Advertisement

Incidence Rate of Herpes Zoster Ophthalmicus

A Retrospective Cohort Study from 1994 through 2018
  • Christina L. Kong
    Affiliations
    F. I. Proctor Foundation, University of California, San Francisco, San Francisco, California
    Search for articles by this author
  • Ryan R. Thompson
    Affiliations
    F. I. Proctor Foundation, University of California, San Francisco, San Francisco, California
    Search for articles by this author
  • Travis C. Porco
    Affiliations
    F. I. Proctor Foundation, University of California, San Francisco, San Francisco, California

    Department of Ophthalmology, University of California, San Francisco, San Francisco, California

    Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California
    Search for articles by this author
  • Eric Kim
    Affiliations
    F. I. Proctor Foundation, University of California, San Francisco, San Francisco, California
    Search for articles by this author
  • Nisha R. Acharya
    Correspondence
    Correspondence: Nisha R. Acharya, MD, MS, F. I. Proctor Foundation, University of California, San Francisco, 513 Parnassus Avenue, S309, San Francisco, CA 94143.
    Affiliations
    F. I. Proctor Foundation, University of California, San Francisco, San Francisco, California

    Department of Ophthalmology, University of California, San Francisco, San Francisco, California

    Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California

    OptumLabs, Visiting Fellow, Cambridge, Massachusetts
    Search for articles by this author
Published:October 09, 2019DOI:https://doi.org/10.1016/j.ophtha.2019.10.001

      Purpose

      To analyze the incidence rate (IR) of herpes zoster ophthalmicus (HZO) and differences by age, gender, race, and region from 1994 through 2018.

      Design

      Retrospective, observational cohort study.

      Participants

      Patients with a new International Classification of Diseases, Ninth or Tenth Edition, codes for herpes zoster (HZ) and HZO from January 1, 1994, through December 31, 2018, in the OptumLabs Data Warehouse (OptumLabs, Cambridge, MA).

      Methods

      OptumLabs Data Warehouse, a longitudinal, real-world data asset with de-identified administrative claims and electronic health record data, was used to identify enrollees with continuous enrollment in the database for 365 days or more. Patients with no history of HZ or HZO and a new code for HZ and HZO were counted as incident cases. The IR of HZO was calculated by year, 10-year age groups, gender, race, and region.

      Main Outcome Measures

      Differences in IR from 1994 through 2018 by 10-year age groups and gender.

      Results

      From 1994 through 2018, 633 474 cases of HZ were reported, with 49 745 (7.9%) having HZO. The incidence of HZO increased from 1994 through 2018 by an estimated 1.1 cases per 100 000 person-years annually (95% confidence interval [CI], 1.0–1.3; P < 0.001). The estimated relative increase was 3.6% annually (95% CI, 3.0%–4.1%). HZO IR increased in all ages over 10 years until 2007, then began declining in individuals younger than 21 and older than 60, stabilizing in individuals 21 to 30 years old, and increasing more slowly among individuals 31 to 60 years old. Men showed an HZO incidence rate ratio (IRR) of 0.74 compared with women. Compared with white patients, the IRRs were 0.70, 0.75, and 0.64 for Asians, black patients, and Hispanics, respectively.

      Conclusions

      The incidence of HZO has increased 3.6% per year from 1994 to 2018 in the United States. Since 2008, HZO incidence declined in individuals younger than 21 years and older than 60 years while increasing at a lower rate in middle-aged adults. Given the continued increase, greater efforts should be made to vaccinate eligible adults 50 years of age and older. More research on earlier vaccination is warranted.

      Abbreviations and Acronyms:

      ACIP (Advisory Committee on Immunization Practices), CDC (Centers for Disease Control and Prevention), CI (confidence interval), HZ (herpes zoster), HZO (herpes zoster ophthalmicus), ICD (International Classification of Diseases), IR (incidence rate), IRR (incidence rate ratio), OLDW (OptumLabs Data Warehouse), RZV (recombinant zoster vaccine), ZVL (zoster vaccine live)
      Herpes zoster (HZ), also known as shingles, occurs in 1 in 3 individuals during their lifetime in the United States and typically manifests as a painful dermatomal rash.
      • Ragozzino M.W.
      • Melton L.J.
      • Kurland L.T.
      • et al.
      Population-based study of herpes zoster and its sequelae.
      Herpes zoster ophthalmicus (HZO), a form of HZ that occurs when varicella zoster virus reactivates along the ophthalmic branch of the trigeminal nerve, will develop in approximately 10%–20% of cases of HZ.
      • Liesegang T.J.
      Herpes zoster ophthalmicus: natural history, risk factors, clinical presentation, and morbidity.
      Of those individuals with HZO, 50% to 71% show ocular involvement with complications such as keratitis, uveitis, retinal necrosis, and loss of vision, leading to significant pain, morbidity, and decreased quality of life.
      • Weinberg J.M.
      Herpes zoster: epidemiology, natural history, and common complications.
      • Yawn B.P.
      • Wollan P.C.
      • St. Sauver J.L.
      • Butterfield L.C.
      Herpes zoster—eye complications: rates and trends.
      • Nithyanandam S.
      • Stephen J.
      • Joseph M.
      • Dabir S.
      Factors affecting visual outcome in herpes zoster ophthalmicus: a prospective study.
      • Vrcek I.
      • Choudhury E.
      • Durairaj V.
      Herpes zoster ophthalmicus: a review for the internist.
      It is also now recognized that approximately 20% of patients with HZO may have a chronic course requiring ongoing treatment.
      • Tran K.D.
      • Falcone M.M.
      • Choi D.S.
      • et al.
      Epidemiology of herpes zoster ophthalmicus recurrence and chronicity.
      The few studies on HZ incidence suggest that rates have been increasing in the 2 decades up to 2012, and a 2018 report from the Center for Disease Control and Prevention (CDC) revealed that HZ rates have continued to rise in adults through 2016.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among older adults.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among children.
      • Wolfson L.J.
      • Daniels V.J.
      • Altland A.
      • et al.
      The impact of varicella vaccination on the incidence of varicella and herpes zoster in the United States: updated evidence from observational databases, 1991–2016.
      However, these studies did not report on HZO incidence. There is no consensus on why HZ has been increasing, although several theories do exist. One theory is that that the introduction of the varicella vaccine in 1996 lowered exposure to wild-type varicella infection in the population, decreasing immune boosting. Others have speculated that the rising HZ incidence may be the result of an increase in immunocompromised conditions or changes in health-seeking behavior, leading to increased diagnosis of HZ.
      • Goldman G.S.
      Cost-benefit analysis of universal varicella vaccination in the U.S. taking into account the closely related herpes-zoster epidemiology.
      • Chan A.Y.
      • Conrady C.D.
      • Ding K.
      • et al.
      Factors associated with age of onset of herpes zoster ophthalmicus.
      • Kawai K.
      • Yawn B.P.
      • Wollan P.
      • Harpaz R.
      Increasing incidence of herpes zoster over a 60-year period from a population-based study.
      Another major factor that may be affecting HZ epidemiology is the introduction of vaccines for HZ: Zostavax (zoster vaccine live [ZVL]; Merck & Co, Inc, Whitehouse Station, NJ) and Shingrix (recombinant zoster vaccine [RZV]; GlaxoSmithKline, Philadelphia, PA). Since the introduction of ZVL in 2006 and RZV in 2017, few reports on the influence of these vaccines on HZ rates have been published, and even fewer have looked at vaccine effect on HZO.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among older adults.
      ,
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among children.
      Aside from a small population study that showed increasing HZO from 1980 through 2007, the available information on HZO epidemiology has been limited.
      • Yawn B.P.
      • Wollan P.C.
      • St. Sauver J.L.
      • Butterfield L.C.
      Herpes zoster—eye complications: rates and trends.
      A retrospective study of a single-site tertiary care center suggests an increase in new HZO cases and an age shift toward younger individuals from 2007 through 2013; however, the referral bias inherent to this type of study makes it difficult to estimate if there has been a change in HZO over time.
      • Davies E.C.
      • Pavan-Langston D.
      • Chodosh J.
      Herpes zoster ophthalmicus: declining age at presentation.
      Overall, knowledge on HZO incidence rates (IRs) and whether they are shifting is lacking.
      Given the significant morbidity associated with HZO, studies on HZO epidemiologic features are crucial for informing public health policy and directing clinician efforts. In particular, the introduction of the vaccines for HZ emphasizes the need to understand who is at greatest risk of HZO and how these trends are changing over time. The objective of this study was to examine the IR of HZO in the American population from 1994 through 2018 and to assess by age, race, gender, and geographic region.

      Methods

      A retrospective, observational cohort study was conducted of more than 200 million records in the de-identified healthcare claims database OptumLabs Data Warehouse (OLDW; OptumLabs, Cambridge, MA).
      Optum Labs
      Optum Labs and Optum Labs Data Warehouse (OLDW) Descriptions and Citation.
      OptumLabs Data Warehouse is a data asset that contains administrative claims and electronic health record data for United States patients enrolled in commercial insurance, Medicare Advantage, or Medicare Part D plans. Comparisons between the OLDW and United States Census Bureau show that the age, gender, race, and geographic distributions of OLDW enrollees are similar to those of the United States population for both the commercially insured and the Medicare Advantage groups. Less representation in the West exists for Medicare Advantage enrollees in OLDW compared with United States Census estimates, with fairly comparable representation in all other regions.
      Optum Labs
      Optum Labs and Optum Labs Data Warehouse (OLDW) Descriptions and Citation.
      To be included in the study, enrollees must have been enrolled continuously for 365 days or more and not have a history of HZ and HZO. Between January 1, 1994, and December 31, 2018, more than 63 million unique people were represented in the OLDW. Cases of HZ and HZO were identified by using International Classification of Diseases (ICD), Ninth and Tenth Editions, codes. Patients with 1 new code for HZ (ICD-9 053.XX, ICD-10 B02.XX) and HZO (ICD-9 0532.X, ICD-10 B023.X) were counted as an incident case. Information was extracted on the patients’ year of birth, gender, race (Asian, black, Hispanic, white, other), and geographic region (Midwest, Northeast, South, West, and other). The exact date of birth is not available through OLDW. Therefore, age was approximated by year of birth (i.e., a patient categorized as 1 year of age in 1994 experienced his or her 1-year birthday in 1994). Data were not available for individuals born before 1930.
      The time trend in the overall incidence was assessed through Huber robust regression of the estimated annual IR on calendar year, using 10-year age-time intervals and broken-stick models of the time ranges from 1994 through 2007 and 2008 through 2018, adjusting for age.
      • Sen A.
      • Srivastava M.
      Regression Analysis: Theory, Methods and Applications.
      The specific time ranges were established to examine HZO IRs before and after the Advisory Committee on Immunization Practices’ (ACIP) recommendation for the ZVL vaccine in 2008.
      • Harpaz R.
      • Ortega-Sanchez I.R.
      • Seward J.F.
      Prevention of herpes zoster: recommendations of the Advisory Committee on Immunization Practices (ACIP).
      A sensitivity analysis using 2006 (the year the Food and Drug Administration approved ZVL) as the cut point was conducted. When comparing gender, race or ethnicity, and region, we calculated the IR ratio using bootstrap resampling of years to compute approximate 95% confidence intervals (CIs; a highly conservative procedure that avoids relying on the very large number of person-years as a basis for inference).
      Between 1994 and 2018, the IR of HZO was calculated by year, 10-year age groups, 10-year birth cohorts, gender, race, and geographic region. Data were stratified into 10-year age groups starting at younger than 1 year of age. Age groups also were aggregated into 10-year age groups to tabulate changes in the slope of HZO incidence overall and by gender from the periods 1994 through 2007 and 2008 through 2018. As a sensitivity analysis, we also standardized the IRs by age, gender, and race or ethnicity using the average of the years 2014 through 2018 in the OLDW database. Standardized IRs were analyzed in the same way using broken-stick Huber robust regression with adjustments for age. Statistical significance testing was conducted at a level of P < 0.01.
      All statistical analyses were conducted in R software version 3.5 (The R Project for Statistical Computing, Vienna, Austria). Only de-identified data were available for analysis, and thus, informed consent was not required for this study. The University of California, San Francisco, Institutional Review Board approved the study, and the described research adhered to the tenets of the Declaration of Helsinki.

      Results

      From January 1, 1994, through December 31, 2018, 633 474 cases of HZ were reported, of which 49 745 (7.9%) included a code specific for HZO. The IR of HZO increased from 1994 through 2018 by an estimated 1.1 cases per 100 000 person-years annually (95% CI, 1.0–1.3; P < 0.001). The estimated relative increase in the HZO IR was 3.6% per year (95% CI, 3.0%–4.1%). For individuals younger than 50 years with known race (representing 88% of the available person-years), the standardized IR increased an average of 0.31 cases per 100 000 person-years (95% CI, 0.20–0.41) from 1994 through 2018. For individuals older than 70 years, the standardized IR changed an average of –0.86 cases per 100 000 person-years (95% CI, –1.36 to –0.37) from 2010 through 2018.
      Figure 1 demonstrates the IR of HZO by 10-year age groups from 1994 through 2018. The age-specific IRs changed from 1994 through 2007 compared with 2008 and beyond, which is explained further in Table 1. Using broken-stick regression, the slope of the HZO IR shifted from –0.24 (95% CI, –0.36 to –0.11) to –0.37 (95% CI, –0.52 to –0.22) per 100 000 person-years per year in individuals in the 0 to 10 years age group, 0.74 (95% CI, 0.47–1.00) to 0.14 (95% CI, –0.17 to 0.46) in the 41 to 50 years age group, and 6.23 (95% CI, 5.05–7.40) to –2.26 (95% CI, –2.96 to –1.57) in individuals 61 to 70 years of age (Table 1). The changes in the slopes of HZO incidences among all age groups from 1994 through 2007 compared with 2008 through 2018 were statistically significant except for in children between 0 and 10 years of age (P = 0.56). Using 2006 as the cut point did not change the overall trends in HZO IR. Results were similar based on the standardized IRs, except that there was no statistically significant evidence of a change in slope for the age group 31 to 40 years of age (P = 0.24).
      Figure thumbnail gr1
      Figure 1Graph showing the incidence rate of herpes zoster ophthalmicus by 10-year age groups from 1994 through 2018.
      Table 1Slopes of Herpes Zoster Ophthalmicus Incidence Rates per Year from 1994 through 2018 by 10-Year Age Groups
      Age (yrs)Year RangeP Value
      P value corresponds to comparison of slopes between the 2 time intervals.
      1994–20072008–2018
      Slope of Herpes Zoster Ophthalmicus Incidence Rate
      Change in HZO incidence rate per 100 000 person-years per year.
      95% Confidence IntervalSlope of Herpes Zoster Ophthalmicus Incidence Rate95% Confidence Interval
      0-10–0.24–0.36–0.11–0.37–0.52–0.220.56
      11-200.490.350.63–0.74–0.91–0.57<0.001
      21-300.430.230.63–0.04–0.290.200.002
      31-400.590.390.800.18–0.070.430.001
      41-500.740.471.000.14–0.170.46<0.001
      51-601.180.811.550.31–0.140.76<0.001
      61-70
      The slope was computed from 2000 through 2007 given that the available data do not include individuals born before 1930 and the full age range could not be captured until 2000.
      6.235.057.40–2.26–2.96–1.57<0.001
      71+
      The slope was not available for ages 71 or older for the first period from 1994 through 2007 because the available data do not include individuals born before 1930.
      –3.76–4.95–2.56
      — = not available.
      P value corresponds to comparison of slopes between the 2 time intervals.
      Change in HZO incidence rate per 100 000 person-years per year.
      The slope was computed from 2000 through 2007 given that the available data do not include individuals born before 1930 and the full age range could not be captured until 2000.
      § The slope was not available for ages 71 or older for the first period from 1994 through 2007 because the available data do not include individuals born before 1930.
      Table 2 shows that HZO IR increased by 10-year age group, starting with an average of 4.8 cases per 100 000 person-years in children ages 1 to 10 and rising to 131.6 in ages 81 to 90 from 1994 to 2018. Incidence rates also increased by decade of birth, with the oldest cohort, born in 1930 through 1939, having the highest average IR of 115.7 cases per 100 000 person-years and individuals born in 2010 through 2019 having an IR of 2.1 cases per 100 000 person-years from 1994 through 2018.
      Table 2Incidence Rates of Herpes Zoster Ophthalmicus by Age Group, Decade of Birth, Gender, Race, and Region from 1994 through 2018
      No. of CasesPerson-YearsIncidence Rate
      Incidence rate is the number of cases per 100 000 person-years.
      Age group (yrs)
       0–1063813 301 1964.8
       11–20130516 767 3977.8
       21–30176813 357 95313.2
       31–40424318 900 67622.4
       41–50699921 701 97832.3
       51–6010 73219 660 93554.6
       61–7010 44912 796 08481.7
       71–8010 0308 834 445113.5
       81–9035812 721 988131.6
       Unknown021950.0
      Decade of birth
       1930–193912 02110 390 005115.7
       1940–194911 65214 780 35578.8
       1950–195910 65320 624 33951.7
       1960–1969748521 497 81834.8
       1970–1979390317 258 87322.6
       1980–1989217815 020 07114.5
       1990–1999142016 063 7768.8
       2000–200939110 447 8903.7
       2010–2019421 959 5252.1
       Unknown021950.0
      Gender
       Female28 93765 077 05944.5
       Male20 65362 390 95133.1
       Other
      Individuals who had both male and female or neither gender listed.
      155576 83726.9
      Race
       Asian14044 674 28030.0
       Black345610 739 35132.2
       Hispanic295810 748 46527.5
       White35 08981 371 55043.1
       Other
      Individuals who were of a race that was not 1 of the 4 listed races.
      16605 899 17628.1
       Unknown517814 612 02535.4
      Region
       Midwest15 03437 808 50039.8
       Northeast628412 683 74549.5
       South19 89852 400 46938.0
       West601517 151 91935.1
       Other
      Individuals who had 2 or more regions or neither region listed.
      25148 000 21431.4
      Incidence rate is the number of cases per 100 000 person-years.
      Individuals who had both male and female or neither gender listed.
      Individuals who were of a race that was not 1 of the 4 listed races.
      § Individuals who had 2 or more regions or neither region listed.
      Compared to females, males had a lower overall HZO IR ratio (IRR) of 0.74 (95% CI, 0.72–0.76). For patients of both genders older than 10 years, the HZO IR slopes increased from 1994 through 2007 and shifted among different age groups from 2008 through 2018 (Table 3). Compared with white patients, Asians showed an IRR of 0.70 (95% CI, 0.67–0.73), black patients showed an IRR of 0.75 (95% CI, 0.72–0.79), and Hispanics showed an IRR of 0.64 (95% CI, 0.60–0.67). Compared with the Northeast, the Midwest showed an IRR of 0.80 (95% CI, 0.75–0.85), the West showed an IRR of 0.71 (95% CI, 0.66–0.76), and the South showed an IRR of 0.77 (95% CI, 0.71–0.81).
      Table 3Slopes of Herpes Zoster Ophthalmicus Incidence Rates per Year from 1994 through 2018 by 10-Year Age Groups and Gender
      GenderAge (yrs)Year Range
      1994–20072008–2018
      Slope of Herpes Zoster Ophthalmicus Incidence Rate
      Change in herpes zoster ophthalmicus incidence rate per 100 000 person-years annually.
      95% Confidence IntervalSlope of Herpes Zoster Ophthalmicus Incidence Rate95% Confidence Interval
      Male0–10–0.01–0.18 to 0.15–0.47–0.67 to –0.27
      11–200.600.41–0.78–0.80–1.03 to –0.58
      21–300.570.30–0.830.04–0.28 to 0.35
      31–400.420.15–0.700.15–0.18 to 0.48
      41–500.790.45–1.130.03–0.39 to 0.44
      51–601.421.01–1.830.05–0.44 to 0.55
      61–70
      The slope was computed from 2000 through 2007 given that the available data do not include individuals born before 1930 and the full age range could not be captured until 2000.
      4.693.31–6.06–1.87–2.69 to –1.05
      71+
      The slope was not available for ages 71+ for the first period from 1994 through 2007 for both men and women because the available data do not include individuals born before 1930.
      –4.35–6.00 to –2.69
      Female0–10–0.06–0.20 to 0.09–0.29–0.47 to –0.12
      11–200.630.45–0.81–0.72–0.93 to –0.50
      21–300.780.51–1.04–0.24–0.56 to 0.08
      31–400.910.61–1.210.19–0.17 to 0.55
      41–500.780.38–1.170.21–0.26 to 0.69
      51–601.140.56–1.720.59–0.11 to 1.29
      61–70
      The slope was computed from 2000 through 2007 given that the available data do not include individuals born before 1930 and the full age range could not be captured until 2000.
      7.695.91–9.46–2.69–3.75 to –1.63
      71+
      The slope was not available for ages 71+ for the first period from 1994 through 2007 for both men and women because the available data do not include individuals born before 1930.
      –3.26–5.02 to –1.50
      — = not available.
      Change in herpes zoster ophthalmicus incidence rate per 100 000 person-years annually.
      The slope was computed from 2000 through 2007 given that the available data do not include individuals born before 1930 and the full age range could not be captured until 2000.
      The slope was not available for ages 71+ for the first period from 1994 through 2007 for both men and women because the available data do not include individuals born before 1930.

      Discussion

      From 1994 through 2018, the HZO IR increased 3.6% per year in this study using a large United States administrative real-world data set. Before the ACIP recommendation for ZVL in 2008, HZO IRs were decreasing in children between 0 to 10 years of age and increasing in all older age groups. However, from 2008 and beyond, HZO IRs began to decline significantly among individuals younger than 21 years and older than 60 years. Herpes zoster ophthalmicus IRs began stabilizing in those 21 to 30 years of age and continued to increase, but at a less significant rate, among individuals between 31 and 60 years of age.
      In children younger than 10 years, the incidence of HZO cases decreased from 1994 through 2007 but dropped more precipitously from 2008 through 2018. The decreasing IRs may be secondary to the introduction of the varicella vaccine in 1996, and the continued rapid decline in the past decade may be associated with the implementation of the 2-dose varicella vaccination program in 2006. Varicella vaccination is widespread, with more than 80% of children older than 7 years having 2-dose coverage in 2012.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among children.
      This decline in HZO in young children is similar to the HZ trends published in a 2018 study by the CDC that found declining rates of HZ among children younger than 18 years from 1998 through 2016.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among children.
      In a multi-database study from 2003 through 2014 looking at varicella vaccination, researchers found that HZ incidence declined by 72% among 0- to 17-year-olds.
      • Weinmann S.
      • Naleway A.L.
      • Koppolu P.
      • et al.
      Incidence of herpes zoster among children: 2003–2014.
      The decline in HZO incidence since 2008 in individuals older than 60 years raises the question of whether the zoster vaccine is having an impact in this age group. The CDC report noted a deceleration in HZ cases among a similar older age group.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among older adults.
      Zoster vaccine coverage is thought to be low, with an upper estimate of 31% in 2015, but at this level, it may be having an impact on incidence in the age group that is vaccine eligible.
      • Williams W.W.
      • Lu P.J.
      • O’Halloran A.
      • et al.
      Surveillance of vaccination coverage among adult populations-United States, 2015.
      Herpes zoster ophthalmicus incidence in the 31- to 60-year-old groups has continued to increase since 2008, although at a slower rate than before that time. The decreasing HZO IRs among the younger and older population suggests that a larger proportion of new HZO cases may be occurring in individuals between 31 and 60 years of age because the overall rate continues to increase. Although changes in incidence are difficult to determine from retrospective studies from single centers, 1 study reported that HZO cases were increasing among individuals younger than 50 years, with the percentage of all HZO cases going from 16.2% between 1996 and 2004 to 29.6% between 2005 and 2012 in this age group.
      • Chan A.Y.
      • Conrady C.D.
      • Ding K.
      • et al.
      Factors associated with age of onset of herpes zoster ophthalmicus.
      Another study from a tertiary care center found that the average age of HZO onset went from 61 years in 2007 to 56 years in 2013.
      • Davies E.C.
      • Pavan-Langston D.
      • Chodosh J.
      Herpes zoster ophthalmicus: declining age at presentation.
      The increase in HZO incidence among the middle-age groups parallels the recent CDC report that found a continued increase in HZ rates among those between 35 and 55 years of age.
      • Harpaz R.
      • Leung J.W.
      The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among older adults.
      It is not possible to attribute the changing HZ and HZO epidemiology to a particular cause with an observational study, but there are several potential explanations. One theory is that intermittent exposure to wild-type varicella helped boost natural immunity and that universal varicella vaccination has decreased such exposure, leading to waning immunity and reactivation of the varicella virus at a younger age.
      • Liesegang T.J.
      Herpes zoster ophthalmicus: natural history, risk factors, clinical presentation, and morbidity.
      ,
      • Goldman G.S.
      Cost-benefit analysis of universal varicella vaccination in the U.S. taking into account the closely related herpes-zoster epidemiology.
      ,
      • Civen R.
      • Chaves S.S.
      • Jumaan A.
      • et al.
      The incidence and clinical characteristics of herpes zoster among children and adolescents after implementation of varicella vaccination.
      ,
      • Donahue J.G.
      • Kieke B.A.
      • Gargiullo P.M.
      • et al.
      Herpes zoster and exposure to the varicella zoster virus in an era of varicella vaccination.
      Yet, studies in Canada and the United Kingdom from a time when there was no comprehensive varicella vaccination program in those countries demonstrated an increase in overall HZ incidence, suggesting that there may be additional factors in effect.
      • Reynolds M.A.
      • Chaves S.S.
      • Harpaz R.
      • et al.
      The impact of the varicella vaccination program on herpes zoster epidemiology in the United States: a review.
      Other investigators have theorized that the live virus within the varicella vaccine has led to increasing cases of HZ, but this has been proven unlikely because the IRs for HZ were similar before and after the introduction of the vaccine.
      • Kawai K.
      • Yawn B.P.
      • Wollan P.
      • Harpaz R.
      Increasing incidence of herpes zoster over a 60-year period from a population-based study.
      The continued rise in HZO in the 31- to 60-year-old groups raises questions about the age recommendations for HZ vaccination. In 2017, RZV was approved by the Food and Drug Administration for adults without contraindications who are 50 years of age or older. Soon after, the ACIP adopted this more efficacious vaccine as the preferred one for this age group.
      • Dooling K.L.
      • Guo A.
      • Patel M.
      • et al.
      Recommendations of the Advisory Committee on Immunization Practices for use of herpes zoster vaccines.
      The American Academy of Ophthalmology also recommends RZV in accordance with the Food and Drug Administration approval.
      Policy statement: recommendations for herpes zoster vaccine for patients 50 years of age and older.
      Zoster vaccine live coverage historically has been low, with only 0.9% to 31.8% coverage by state in 2014 and 31% nationwide in 2015.
      • Williams W.W.
      • Lu P.J.
      • O’Halloran A.
      • et al.
      Surveillance of vaccination coverage among adult populations-United States, 2015.
      ,
      • Lu P.J.
      • O’Halloran A.
      • Williams W.W.
      • Harpaz R.
      National and state-specific shingles vaccination among adults aged ≥60 years.
      Coverage data for RZV is not yet known because of its recent approval. Given the changing epidemiologic trends and risks associated with HZO, it is crucial to continue advocating for HZ vaccination in eligible adults.
      • Cohen E.J.
      Prevention of herpes zoster: we need to do better.
      In addition, further study into whether HZ vaccination is indicated in individuals younger than 50 years is needed.
      This study demonstrated that females and white patients are at higher risk of developing HZO. Through 2007, HZO IRs increased for both males and females older than 10 years. Since 2008, the IRs have been declining in the youngest and oldest age groups for both males and females. A previous study from Hawaii failed to find evidence of a difference in HZO incidence by gender but did not stratify by age. The same study also showed that HZO was more common among non-Pacific Islanders but did not differentiate further by race and ethnicity.
      • Borkar D.S.
      • Tham V.M.
      • Esterberg E.
      • et al.
      Incidence of herpes zoster ophthalmicus: results from the Pacific Ocular Inflammation Study.
      The risk of HZO by gender and race aligns with studies on HZ that found that women and white patients are at higher risk.
      • Kawai K.
      • Yawn B.P.
      • Wollan P.
      • Harpaz R.
      Increasing incidence of herpes zoster over a 60-year period from a population-based study.
      ,
      • Johnson B.H.
      • Palmer L.
      • Gatwood J.
      • et al.
      Annual incidence rates of herpes zoster among an immunocompetent population in the United States.
      ,
      • Yawn B.P.
      • Saddier P.
      • Wollan P.C.
      • et al.
      A population-based study of the incidence and complication rates of herpes zoster before zoster vaccine introduction.
      Herpes zoster ophthalmicus rates were found to be highest in the Northeastern United States and lowest in the West, which may have some relation to ZVL coverage. A 2014 analysis demonstrated that vaccination rates were the second lowest in the Northeast at 30.3% and highest in the West at 37.4%.
      • Lu P.J.
      • O’Halloran A.
      • Williams W.W.
      • Harpaz R.
      National and state-specific shingles vaccination among adults aged ≥60 years.
      Little information is available on the role of geographic location and risk of HZ or HZO, although some reports have indicated a possible association between ultraviolet exposure and HZ risk.
      • Zak-Prelich M.
      • Borkowski J.L.
      • Alexander F.
      • Norval M.
      The role of solar ultraviolet irradiation in zoster.
      The correlation between HZO and geographic region seen in this study is hypothesis generating and requires further study.
      This study has certain limitations. The observational nature of this analysis complicates any measures of causation from being made in relation to the trends seen in HZO incidence, given potential unmeasured confounders. We are not able to compare HZO incidence between the periods before and after the introduction of the varicella vaccine in 1996, which would have provided additional clarification on changing HZO patterns. Analyses were conducted at the level of aggregate data and cannot be used to infer individual-level changes in risk. Different choices of cut point for the broken-stick regression or the use of more general polynomial trend models would result in modest differences in estimated HZO IR slopes. Statistical estimates of trends in the youngest age groups are less reliable because of the small numbers of cases. Owing to the focus on patients in commercial healthcare programs, Medicare Part D and Medicare Advantage, the presented data may underestimate the true incidence of HZO if some patients left the healthcare plan at 65 years of age. However, this would have been the case throughout the study period and would not affect the ability to detect a change in incidence. The database also does not account for individuals who are either uninsured or have other forms of insurance. However, OLDW is relatively generalizable to the United States population given the demographic similarities to the United States Census. Coding for HZO diagnosis has been found to be highly accurate, but undercoding for HZO or changes in health-seeking behavior could affect results.
      • Pimentel M.A.
      • Browne E.N.
      • Janardhana P.M.
      • et al.
      Assessment of the accuracy of using ICD-9 codes to identify uveitis, herpes zoster ophthalmicus, scleritis, and episcleritis.
      This study counted a single code for HZO as an incident diagnosis if a patient met the enrollment eligibility criteria. This criterion was used to include patients who may not have returned for a subsequent evaluation after their initial diagnosis of HZO, which may occur particularly when there is no intraocular involvement. Overall, such limitations would be unlikely to account fully for the magnitude of change seen in the HZO IR.
      In conclusion, after the ACIP recommended ZVL in 2008, HZO rates have declined in the youngest and oldest age groups while continuing to increase among individuals between 31 to 60 years of age. Given the potential shift in HZO burden toward middle-aged individuals, it is crucial for clinicians to support vaccination efforts for individuals 50 years of age and older. These results also raise the question of whether HZ vaccine recommendations should be re-evaluated for individuals in younger age groups.

      Acknowledgments

      The authors thank Nina Veeravalli, OptumLabs, for extensive help with review of the data.

      References

        • Ragozzino M.W.
        • Melton L.J.
        • Kurland L.T.
        • et al.
        Population-based study of herpes zoster and its sequelae.
        Med. 1982; 61: 310-316
        • Liesegang T.J.
        Herpes zoster ophthalmicus: natural history, risk factors, clinical presentation, and morbidity.
        Ophthalmology. 2008; 115: 3-12
        • Weinberg J.M.
        Herpes zoster: epidemiology, natural history, and common complications.
        J Am Acad Dermatol. 2007; 57: 130-135
        • Yawn B.P.
        • Wollan P.C.
        • St. Sauver J.L.
        • Butterfield L.C.
        Herpes zoster—eye complications: rates and trends.
        PLoS One. 2013; 32: 736-740
        • Nithyanandam S.
        • Stephen J.
        • Joseph M.
        • Dabir S.
        Factors affecting visual outcome in herpes zoster ophthalmicus: a prospective study.
        Clin Exp Ophthalmol. 2010; 38: 845-850
        • Vrcek I.
        • Choudhury E.
        • Durairaj V.
        Herpes zoster ophthalmicus: a review for the internist.
        Am J Med. 2017; 130: 21-26
        • Tran K.D.
        • Falcone M.M.
        • Choi D.S.
        • et al.
        Epidemiology of herpes zoster ophthalmicus recurrence and chronicity.
        Ophthalmology. 2016; 123: 1469-1475
        • Harpaz R.
        • Leung J.W.
        The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among older adults.
        Clin Infect Dis. 2019; 69: 341-344
        • Harpaz R.
        • Leung J.W.
        The epidemiology of herpes zoster in the United States during the era of varicella and herpes zoster vaccines: changing patterns among children.
        Clin Infect Dis. 2018; 404
        • Wolfson L.J.
        • Daniels V.J.
        • Altland A.
        • et al.
        The impact of varicella vaccination on the incidence of varicella and herpes zoster in the United States: updated evidence from observational databases, 1991–2016.
        Clin Infect Dis. 2019 Apr 24; (pii: ciz305) ([Epub ahead of print])https://doi.org/10.1093/cid/ciz305
        • Goldman G.S.
        Cost-benefit analysis of universal varicella vaccination in the U.S. taking into account the closely related herpes-zoster epidemiology.
        Vaccine. 2005; 23: 3349-3355
        • Chan A.Y.
        • Conrady C.D.
        • Ding K.
        • et al.
        Factors associated with age of onset of herpes zoster ophthalmicus.
        . 2015; 34: 535-540
        • Kawai K.
        • Yawn B.P.
        • Wollan P.
        • Harpaz R.
        Increasing incidence of herpes zoster over a 60-year period from a population-based study.
        Clin Infect Dis. 2016; 63: 221-226
        • Davies E.C.
        • Pavan-Langston D.
        • Chodosh J.
        Herpes zoster ophthalmicus: declining age at presentation.
        Br J Ophthalmol. 2016; 100: 312-314
        • Optum Labs
        Optum Labs and Optum Labs Data Warehouse (OLDW) Descriptions and Citation.
        Optum Labs, Cambridge, MA2018
        • Optum Labs
        Optum Research Data Assets.
        Optum Labs, Cambridge, MA2014 (Accessed 20.09.19)
        • Sen A.
        • Srivastava M.
        Regression Analysis: Theory, Methods and Applications.
        Springer, Berlin, Heidelberg1990
        • Harpaz R.
        • Ortega-Sanchez I.R.
        • Seward J.F.
        Prevention of herpes zoster: recommendations of the Advisory Committee on Immunization Practices (ACIP).
        MMWR Morb Mortal Wkly Rep. 2008; 57: 1-43
        • Weinmann S.
        • Naleway A.L.
        • Koppolu P.
        • et al.
        Incidence of herpes zoster among children: 2003–2014.
        Pediatrics. 2019; 144
        • Williams W.W.
        • Lu P.J.
        • O’Halloran A.
        • et al.
        Surveillance of vaccination coverage among adult populations-United States, 2015.
        MMWR Surveill Summ. 2017; 66: 7
        • Civen R.
        • Chaves S.S.
        • Jumaan A.
        • et al.
        The incidence and clinical characteristics of herpes zoster among children and adolescents after implementation of varicella vaccination.
        Pediatr Infect Dis J. 2009; 28: 954-959
        • Donahue J.G.
        • Kieke B.A.
        • Gargiullo P.M.
        • et al.
        Herpes zoster and exposure to the varicella zoster virus in an era of varicella vaccination.
        Am J Public Health. 2010; 100: 1116-1122
        • Reynolds M.A.
        • Chaves S.S.
        • Harpaz R.
        • et al.
        The impact of the varicella vaccination program on herpes zoster epidemiology in the United States: a review.
        J Infect Dis. 2008; 197: S224-S227
        • Dooling K.L.
        • Guo A.
        • Patel M.
        • et al.
        Recommendations of the Advisory Committee on Immunization Practices for use of herpes zoster vaccines.
        MMWR Morb Mortal Wkly Rep. 2018; 67: 103-108
      1. Policy statement: recommendations for herpes zoster vaccine for patients 50 years of age and older.
        Ophthalmology. 2018; 125: 1813-1816
        • Lu P.J.
        • O’Halloran A.
        • Williams W.W.
        • Harpaz R.
        National and state-specific shingles vaccination among adults aged ≥60 years.
        Am J Prev Med. 2017; 52: 362-372
        • Cohen E.J.
        Prevention of herpes zoster: we need to do better.
        JAMA Ophthalmol. 2013; 131: 396-398
        • Borkar D.S.
        • Tham V.M.
        • Esterberg E.
        • et al.
        Incidence of herpes zoster ophthalmicus: results from the Pacific Ocular Inflammation Study.
        PLoS One. 2017; 32: 736-740
        • Johnson B.H.
        • Palmer L.
        • Gatwood J.
        • et al.
        Annual incidence rates of herpes zoster among an immunocompetent population in the United States.
        BMC Infect Dis. 2015; 15: 1-5
        • Yawn B.P.
        • Saddier P.
        • Wollan P.C.
        • et al.
        A population-based study of the incidence and complication rates of herpes zoster before zoster vaccine introduction.
        Mayo Clin Proc. 2007; 82: 1341-1349
        • Zak-Prelich M.
        • Borkowski J.L.
        • Alexander F.
        • Norval M.
        The role of solar ultraviolet irradiation in zoster.
        Epidemiol Infect. 2002; 129: 593-597
        • Pimentel M.A.
        • Browne E.N.
        • Janardhana P.M.
        • et al.
        Assessment of the accuracy of using ICD-9 codes to identify uveitis, herpes zoster ophthalmicus, scleritis, and episcleritis.
        JAMA Ophthalmol. 2016; 134: 1001-1006