Secondary Logo

Journal Logo

Original Studies

Measles Vaccine Effectiveness and Duration of Vaccine-induced Immunity in the Absence of Boosting from Exposure to Measles Virus


Author Information
The Pediatric Infectious Disease Journal: December 1996 - Volume 15 - Issue 12 - p 1082-1086
  • Free



Peter Panum's1 description of the epidemiology of measles in the 1846 Faroe Islands outbreak provides strong evidence that immunity after measles disease is lifelong. Whether vaccine-induced immunity is also lifelong is unknown. At least 95% of the individuals who receive a single dose of measles vaccine at the appropriate age develop measles antibodies.2 Most vaccinated individuals continue to have detectable antibodies, although titers of vaccine-induced antibodies decline over time.3 Mathias et al.4 have demonstrated that secondary vaccine failure (i.e. an adequate response initially developed, but immunity was lost over time) could occur among persons who developed hemagglutination inhibition and neutralizing antibody responses after measles vaccination. However, most studies that have evaluated the duration of vaccine-induced immunity suggest that, after seroconversion, live measles vaccine provides long lasting immunity. Although data from animal studies suggest that antigenic stimulation from periodic exposure to measles virus may provide boosting that is necessary for such long term immunity,5 it is not known whether all or some humans need periodic exposure to measles virus after vaccination to maintain protection against measles. In settings where measles elimination has been achieved and exposure to measles virus ceases, it is important to know whether vaccination alone will provide long term immunity and whether vaccinees will be protected against measles if the virus is reintroduced after a long absence.

The duration of vaccine-induced immunity in the absence of boosting can be studied only in a population where vaccinated persons have not been exposed to measles virus for many years. During 1993 a large measles outbreak in an isolated island population of Palau presented an opportunity for such an investigation. The objectives of the investigation were to examine the risk of developing measles among 0-, 1- and 2-dose-vaccine recipients and to estimate the effectiveness of measles vaccine and duration of vaccine-induced immunity in the absence of boosting from exposure to measles virus.


The Republic of Palau consists of several hundred islands situated in the western Pacific Ocean. Ninety-three percent of its population of 15 122 (1990 census) lives on the islands of Koror and Babeldaob. Measles vaccine has been used in Palau since 1969.6 Since 1989, 2 doses of measles vaccination have been recommended at ages 15 months and 4 to 6 years. At the time of the outbreak in 1993, 1-dose measles vaccination coverage among children age 2 years and 2-dose vaccination coverage among children age 6 years were 94 and 69%, respectively.

On May 18, 1993, a case of measles occurred in a Japanese tourist visiting Palau. During the ensuing 5 months a total of 93 confirmed measles cases were reported to the Bureau of Public Health. This was the first documented measles outbreak since 1966.6 Thus it is likely that measles virus had been absent from Palau for 27 years. The overall attack rate during the 1993 outbreak was 6.6 cases/1000 population. The highest attack rate (69/1000 population) was observed among infants (age <12 months). The lowest attack rates (2/1000 population) were observed among children age 5 to 9 years and persons age 25 years and older. Forty-six percent of the cases had a history of measles vaccination.


Case definition. A confirmed measles case was defined as a person who had (1) serologic evidence of acute measles infection or (2) at least a 3-day history of generalized maculopapular rash, fever (>101°F if measured), any one or more of the three symptoms of cough, coryza or conjunctivitis and was epidemiologically linked to another measles case. A primary case was defined as a confirmed case having the first rash onset in the household. A co-primary case was defined as a confirmed measles case in the household who developed rash within 6 days after rash onset of the primary case. A secondary case was defined as a confirmed measles case in the household who developed rash 7 to 18 days after rash onset of the primary case.

Data collection. Households enrolled in the study were selected by reviewing the measles case reports at the Palau Bureau of Public Health. Households of cases who recently had rash onset (but at least 3 weeks before the interview) and were living in Koror or Airai (the village closest on Babeldaob to Koror) were selected to reduce recall problems and reach as many household contacts as possible. Comparison of the selected cases with the remaining cases in the outbreak did not reveal a significant difference regarding demographic characteristics, such as sex and age.

Trained interviewers visited the households and interviewed the family members to identify the first person with rash onset in the household. This person was designated as the primary case. Using a standard questionnaire, each primary case was interviewed regarding symptoms, complications and exposure to other measles cases. They were also asked to give the names of the household members who were exposed to them during the period beginning 3 days before rash onset and ending 3 days after the rash onset (primary case's infectious period).

During the same visit each household contact was interviewed using a standardized questionnaire. Information collected included: (1) duration of exposure to the primary case during the infectious period; (2) symptoms suggesting viral illness (i.e. rash, fever, cough, conjunctivitis, coryza, muscle ache, headache, sore throat, diarrhea and vomiting) during the 3 weeks after their exposure; and (3) dates of measles vaccination. Blood was obtained from persons who reported having at least one symptom of viral illness. Health records of all contacts who reported having measles or were uncertain whether they had measles before 1993 were reviewed to verify their exposure to measles before this epidemic. Vaccination cards were reviewed when available. Vaccination status of all contacts was verified at the Bureau of Public Health and the Belau National Hospital by reviewing written records with dates of vaccination. A valid dose of measles vaccine in a contact was defined as a dose received on or before the day of fever onset in the primary household case. The person was considered unvaccinated if no vaccination record could be located or if the record did not list a dose of measles vaccine.

Blood specimens collected from suspected cases among household contacts were tested at the measles laboratory of the CDC by direct capture IgM and indirect IgG enzyme immunoassays.7

Data analysis. To reduce the misclassification of measles, measles exposure and measles immunity, the following household contacts were excluded from the analysis: co-primary cases; persons born before 1966 or born outside of Palau; persons younger than 12 months of age; persons exposed to the primary household case for <2 h; and persons vaccinated before their first birthday. Persons born before 1966 were excluded because they were thought to be immune to measles from disease during the previous outbreak. The threshold time (2 h) for exposure was selected based on the minimum time a household contact who developed measles was exposed to the primary case.

Household contacts were stratified by their vaccination status. Measles secondary attack rates among unvaccinated household contacts and contacts vaccinated with one or two doses of measles vaccine were calculated. Measles attack rates for unvaccinated household contacts and contacts vaccinated with one dose of measles vaccine were compared to obtain the relative risk (RR) for measles disease of those vaccinated. Vaccine effectiveness (VE) was calculated by the formula VE = (1 - RR) × 100.8 To assess significance of more than one dose of measles in preventing measles, attack rates among household contacts who had received one dose were compared with the attack rates among contacts who had received more than one dose of measles vaccine. Differences in attack rates were tested for statistical significance by the chi square test of Fisher's exact test where appropriate.9

Household contacts who had received only one dose of measles vaccine were further stratified by the number of years since vaccination. Measles attack rates were calculated and compared among household contacts who had different intervals since vaccination to assess the RR and its 95% confidence intervals (95% CI) for measles. Relative risk and 95% CI were calculated, using different intervals since vaccination (e.g. 2, 4, 5 years) as reference values.10


Study population. Eleven primary cases reported a total of 168 persons with household contact (Fig. 1). Of these, 147 (88%) were located and interviewed. Vaccination records of 129 (88%) contacts who were interviewed were located at the Bureau of Public Health or the Belau National Hospital. Sixty-nine (47%) contacts were excluded from the analysis, including 2 who were co-primary cases, 46 who were born before 1966, 11 who were born outside of Palau, 4 who were exposed to the primary household case for less than 2 h, 5 who were <12 months old and 1 who was vaccinated before 12 months of age (Fig. 1).

Of the remaining 78 household contacts, 5 were unvaccinated, 35 had received 1 and 38 had received >1 dose of measles vaccine (5 of which had received 3 doses). Median ages were: unvaccinated contacts, 1 year; contacts who had received one dose of vaccine, 10 years; and those who had received >1 dose, 13 years. Approximately one-half of all household contacts were female. The median duration of exposure to the primary case ranged between 1 and 3 days in each group.

Of the 78 household contacts 24 (31%) reported having symptoms during 3 weeks after the rash onset of the primary case. Eight had symptoms that met the clinical case definition for measles; 7 of these were serologically tested and had evidence of acute measles infection. Of the remaining 16 contacts who had symptoms that did not meet the measles case definition, 7 were tested 5 to 15 days after their symptoms and had evidence of measles immunity but not acute measles infection.

Vaccine effectiveness. The secondary measles attack rate was 80% (4 of 5) among unvaccinated household contacts and 11.4% (4 of 35) among contacts who had received one dose of measles vaccine (Table 1). None of the household contacts who had received more than one dose of measles vaccine developed measles.

The effectiveness of one dose of measles vaccine was 86% (95% CI = 60 to 95%). Receiving an additional dose of measles vaccine further reduced the risk of measles compared with those who had received only one dose (P = 0.048).

Duration of vaccine-induced immunity. Among the 35 household contacts who had received 1 dose, the interval since receiving the measles vaccine ranged between 0 days and 23 years (mean, 9.5 years; median, 8 years). The four secondary cases had received their last measles vaccine 1, 3, 18 and 20 years before their exposure to the primary case.

The attack rates among persons vaccinated 1 to 5 years and 16 to 20 years before exposure to measles were 15.4 and 33.3%, respectively. No cases occurred among contacts in other interval groups. The risk for developing measles among one-dose recipients who were vaccinated >15 years before exposure was 1.6 times higher than those vaccinated within 5 years (Table 2). This difference was not statistically significant (95% CI = 0.3 to 9.4). The difference between attack rates was not significant for 2-, 3-, or 4-year intervals used in the analysis.


Results of this study in an isolated community with no known exposure to measles virus during a 27-year period suggest that the effectiveness of one dose of measles vaccine is 86% (95% CI = 60 to 95%) and a second dose further reduces the risk of developing measles disease. We found no evidence that waning immunity was an important problem in this limited population with no known exposure to circulating wild measles virus. However, our study was limited by a small final sample size. To detect a true relative risk of 1.6 in our analysis of duration of immunity, the power of our study was 13%. To raise the power of the study to 80% with an alpha level of 0.5, ≈6 times as many household contacts would be necessary.

The study results indicate that overall 80% of the unvaccinated individuals were susceptible to measles after household exposure during this outbreak. The only unvaccinated household contact who did not develop measles was a 1-year-old child who, according to public health records, was vaccinated 1 month after being exposed to the primary case. It is possible that she may have been vaccinated before being exposed to the primary case and date of vaccination was recorded with an error. Although the vaccine effectiveness estimate in this study is lower than the 95% effectiveness estimate recently obtained in the United States,11 it is consistent with the effectiveness estimates previously obtained in two islands in the western Pacific Ocean in 1978 and 1992. Those estimates were 84% (95% CI = 74 to 89%)6 and 80% (95% CI = 67 to 88%) (Dr. Steven Auerbach, personal communication), respectively. Lower vaccine effectiveness in the area may be the result of loss of potency caused by improper handling and storage practices. Although no clustering among cases based on dates of vaccination was observed, it is likely that vaccines are exposed to warm temperatures and lose potency during frequent power shortages or transportation to the islands.

We found no statistically significant trend in increasing attack rates among household contacts as the time since vaccination increased. Four measles cases occurred among people who were vaccinated 1 to 5 (2) and 16 to 20 (2) years before being exposed to measles virus. No cases were observed among people who were vaccinated <1, 6 to 15 or >20 years before being exposed to measles. Although the household secondary attack rates (14%) among people vaccinated >15 years ago was higher than the attack rates (4%) among people who were vaccinated 1 to 5 years ago, relative risk was small and not statistically significant (RR = 1.6, 95% CI 0.3 to 9.4). The findings did not change as different time intervals were used in the analysis. Although our rigorous efforts to reduce potential bias resulted in a small sample size and no statistically significant difference, this study did provide unique data on persistence of vaccine-induced immunity in the absence of boosting.

Given the rare occurrence of outbreaks in isolated communities where measles virus has not been circulating for many years, very few studies have been conducted to evaluate the duration of vaccine-induced immunity in the absence of exposure to measles virus.12, 13 In two isolated Pacific atoll populations people vaccinated with Edmonston B vaccine were followed for 5 years (1963-1968) during which they were not exposed to measles.12 Although a 2- to 3-fold decline was recorded in the serum hemagglutination inhibition antibody titers 5 years after immunization, neutralizing antibody was detected suggesting persisting immunity to measles. In Iceland 52 individuals who seroconverted after receiving Edmonston B vaccine were tested about 3 years later, in 1965.13 An approximately 2-fold decrease in hemagglutination inhibition titer was observed during this time. However, all sera, except one, had measurable antibody titer. In another study 88% of vaccinated individuals in Greenland had detectable antibody titers 11 years after vaccination and no known exposure to measles virus.14 Although these studies examined laboratory evidence for immunity, we examined actual protection against measles among a population which had a measles outbreak after almost 30 years of measles-free period.

In this study no cases occurred among people who had received two doses of measles vaccine. Previous studies have shown that revaccination reduces the susceptibility for measles compared with a single dose of measles vaccine.15-18 Persons with no detectable antibody before revaccination usually develop a strong antibody response to revaccination. Although some persons with preexisting antibodies may have an increase in antibody titer (“boosting”) when given a second dose of measles vaccine,19 the preponderance of data suggests that an anamnestic response (characterized by a rapid but often transient increase in IgG antibody with little or no detectable IgM antibody) accurs after a second dose of measles vaccination.20 This study confirms the high effectiveness of the second dose of measles vaccine, even in a setting in which the effectiveness of one dose of measles vaccine is lower than expected and reaffirms the current two-dose vaccination schedule as the best approach to control measles in the US and its territories.


We acknowledge Hana Ngiruchelbad, Nancy Mengloi, Rosemary Kiep, Henrietta Merei and Florencia Kitalong at the Bureau of Public Health for investigating measles cases and conducting household interviews; practitioners at Belau National Hospital, in particular Dr. Samir Patel, for their technical assistance; and Dr. Stephen Redd, Dr. Melinda Wharton and Dr. Stephen Hadler, National Immunization Program, Centers for Disease Control and Prevention, for reviewing the manuscript.

FIG. 1
FIG. 1:
Study population, Palau, 1993.


1. Panum LP. Observations made during the epidemic of measles on the Faroe Islands in the year 1846. Baltimore: Williams & Wilkins, 1938-1939.
2. Centers for Disease Control. Measles prevention: recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR 1989;38(n.S9):1-13.
3. Markowitz LE, Preblud SR, Fine PEM, Orenstein WA. Duration of live measles vaccine-induced immunity. Pediatr Infect Dis J 1990;9:101-10.
4. Mathias RG, Meekison WG, Arcand TA, Schechter MT. The role of secondary vaccine failures in measles outbreaks. Am J Public Health 1989;79:475-8.
5. Gary D, Skarvall H. B-cell memory is short-lived in the absence of antigen. Nature 1988;336:70-3.
6. McIntyre RC, Preblud SR, Polloi A, Korean M. Measles and measles vaccine efficacy in a remote island population. Bull WHO 1982;60:767-75.
7. Hummel KB, Erdman DD, Heath J, Bellini WJ. Baculovirus expression of the nucleoprotein gene of measles virus and utility of the recombinant protein in diagnostic enzyme immunoassays. J Clin Microbiol 1992;30:2874-80.
8. Orenstein WA, Bernier RH, Hinman AR. Assessing vaccine efficacy in the field. Epidemiol Rev 1988;10:212-41.
9. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: Wiley, 1981.
10. Greenland S, Robins JM. Estimation of a common effect parameter from sparse follow-up data. Biometrics 1985;41:55-68.
11. King GE, Markowitz LE, Patriarca PA, Dales LG. Clinical efficacy of measles vaccine during the 1990 measles epidemic. Pediatr Infect Dis J 1991;10:883-8.
12. Brown P, Gajdusek DC, Tsai T. Persistence of measles antibody in the absence of circulating natural virus five years after immunization of an isolated virgin population with Edmonston B vaccine. Am J Epidemiol 1969;90:514-8.
13. Black FL, Gudnadottir M. Measles vaccination in Iceland: Three-year follow-up of antibody titres in adults and children. Bull WHO 1966;35:955-7.
14. Pedersen IR, Mordhorst CH, Ewald T, von Magnus H. Longterm antibody response after measles vaccination in an isolated Arctic society in Greenland. Vaccine 1986;4:173-8.
15. Watson JC, Pearson JA, Markowitz LE, et al. An evaluation of measles revaccination among school-entry-aged children. Pediatrics 1996;97:613-8.
16. Nkowane BM, Bart SW, Orenstein WA, Baltier M. Measles outbreak in a vaccinated school population: epidemiology, chains of transmission and the role of vaccine failures. Am J Public Health 1987;77:434-8.
17. Hutchins SS, Markowitz LE, Mead P, et al. A school-based measles outbreak: the effect of a selective revaccination policy and risk factors for vaccine failure. Am J Epidemiol 1990;132:157-68.
18. Robertson SE, Markowitz LE, Berry DA, Dini EF, Orenstein WA. A million dollar measles outbreak: epidemiology, risk factors, and a selective revaccination strategy. Public Health Rep 1992;107:24-31.
19. Christenson B, Bottiger M. Measles antibody: comparison of long-term vaccination titres, early vaccination titres and naturally acquired immunity to and booster effects on the measles virus. Vaccine 1994;12:129-33.
20. Markowitz LE, Katz SL. Measles vaccine. In: Plotkin SA, Mortimer EA, eds. Vaccines. 2nd ed. Philadelphia: Saunders, 1994:229-76.

Measles; measles vaccine; vaccine effectiveness; duration of vaccine-induced immunity; immunization; outbreak; Palau; Micronesia

© Williams & Wilkins 1996. All Rights Reserved.