.1600.113
      40–44 years0.383**0.613**0.504**0.599**0.485**0.344*0.317*
      45–49 years0.501**0.654**0.553**0.643**0.550**0.390**0.397**
      50–54 years0.546**0.512**0.400**0.503**0.399**0.308*0.318*
      55–59 years−0.185−0.359*−0.358*−0.349*−0.349*−0.206−0.187
      60–64 years−0.493**−0.665**−0.580**−0.656**−0.574**−0.405**−0.403**
      65–69 years−0.435**−0.618**−0.526**−0.602**−0.522**−0.381**−0.373*
      70–74 years−0.085−0.229−0.084−0.218−0.065−0.107−0.069
      75–79 years0.146−0.0550.030−0.0590.0680.0270.087
      80–84 years−0.337*−0.579**−0.475**−0.561**−0.450**−0.364*−0.328*
      85 years or more−0.492**−0.593**−0.509**−0.581**−0.483**−0.407**−0.383**
      Household with 5 or more members (ratio)−0.489**−0.252−0.323*−0.250−0.367*−0.296*−0.323*
      Day/night population (ratio)−0.1300.081−0.0080.0940.012−0.103−0.097
      Days 1st diagnosis - Mar 290.606**0.533**0.432**0.533**0.452**0.2910.298
      Inward population (ratio)0.370*0.642**0.507**0.638**0.486**0.339*0.300*
      Industry worker (%)
      Primary sector−0.466**−0.684**−0.579**−0.672**−0.553**−0.402**−0.383**
      Secondary sector−0.1360.032−0.0110.016−0.047−0.059−0.067
      Tertiary sector0.400**0.374**0.385**0.383**0.404**0.316*0.309*
      *p<0.05, **p<0.01.
      Fig 3
      Fig. 3Correlation between prevalence of patients with COVID-19 on March 29 and related factors. Scatter plot of the prevalence of patients with COVID-19 (ordinate) against related factors (abscissa). (A) Significant positive correlation of age groups 20–24, 25–29, 30–34, 40–44, 45–49 and 50–54 years with COVID-19 prevalence. (B) Workers in the tertiary sector of industry were significantly correlated with SARS-CoV-2 infection in a positive manner. Dotted squares indicate significant correlation (Spearman's rank correlation, p<0.01). Regression lines are shown for reference.

      Discussion

      In this study, we evaluated the relationship between BCG vaccination and SARS-CoV-2 infection in Japanese patients with COVID-19 infection. To the best of our knowledge, this is the first study to demonstrate an impact of routine infant BCG vaccine coverage among younger people, especially those born between 1995 and 2004, on local COVID-19 spread in Japan. The current data are in accordance with the concept of silent spreaders in which asymptomatic or mild cases have a substantial role in the dissemination of SARS-CoV-2.
      • Li R.
      • Pei S.
      • Chen B.
      • Song Y.
      • Zhang T.
      • Yang W.
      • Shaman J
      Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2).
      Moreover, our study findings suggest that infant BCG vaccination has a protective effect against mass infection.
      Diagnosis of COVID-19 was made based upon the Infectious Disease Surveillance System in Japan, in accord with the Act on the Prevention of Infectious Diseases and Medical Care for Patients with Infectious Diseases (the Infectious Diseases Control Law). Because in principle, the Act is applicable to individuals with flulike symptoms such as fever, cough, and fatigue, which are suggestive of SARS-CoV-2 infection, all patients officially notified on March 29 were symptomatic. Along with the disease spread, the Act has been applied to asymptomatic ones; the number of official notifications thereafter included PCR+ individuals both with and without clinical symptoms. The number of PCR+ individuals is lower than the actual number because of limited PCR testing in Japan; however, considering the relatively good correlation between the prevalence of PCR+ individuals and the PCR positivity rate, this likely has little involvement in nonparametric analyses (Table 4). Nevertheless, this difference may cause inconsistent results for the case fatality rate among prefectures.
      Table 4Prevalence of PCR+ individuals (descending order), number of PCR tests conducted, and positivity rate on April 24 in each prefecture.
      (A) Prevalence of PCR+ (per million)(B) Number of PCR testing (per million)(C) Positivity ratio (%)
      Tokyo271.09710.9738.10
      Ishikawa194.231245.8415.60
      Osaka164.30782.3719.20
      Toyama155.241795.248.60
      Fukui155.041696.389.10
      Chiba122.78780.0214.80
      Fukuoka114.351559.237.30
      Kyoto110.381394.837.90
      Hyogo110.321178.529.40
      Hokkaido106.321014.3810.50
      Saitama104.64803.4113.00
      Kochi100.571832.865.50
      Kanagawa99.38553.0116.60
      Okinawa91.851184.397.80
      Gifu73.111137.716.40
      Gunma71.721246.935.80
      Shiga65.86808.078.20
      Aichi62.89832.167.60
      Yamanashi62.422352.512.70
      Yamagata60.551786.243.40
      Nara56.76876.036.50
      Wakayama55.612701.602.10
      Ibaraki54.571401.113.90
      Oita52.452543.712.10
      Hiroshima50.411504.793.30
      Saga39.07868.134.50
      Miyagi36.27677.035.40
      Fukushima35.41761.804.60
      Ehime34.76741.124.70
      Nagano32.48750.364.30
      Kagawa29.111483.372.00
      Niigata28.941109.082.60
      Tochigi27.24859.713.20
      Mie25.13922.952.70
      Kumamoto24.471492.891.60
      Shimane23.531102.942.10
      Yamaguchi22.63888.322.50
      Aomori17.42409.344.30
      Akita16.31801.222.00
      Shizuoka16.12646.622.50
      Miyazaki15.73934.321.70
      Nagasaki11.931288.590.90
      Okayama10.54548.471.90
      Tokushima6.79524.461.30
      Kagoshima6.20687.110.90
      Tottori5.361658.930.30
      Iwate0.00225.620.00
      Spearman's rank correlation between (A) and (B): ρ=0.320, p<0.05, between (A) and (C): ρ=0.912, p<0.001.
      The significant negative correlation between COVID-19 infection and the population age 65 years and over suggests that in Japan, this population is irrelevant to disease spread. In contrast, the case fatality rate is high among those age 70 years or older (Fig. 4). Acute respiratory distress syndrome subsequent to SARS-CoV-2 respiratory infection is attributable to a high level of inflammatory cytokines, which increase with age, especially in men.
      • Hirano T.
      • Murakami M.
      COVID-19: a new virus, but a familiar receptor and cytokine release syndrome.
      ,
      • Ferrucci L.
      • Corsi A.
      • Lauretani F.
      • et al.
      The origins of age-related proinflammatory state.
      BCG vaccination induces trained immunity that enhances host defence mechanisms against nonspecific pathogens.
      • Kleinnijenhuis J.
      • Quintin J.
      • Preijers F.
      • et al.
      Bacille Calmette-Guerin induces NOD2-dependent nonspecific protection from reinfection via epigenetic reprogramming of monocytes.
      • Kleinnijenhuis J.
      • Quintin J.
      • Preijers F.
      • et al.
      BCG-induced trained immunity in NK cells: role for non-specific protection to infection.
      • Arts R.J.W.
      • Moorlag S.J.C.F.M.
      • Novakovic B.
      • et al.
      BCG vaccination protects against experimental viral infection in humans through the induction of cytokines associated with trained immunity.
      BCG-vaccinated neonates (BCG-Denmark 0.05 mL intradermally within 10 days of birth) showed increased interleukin 6 in an unstimulated state but decreased production of interleukins and chemokines following stimulation with pathogens.
      • Freyne B.
      • Donath S.
      • Germano S.
      • et al.
      Neonatal BCG vaccination influences cytokine responses to toll-like receptor ligands and heterologous antigens.
      This efficacy possibly contributed to the protection against SARS-CoV-2 spread, infection, and mortality in communities with high BCG vaccine coverage among younger people. In Japan, routine BCG vaccination has been the principal element in tuberculosis control. The incidence of tuberculosis in Japan has drastically decreased from 698.4 in 1951 to 17.7 per 100,000 in 2011.
      • Katsuda N.
      • Hirosawa T.
      • Reyer J.A.
      • Hamajima N
      Roles of public health centers (hokenjo) in tuberculosis control in Japan.
      After the introduction of BCG to the country in 1924, data on tuberculosis prevention and adverse event profiles had been accumulated mainly among young nurses or students.

      Toida I.History of BCG vaccination. Information and review of tuberculosis and respiratory disease research. 2004;48:15–40. Japanese.

      Group BCG vaccination commenced in 1946 for young people in their teens and 20 s, and in 1949 for all citizens younger than 30 years old; vaccination was repeated each year until positive tuberculin skin test result was obtained.

      Toida I.History of BCG vaccination. Information and review of tuberculosis and respiratory disease research. 2004;48:15–40. Japanese.

      ,
      • Obayashi Y
      Dried BCG vaccine.
      In 1949, a freeze-dried BCG vaccine that passed quality control assays replaced the previous liquid form.
      • Obayashi Y
      Dried BCG vaccine.
      A multiple puncture method was adopted in 1967, to avoid adverse skin reactions to intradermal vaccination.
      • Yamamoto S.
      • Yamamoto T.
      Historical review of BCG vaccine in Japan.
      ,

      Toida I.History of BCG vaccination. Information and review of tuberculosis and respiratory disease research. 2004;48:15–40. Japanese.

      ,
      • Obayashi Y.
      • Shimao T.
      • Ebina T.
      • et al.
      Comparison among various multi-puncture BCG vaccination methods.
      In 1974, the timing of routine infant vaccination was set at less than 4 years old, followed by revaccination twice during the first year of elementary school and during the first or second year of junior high school after tuberculin skin test screening.

      Toida I.History of BCG vaccination. Information and review of tuberculosis and respiratory disease research. 2004;48:15–40. Japanese.

      In 1999, a state of emergency concerning tuberculosis was declared because gradual mitigation of tuberculosis prevention measures during previous periods brought about a resurgence of newly infected patients.
      • Rahman M.
      • Takahashi O.
      • Goto M.
      • Fukui T
      BCG vaccination and tuberculosis in Japan.
      In 2003, revaccination was abandoned because of insufficient evidence of efficacy.

      Declaration of State of Emergency Concerning Tuberculosis. July 1999. https://www.mhlw.go.jp/www1/houdou/1107/h0726-2_11.html.

      The target age of routine infant BCG vaccination was set at younger than 6 months old in 2005 and then at 1 year old in 2013. High COVID-19 mortality among elderly people may be owing to a lack of benefit received from immunization as routine BCG vaccination for infants using qualified freeze-dried BCG commenced in 1949, and BCG vaccination was not administered in those with a positive tuberculin skin test. Current younger generations born in the late 1990s and early 2000s, before establishment of the reinforced tuberculosis control system, have relatively low BCG vaccine coverage. It is possible that their parents, age around 50 years (considering mean age of the mother at the first childbirth was 28 years old in 2000
      Part 1. Current Status of Countermeasures against Declining Birthrate.
      ), represent the significant positive correlation with COVID-19 infection in these age groups.
      Fig 4
      Fig. 4Number of PCR+ individuals (top), seriously ill patients and deaths owing to COVID-19 (middle), and their proportions (bottom) in each age group on May 1. A drastic increase in case fatality rate was observed with age, especially among individuals age 70 years or older. Shaded squares show age groups born before commencement of routine BCG vaccination for infants using qualified freeze-dried BCG.
      Several studies have compared the incidence and severity of COVID-19 among countries with different BCG vaccination policies and found a protective effect of universal BCG vaccination, especially during the past 15 years, and high vaccination coverage in young people under 25 years of age.
      • Sharma A.
      • Kumar Sharma S.
      • et al.
      BCG vaccination policy and preventive chloroquine usage: do they have an impact on COVID-19 pandemic? Version 2.
      • Klinger D.
      • Blass I.
      • Rappoport N.
      • Linial M
      Significantly improved COVID-19 outcomes in countries with higher BCG vaccination coverage: a multivariable analysis.
      • Ebina-Shibuya R.
      • Horita N.
      • Namkoong H.
      • Kaneko T
      National policies for paediatric universal BCG vaccination were associated with decreased mortality due to COVID-19.
      A cohort study in Rhode Island in the United States also demonstrated the potential of BCG in preventing severe COVID-19 requiring hospitalization
      • Weng C.H.
      • Saal A.
      • Butt W.W.
      • et al.
      Bacillus Calmette-Guérin vaccination and clinical characteristics and outcomes of COVID-19 in Rhode Island, United States: a cohort study.
      ; these findings are in accord with those of the present study. Two within-country studies from Israel and Germany failed to demonstrate the relevance of BCG vaccination to COVID-19 prevention; however, those studies do not contradict our study findings considering that universal immunization was discontinued in these countries in 1982 and 1990, respectively, and that the number of asymptomatic silent spreaders could not be evaluated.
      • Hamiel U.
      • Kozer E.
      • Youngster I
      SARS-CoV-2 Rates in BCG-vaccinated and unvaccinated young adults.
      ,
      • Bluhm R.
      • Pinkovskiy M.
      The spread of COVID-19 and the BCG vaccine: a natural experiment in reunified Germany.

      Conclusions

      In summary, the current study demonstrated that routine infant BCG vaccine coverage in young people showed a protective effect against local COVID-19 spread in Japan. The possible relevance of infant BCG vaccination to high mortality among elderly patients with COVID-19 should be elucidated.

      Funding

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Declaration of Competing Interest

      None.

      Acknowledgments

      We thank Analisa Avila, ELS, of Edanz Group (https://en-author-services.edanzgroup.com/ac) for editing a draft of this manuscript.

      References

      1. Ministry of Health, Labour and Welfare of Japan. About Coronavirus Disease 2019 (COVID-19). 2020. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000164708_00001.html. Accessed 28 May 2020.

        • de Vrieze
        Can a century-old TB vaccine steel the immune system against the new coronavirus?.
        Science. 2020; (Accessed 24 April 2020)
        • Redelman-Sidi G
        Could BCG be used to protect against COVID-19.
        Nat Rev Urol. 2020 Apr 27; https://doi.org/10.1038/s41585-020-0325-9
        • Yamamoto S.
        • Yamamoto T.
        Historical review of BCG vaccine in Japan.
        Jpn J Infect Dis. 2007; 60: 331-336
      2. Statistics Bureau, Ministry of Internal Affairs and Communications of Japan. e-Stat, Portal Site of Official Statistics of Japan. https://www.e-stat.go.jp. Accessed 24 April2020.

        • Li R.
        • Pei S.
        • Chen B.
        • Song Y.
        • Zhang T.
        • Yang W.
        • Shaman J
        Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2).
        Science. 2020; 368: 489-493
        • Hirano T.
        • Murakami M.
        COVID-19: a new virus, but a familiar receptor and cytokine release syndrome.
        Immunity. 2020; https://doi.org/10.1016/j.immuni.2020.04.003
        • Ferrucci L.
        • Corsi A.
        • Lauretani F.
        • et al.
        The origins of age-related proinflammatory state.
        Blood. 2005; 105: 2294-2299
        • Kleinnijenhuis J.
        • Quintin J.
        • Preijers F.
        • et al.
        Bacille Calmette-Guerin induces NOD2-dependent nonspecific protection from reinfection via epigenetic reprogramming of monocytes.
        Proc Natl Acad Sci U S A. 2012; 109: 17537-17542
        • Kleinnijenhuis J.
        • Quintin J.
        • Preijers F.
        • et al.
        BCG-induced trained immunity in NK cells: role for non-specific protection to infection.
        Clin Immunol. 2014; 155: 213-219
        • Arts R.J.W.
        • Moorlag S.J.C.F.M.
        • Novakovic B.
        • et al.
        BCG vaccination protects against experimental viral infection in humans through the induction of cytokines associated with trained immunity.
        Cell Host Microbe. 2018; 23 (89-100.e5)
        • Freyne B.
        • Donath S.
        • Germano S.
        • et al.
        Neonatal BCG vaccination influences cytokine responses to toll-like receptor ligands and heterologous antigens.
        J Infect Dis. 2018; 217: 1798-1808
        • Katsuda N.
        • Hirosawa T.
        • Reyer J.A.
        • Hamajima N
        Roles of public health centers (hokenjo) in tuberculosis control in Japan.
        Nagoya J Med Sci. 2015; 77: 19-28
      3. Toida I.History of BCG vaccination. Information and review of tuberculosis and respiratory disease research. 2004;48:15–40. Japanese.

        • Obayashi Y
        Dried BCG vaccine.
        World Health Organization Monograph, Geneva1955 (Series No.28)
        • Obayashi Y.
        • Shimao T.
        • Ebina T.
        • et al.
        Comparison among various multi-puncture BCG vaccination methods.
        Kekkaku. 1964; 39 (Japanese): 369-370
        • Rahman M.
        • Takahashi O.
        • Goto M.
        • Fukui T
        BCG vaccination and tuberculosis in Japan.
        J Epidemiol. 2003; 13: 127-135
      4. Declaration of State of Emergency Concerning Tuberculosis. July 1999. https://www.mhlw.go.jp/www1/houdou/1107/h0726-2_11.html.

      5. Part 1. Current Status of Countermeasures against Declining Birthrate.
        A 2018 Declining Birthrate White Paper (Summary). Cabinet Office, Government of Japan, 2018
        • Sharma A.
        • Kumar Sharma S.
        • et al.
        BCG vaccination policy and preventive chloroquine usage: do they have an impact on COVID-19 pandemic? Version 2.
        Cell Death Dis. 2020; 11: 516
        • Klinger D.
        • Blass I.
        • Rappoport N.
        • Linial M
        Significantly improved COVID-19 outcomes in countries with higher BCG vaccination coverage: a multivariable analysis.
        Vaccines (Basel). 2020; 8: E378
        • Ebina-Shibuya R.
        • Horita N.
        • Namkoong H.
        • Kaneko T
        National policies for paediatric universal BCG vaccination were associated with decreased mortality due to COVID-19.
        Respirology. 2020; https://doi.org/10.1111/resp.13885
        • Weng C.H.
        • Saal A.
        • Butt W.W.
        • et al.
        Bacillus Calmette-Guérin vaccination and clinical characteristics and outcomes of COVID-19 in Rhode Island, United States: a cohort study.
        Epidemiol Infect. 2020; 148: e140
        • Hamiel U.
        • Kozer E.
        • Youngster I
        SARS-CoV-2 Rates in BCG-vaccinated and unvaccinated young adults.
        JAMA. 2020; 323: 2340-2341
        • Bluhm R.
        • Pinkovskiy M.
        The spread of COVID-19 and the BCG vaccine: a natural experiment in reunified Germany.
        Covid Econ. 2020; 19: 87-114