University of Southern California Dissertations and Theses

Three essays on modifiable determinants of shingles: risk factors for shingles incidence and factors affecting timing of vaccine uptake

(USC Thesis Other)

Three essays on modifiable determinants of shingles: risk factors for shingles incidence and factors affecting timing of vaccine uptake

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THREE ESSAYS ON MODIFIABLE DETERMINANTS OF SHINGLES
RISK FACTORS FOR SHINGLES INCIDENCE AND
FACTORS AFFECTING TIMING OF VACCINE UPTAKE

By
Hyewon Kang

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GERONTOLOGY)
December 2021

Copyright 2021 Hyewon Kang

ii

ACKNOWLEDGEMENTS
I am very pleased to finish my journey to obtain a Ph.D. degree. It started with my curiosity into
the unknown and a vague dream of becoming a scholar. As I wrap up the Ph.D. program, I consider
getting a degree as a commitment toward life in the face of recurrent challenges beyond obtaining
knowledge. Some people may say the nature of academics is dry, but I have learnt the true meaning
of courage, perseverance, growth, care, and love in the program. There were moments I could not
pursue a Ph.D. degree any longer in my strength, but I have been surrounded by people who hold
my hands and push me to take the next step. I want to acknowledge my mentors, family, and
friends who have made this dissertation possible.
My deepest gratitude is to my chair and mentor, Dr. Eileen Crimmins. She has been dedicated to
developing my potentials and abilities. Thank you for teaching me how to think a big picture and
solve research questions in every steps. She has been patiently reading a piece of my writing
repeatedly and guiding me into the best empirical analysis for the questions I want to address. I
often felt frustrated with my progress given her generous dedication and mentorship toward me.
She was supportive and accessible whenever I struggled to balance between family and my
research in major life events such as birth of children, surgery, and remote working in the context
of COVID-19 pandemic. I frequently received a question about the secrets of surviving in a Ph.D.
program along with three little kids. My answer is simple. My faculty advisor is Dr. Crimmins,
who shows a life-long commitment to Geroscience who stands on mountains and stormy seas. The
only thing I had to do is just to follow Dr. Crimmins who already laid the groundwork for young
scholars.
With respect to this, I am grateful to Dr. Jennifer Ailshire. Since I met her in the meeting to ask
for a research assistant position, she has helped me become an independent social scientist. Dr.
iii

Ailshire always paid attention to my ideas and analysis and helped how my research could
contribute to the field by addressing the unanswered questions. Especially, her heartful care has
been a great social support to move my research projects forward. I will be transition to a next
stage of my career armed with skills in data management I learned from your projects. Dr. Jungki
Kim, thank you for your unfailing support throughout my Ph.D. program. I relied on you for
professional and personal guidance. Dr. Jessica Ho, thank you for sharing your keen understanding
of complex data and new perspectives on my projects. I deeply value the advice and
encouragement you have given me when I felt less confident.
Lastly, everything from the beginning to the end of my Ph.D. degree completion would have been
impossible without my life companion, Yeokwang An. I am proud of our academic and personal
achievements while we struggle to raise three children. I hope we trust and support each other in
the up and down life. Thank you for my triple-Js (Jisu, Jiho, and Jia) coming to me. I often felt
sorry for being overwhelmed with many duties as a mom, wife, and doctoral student, but you have
been grown like twinkling stars. Please remember that mommy with a career loves each of you
very much. I also appreciate the support from family members in Korea who pray for me not to
give up on this tough journey.

iv

TABLE OF CONTENTS
ACKNOWLEDGEMENTS……………………………………………………………………….ii
LIST OF TABLES………………………………………………………………………………..vi
LIST OF FIGURES……………………………………………………………………………...vii
ABSTRACT……………………………………………………………………………………viii
CHAPTER 1. INTRODUCTION TO DISSERTATION…………………………………………1
REFERENCES………………………………………………………………………..…7
CHAPTER 2. STRESS: A RISK FACTOR FOR SHINGLES AMONG OLDER ADULTS.….11
INTRODUCTION……………………………………………………………………….12
METHODS………………………………………………………………………………17
RESULTS……………………………………………………………………………..…22
DISCUSSION……………………………………………………………………………23
REFERENCES…………………………………………………………………………..29
TABLES/FIGURES……………………………………………………………………...35
CHAPTER 3. AGE, CHRONIC DISEASES, CRP: AN INTEGRATED APPROACH FOR
INCIDENCE OF SHINGLES AMONG OLDER AMERICANS………………………………38
INTRODUCTION……………………………………………………………………….39
METHODS………………………………………………………………………………41
RESULTS………………………………………………………………………………..45
DISCUSSION……………………………………………………………………………47
REFERENCES…………………………………………………………………………..52
TABLES/FIGURES……………………………………………………………………...57
CHAPTER 4. SHINGLES VACCINE UPTAKE AMONG OLDER ADULTS: IDENTIFYING
EARLY, LATE, AND NON-ADOPTERS………………………………………………………62
INTRODUCTION……………………………………………………………………….63
METHODS………………………………………………………………………………66
RESULTS………………………………………………………………………………..70
v

DISCUSSION……………………………………………………………………………72
REFERENCES…………………………………………………………………………..77
TABLES/FIGURES…………………………………………………………………..….82
CHAPTER 5. CONCLUSION…………………………………………………………………..84
REFERENCES…………………………………………………………………………..89

vi

LIST OF TABLES
Table 2.1 - Descriptive Statistics for 28,727 person-interval observations, Health and Retirement
Study, 2006-2016……………………………..………………………………………………….35
Table 2.2 - Discrete-time Logistic Regression Models Predicting Risks for Incidence of Shingles
associated with Chronic Problems and Stressful Life Events………………..…………………..37
Table 3.1 - Descriptive Statistics for the Full Sample and by Age groups at Baseline in the
2010/2012 Health and Retirement Study …………………………………………………………57
Table 3.2 - Discrete-time Logistic Regression Models Examining the Associations between Age,
Number of Chronic Diseases, CRP, and Incidence of Shingles, 2010-2018 HRS………………59
Table 3.3 - Discrete-time Logistic Regression Models Examining the Associations between Age,
Discrete Chronic Diseases, CRP, and Incidence of Shingles, 2010-2018 HRS…………………60
Table 4.1 - Descriptive Statistics for the Full Sample and by Vaccine Adopter Category, Health
and Retirement Study, 2006-2016……………………….………………………………………82
Table 4.2 - Multinomial Logistic Regression Models Predicting Early Vaccine Adopters and
Late Vaccine Adopters…………………………………………………………………………...83

vii

LIST OF FIGURES
Figure 2.1 - Shingles Incidence Rate by Stress Measures……………………………………….36
Figure 3.1 - Incidence Shingles Rate by Age Groups Between 2010/2012 and 2016/2018 ……..58

viii

ABSTRACT
This dissertation examines multifaceted factors affecting older adults’ shingles experience
that can be prevented and controlled by identifying risk factors for incidence of shingles and factors
affecting the timing of shingles vaccine uptake.
A growing literature has linked psychological and biomedical conditions in old age to
shingles, offering potential mechanisms through which older adults experience the disease.
However, prior studies at the population level fail to appropriately conceptualize and measure
psychological stress, leading to inconsistent findings. Indeed, previous research overlooks that age
and age-related health conditions are highly interrelated, providing a limited insights on the
relative importance of a set of biological and morbid conditions associated with age on incidence
of shingles. Therefore, it remains unclear how a broad range of age-related psychological,
biological, and morbid characteristics is associated with the incidence of shingles. Further, this
dissertation recognizes that the dichotomous view on shingles vaccine uptake or not in previous
literature has limitations in examining the heterogeneity within vaccine recipients. Understanding
factors affecting early or late vaccine uptake is critical to tackling slow dissemination of shingles
vaccination in older adults.
This dissertation provides important epidemiological and public health findings in a
nationally representative sample of older adults, which significantly contribute to shingles
literature. The first study examines the associations between reported stress and shingles incidence
by incorporating event-based and chronic stress from multiple life domains. Research suggests
that problems reflecting chronic stress were associated with the incidence of shingles, whereas
stressful events were not associated with shingles onset. The results demonstrate that psychological
stress triggers shingles, and types of stress differentially operate in shingles risks. The second study
ix

integrates age and age-related health conditions, including chronic diseases and chronic
inflammation, to observe its relative contribution to the incidence of shingles. We found that age,
chronic disease, and chronic inflammation are significantly associated with increased risk of
shingles. However, chronic diseases account for age differentials in shingles risk in the model that
combined age, chronic disease, and chronic inflammation. The findings imply that age-related
physiological and morbid conditions are the main drivers of increased risk for shingles incidence
in older adults, which challenges the interpretation of age as a strong risk factor for shingles. The
third study examines how individual and areal-level factors affect the acceptance and timing of
shingles vaccination over ten years. Findings suggest significant differences between early and
late vaccine recipients in terms of socio-demographic, health care, personality, and geography
factors.
This dissertation lays an empirical foundation for preventing and controlling factors that
affect older adults’ likelihood of getting shingles. The adverse effect of chronic life problems on
the incidence of shingles in this work calls for a better understanding of the conditions where older
adults live and age, in addition to stress management. Considering chronic disease and chronic
inflammation are common aging conditions, health promotion has the potential to retard shingles
onset among older adults. Additionally, this research uncovers the disparity in early shingles
vaccine uptake by race/ethnicity, socioeconomic factors, and geography, which should be
acknowledged in vaccination promotion strategies for a newer shingles vaccine.

1

Chapter 1. Introduction
Shingles (Herpes Zoster) is a disease that occurs more frequently in the older population
and is a burden for patients and the health care system. Understanding risk factors and improving
a shingles vaccine uptake is essential for a better prevention of shingles among older population.
While researchers have identified age, being female, being white, and having immunosuppressive
health conditions as risk factors (Marra et al., 2020), psychological stress and comorbid conditions
that are prevalent in old age have not been successfully evaluated with shingles incidence. A
vaccine has been available for older adults as a preventive measure since 2006; and a more
effective two-dose vaccine was introduced in 2017. An increase in shingles vaccine uptake is one
of the goals of Healthy People 2020, but the uptake has been low. This dissertation is designed to
uncover the determinants of shingles risk, including psychosocial factors, age and related health
conditions (chronic disease and inflammation), and the factors associated with the acceptance and
the timing of the vaccination. This dissertation focuses on a timely and important subject for
preventing shingles among the older population.
Incident Shingles and Risk Factors
A recent review systematically reviewed risk factors for shingles, which include age,
gender, race/ethnicity, family history, comorbidities, physical trauma, psychological stress, and
statin use (Kawai & Yan, 2017). The direction of several factors such as psychological stress and
chronic diseases remain inconsistent (Kawai & Yawn, 2017), where methodological limitations
play a role.
Previous literature often determined risk factors by examining variables related to the
prevalence of shingles at a particular time-point to suggest possible associations between risk
2

factors and shingles or generate a new hypothesis (Marra et al., 2020; Kawai & Tawn, 2017), but
associations with prevalence cannot determine the temporal relationship between these factors and
shingles. Another frequent analytic approach uses a case-control design where the risk factors for
shingles incidence are retrospectively compared with matched individuals who did not develop the
condition (Marin et al., 2016; Schmader et al., 1990; Lasserre et al., 2012). However, confounding
factors and recall bias are barriers to establishing the relationship between potential triggers and
shingles risks in case-control studies (Song & Chung, 2010). Risk factors for the incidence of
shingles can be better identified with a longitudinal study, which allows us to examine temporal
patterns between risk factors and the onset of shingles.
Vaccine Uptake and Its Timing
A shingles vaccine is an effective preventive medicine to protect older adults against
shingles. However, the shingles vaccination rate was 33% in adults aged 60 and older in 2016,
leaving two-thirds of older adults unvaccinated after ten years of vaccine availability (Dooling et
al., 2018). Slow dissemination of shingles vaccines and other vaccines recommended for older
adults has become public health agenda in the U.S. (Gierke et al. 2021). Evaluating older adults’
vaccination progress identifies three distinct groups of early, late, and non-vaccine adopters. Delay
or avoidance of vaccine uptake can increase unnecessary disease risk, yet, to our knowledge, no
studies to date have examined the timing of vaccination of any kind, including shingles vaccine
among the older population. Promoting vaccine uptake through targeted approaches for those who
delay or avoid a vaccine will improve population health free from shingles.
In the following sections, I outline the content of this dissertation and its contributions to
the literature.
3

Stress: A Risk Factor for Shingles among Older Adults
Some studies have reported that those who have experienced stress are more susceptible to
shingles (Schmader et al., 1990; Marin et al., 2016; Lasserre et al., 2012); but adverse effects of
stress were not found in all studies (Schmader et al., 1998 a,b; Harpaz et al., 2014). Inconsistent
findings for stress are mainly attributable to three methodological limitations in the literature. First,
studies are based on varying definitions of stress. Research has found that adverse life events and
chronic stressors are associated with weakened immune functioning (Segerstrom et al., 2004),
suggesting a linkage with shingles. However, previous studies primarily examined acute and
event-based stressors with less attention to chronic stress (Marin et al., 2016; Lasserre et al., 2012;
Harpaz et al., 2014). Although a few studies have examined both event-based and chronic stress,
the scope of chronic stress measures has been limited to the interpersonal life domain (Schmader
et al., 1998 a,b). Stress literature emphasizes that it is important to incorporate sources of stressors
from multiple life domains including work, health, finances, and housing to better approximate
total burden (Turner, Weaton, & Lloyd, 1995). Therefore, the lack of differential types and reduced
scope of stressors in existing work are barriers to understanding the genuine relationship between
stress and shingles. Secondly, the stress assessment window is sometimes not entirely reflective
of the timescale of event-based and chronic stress. Event-based stress is short-term and time-
limited (Perlin, 1989), but intervals between stressful events and the onset of shingles in
prospective population-based studies are often not proximate, but as long as 3 years (Schmader et
al., 1998 a,b). Case-control studies have investigated self-reported stress three to six months prior
to the onset of shingles (Marin et al., 2016, Schmader et al., 1990; Lasserre et al., 2012), but
shingles experience can affect the retrospective reporting of perceived stress. No information is
available for the duration of chronic stress (Schamader 1998 a,b), which makes it difficult to
4

determine whether the risk period for shingles overlaps chronicity of stress. Lastly, prior research
neglected sociodemographic and vaccination factors as potential confounders that can affect stress
level and shingles risks (Schmader et al., 1998a, b; 1990; Marin et al., 2016; Lasserre et al., 2012;
Harpaz et al., 2014), leaving the question of the contribution of psychological stress into incident
shingles unanswered.
To address these limitations, the first chapter of this dissertation examines the associations
between a broad spectrum of stress and shingles incidence by incorporating both event-based and
chronic stress from multiple life domains using data from the 2006-2016 Health and Retirement
Study. Distinguishing stressful events and chronic stressors will improve understanding of how
different types of stressors are associated with shingles incidence. Furthermore, the risk period of
incidence of shingles reflects the timescale of stressful events and chronic stressors. Finally, this
study helps to determine the temporal relationship between reported stress and incidence of
shingles using person-period analytic data.
Age, Chronic Diseases, CRP: An Integrated Approach for Incidence of Shingles among Older
Americans
The most well-established risk factor for shingles is age (Marra et al., 2020). Age-related
decline in immune functions known as immunosenescence is known to be the underlying reason
for older adults’ increased susceptibility to shingles (Oxman et al., 2009). Another major risk
factors include chronic diseases. Studies have found that older adults with a cardiovascular and
respiratory disease, diabetes, cancer, and arthritis frequently experience shingles (Marra et al.,
2020). Although previous literature considers age and chronic diseases as separate potential
triggers of shingles, research in biology of aging suggests that age-related immune system changes
concurrently occur with low-grade, chronic inflammatory process (inflammaging) which underlies
5

the emergence of chronic diseases (Fulop et al., 2016; 2020). Although age-related health
conditions are highly related (Salive, 2012), no studies to date have integrated age, chronic diseases,
and inflammation for prediction of the onset of shingles.
Given the gap in the literature, the second chapter of this dissertation explores the
independent and interdependent role of age groups (50-64, 65-79, 80+), seven chronic conditions
(hypertension, diabetes, cancer, lung disease, heart disease, stroke, and arthritis), and chronic
inflammation measured by CRP as potential risk factors for incident shingles. This novel
integrative approach may shed light on important pathways to incident shingles within the older
population. Additionally, this study is based on the 2010-2018 Health and Retirement Study, which
will allow to calculate national estimates of shingles incidence rates and clarify the directionality
of chronic diseases and shingles.
Shingles Vaccine Uptake among Older Adults: Identifying Early, Late, and Non-adopters
Previous studies mainly examined the differences between shingles vaccine recipients and
non-recipients at one-time point (Lu et al., 2009; 2017, Tseng et al., 2012; Zhang et al., 2017; Shen
et al., 2019), therefore, distinguishing early or late adopters within vaccine recipients has not been
possible. Accelerating shingles vaccine uptake among older adults requires targeted and time-
sensitive approaches.
To address the gap in the literature, the third chapter of this dissertation examines the
factors affecting the acceptance and the timing of vaccination over ten years among older adults
who were first eligible to get vaccinated using the 2006-2016 Health and Retirement Study. We
explore hypotheses that early adopters are more likely to be socioeconomically advantaged,
6

conscientious, recipients of other vaccines, socially active, and reside in areas with higher
vaccination rate than later adopters.
Three studies in this dissertation have significant ramification on the various determinant
of incident shingles and vaccination that can be modified and controlled, which further supports
shingles-free healthy aging in older Americans.

7

Bibliography
Braveman, P., Egerter, S., & Williams, D. R. (2011). The social determinants of health: coming of
age. Annual review of public health, 32.
Crimmins, E. (2020). Social hallmarks of aging: Suggestions for geroscience research. Ageing
Research Reviews, 101136.
Fülöp, T., Dupuis, G., Witkowski, J. M., & Larbi, A. (2016). The role of immunosenescence in
the development of age-related diseases. Revista de investigacion clinica, 68(2), 84-91.
Fulop, T., Larbi, A., Hirokawa, K., Cohen, A. A., & Witkowski, J. M. (2020, September).
Immunosenescence is both functional/adaptive and dysfunctional/maladaptive. In Seminars in
Immunopathology (pp. 1-16). Springer Berlin Heidelberg.
Gierke R., Patricia W., Kobayashi M., (2015). Pneumococcal Disease. Epidemiology and
Prevention of Vaccine-Preventable Diseases.
Harpaz, R., Leung, J. W., Brown, C. J., & Zhou, F. J. (2014). Psychological stress as a trigger for
herpes zoster: might the conventiona l wisdom be wrong?. Clinical Infectious Diseases, 60(5),
781-785.
Harpaz, R., Ortega-Sanchez, I. R., & Seward, J. F. (2008). Prevention of herpes zoster:
recommendations of the Advisory Committee on Immunization Practices (ACIP). Morbidity and
Mortality Weekly Report: Recommendations and Reports, 57(5), 1-30.
Kawai, K., & Yawn, B. P. (2017, December). Risk factors for herpes zoster: a systematic review
and meta-analysis. In Mayo clinic proceedings (Vol. 92, No. 12, pp. 1806-1821). Elsevier.
8

Kawai, K., Yawn, B. P., Wollan, P., & Harpaz, R. (2016). Increasing Incidence of Herpes Zoster
Over a 60-year Period From a Population-based Study. Clinical infectious diseases : an official
publication of the Infectious Diseases Society of America, 63(2), 221–226.
https://doi.org/10.1093/cid/ciw296
Lasserre, A., Blaizeau, F., Gorwood, P., Bloch, K., Chauvin, P., Liard, F., ... & Hanslik, T. (2012).
Herpes zoster: family history and psychological stress—case–control study. Journal of clinical
virology, 55(2), 153-157.
Lasserre, A., Blaizeau, F., Gorwood, P., Bloch, K., Chauvin, P., Liard, F., ... & Hanslik, T. (2012).
Herpes zoster: family history and psychological stress—case–control study. Journal of clinical
virology, 55(2), 153-157.
Lu P-j, Euler GL, Jumaan AO, Harpaz R. Herpes zoster vaccination among adults aged 60 years
or older in the United States, 2007: Uptake of the first new vaccine to target seniors. Vaccine
2009;27(6):882-887.
Lu PJ, O’Halloran A, Williams WW, Harpaz R. National and state-specific shingles vaccination
among adults aged ≥ 60 years. American journal of preventive medicine. 2017 Mar 1;52(3):362-
72.
Marin, M., Harpaz, R., Zhang, J., Wollan, P. C., Bialek, S. R., & Yawn, B. P. (2016, May). Risk
factors for herpes zoster among adults. In Open forum infectious diseases (Vol. 3, No. 3, p.
ofw119). Oxford University Press.
Marin, M., Harpaz, R., Zhang, J., Wollan, P. C., Bialek, S. R., & Yawn, B. P. (2016, May). Risk
factors for herpes zoster among adults. In Open forum infectious diseases (Vol. 3, No. 3, p.
ofw119). Oxford University Press.
9

Marra, F., Parhar, K., Huang, B., & Vadlamudi, N. (2020, January). Risk factors for herpes zoster
infection: a meta-analysis. In Open forum infectious diseases (Vol. 7, No. 1, p. ofaa005). US:
Oxford University Press.
National Center for Health Statistics. Health, United States, 2017: With special feature on
mortality. Hyattsville, MD. 2018.
Oxman, M. N. (2009). Herpes zoster pathogenesis and cell-mediated immunity and
immunosenescence. The Journal of the American Osteopathic Association, 109(6_suppl_2), S13-
S17.
Pearlin, L. I. (1989). The sociological study of stress. Journal of health and social behavior, 241-
256.
Pearlin, L. I., Lieberman, M. A., Menaghan, E. G., & Mullan, J. T. (1981). The stress process.
Journal of Health and Social Behavior, 22, 337–356. doi:10.2307/2136676
Salive, M. E. (2012). Multimorbidity in older adults. Epidemiologic reviews, 35(1), 75-83.
Schmader K, George LK, Burchett BM, Hamilton JD, Pieper CF. (1998) Race and stress in the
incidence of herpes zoster in older adults. Journal of the American Geriatrics Society.
Aug;46(8):973-7.
Schmader K, George LK, Burchett BM, Pieper CF. (1998) Racial and psychosocial risk factors for
herpes zoster in the elderly. The Journal of infectious diseases. Nov 1;178(Supplement_1):S67-70.
Schmader, K., Studenski, S., MacMillan, J., Grufferman, S., & Cohen, H. J. (1990). Are stressful
life events risk factors for herpes zoster?. Journal of the American Geriatrics Society, 38(11), 1188-
1194.
10

Schmader, K., Studenski, S., MacMillan, J., Grufferman, S., & Cohen, H. J. (1990). Are stressful
life events risk factors for herpes zoster?. Journal of the American Geriatrics Society, 38(11), 1188-
1194.
Shen, A. K., Warnock, R., Selna, W., MaCurdy, T. E., Chu, S., & Kelman, J. A. (2019).
Vaccination among Medicare-fee-for service beneficiaries: characteristics and predictors of
vaccine receipt, 2014–2017. Vaccine, 37(9), 1194-1201.
Song, J. W., & Chung, K. C. (2010). Observational studies: cohort and case-control studies. Plastic
and reconstructive surgery, 126(6), 2234.
Tseng HF, Chi M, Smith N, Marcy SM, Sy LS, Jacobsen SJ. Herpes zoster vaccine and the
incidence of recurrent herpes zoster in an immunocompetent elderly population. J Infect Dis
2012;206(2):190-6.
Turner, R. J., Wheaton, B., & Lloyd, D. A. (1995). The epidemiology of social stress. American
Sociological Review, 60, 104–125. doi:10.2307/2096348
Zhang D, Johnson K, Newransky C, Acosta CJ. Herpes Zoster Vaccine Coverage in Older Adults
in the U.S., 2007-2013. Am J Prev Med 2017;52(1):e17-e23

11

Chapter 2. Stress: A Risk Factor for Shingles among Older Adults

Abstract
Objective: The role of stressful events and chronic stress as potential triggers for shingles among
older adults is not clear. This paper examines associations between reported stress and shingles
incidence in the Health and Retirement Study, a nationally representative sample of people over
age 50.
Method: We used data on 28,727 person-interval observations that were derived from interviews
with 11,376 respondents occurring every two years from 2006-2016. Using discrete-time logistic
regression modeling, we examined associations of shingles onset with stressful events and chronic
stressors.
Results: Problems reflecting chronic stress were associated with an increased risk of shingles even
after adjustments for sociodemographic characteristics, health, and vaccination. However,
stressful events were not associated with shingles onset.
Discussion: This study adds to a growing body of research on shingles’ risk factors by
demonstrating the adverse effects of chronic stressors. Older adults’ reserve psychological
capacity and emotion regulation help them better deal with stressful events and minimalize its
adverse effects on shingles.

12

Introduction
Shingles is a painful disease that significantly impacts morbidity and quality of life among
older adults (Harpaz et al., 2008). Stress in later life, mainly stressful events and chronic stress has
been often considered as a potential trigger of shingles as stress leads to weak immune responses
(Segerstrom & Miller, 2004). Research suggests that when immune function declines, the latent
varicella-zoster virus acquired with childhood chickenpox becomes reactivated, causing shingles
(Oxman, 2009).
Despite the biological link between stress and shingles, the adverse effect of stress on
shingles is supported by some studies, not all. There were key limitations in examining the
associations between stress and shingles, which may contribute to mixed results of stress. First,
prior studies included the lack of differential types and reduced scope of stressors (Schmadser et
al., 1998a, b; 1990; Marin et al., 2016; Lasserre et al., 2012; Harpaz et al., 2014), which do not
fully capture older adults’ stress burden derived from stressful transitions and recurrent difficulties
in various life aspects. Second, the stress assessment window is not entirely reflective of the
timescale of event-based and chronic stress, leading to questions on whether shingles is a timely
reflection of stress exposures (Schmadser et al., 1998a, b; 1990; Marin et al., 2016; Lasserre et al.,
2012; Harpaz et al., 2014). Lastly, prior research overlooked confounding factors that may affect
stress levels and shingles risk including sociodemographic and vaccination factors (Schmadser et
al., 1998a, b; 1990; Marin et al., 2016; Lasserre et al., 2012; Harpaz et al., 2014), which can bias
the reported relationships between stress and shingles. Taken together, there remains a gap in the
literature as to conceptualization and measurement of stress. The purpose of this study is to
examine the associations between a broader spectrum of stress indicators and shingles onset by
13

incorporating both stressful events and chronic stressors across multiple life domains in a
nationally representative sample of U.S. older adults
Background
Shingles and Its Correlates
Shingles, a painful disease accompanied by skin rash and blisters, is a viral infection that
results from the varicella-zoster virus (VZV). Greater than 90% of older adults have evidence of
VZV infection and are at risk of shingles in the U.S (Holmes et al., 1996; Harpaz et al., 2008).
After individuals have had childhood chickenpox, VZV maintains a dormant state in the nervous
system, but it becomes active again and triggers shingles when immune function weakens (Oxman,
2009). Shingles causes significant public health and economic burden. An estimated one million
Americans are diagnosed with new shingles cases each year (Insinga et al., 2005), which incurs
$782 million of economic burden (Ozawa et al., 2016). Shingles vaccines are available for older
adults to reduce the risk of getting the disease: the first shingles vaccine was approved in 2006,
and a new and more effective vaccine was introduced in 2017. Risk factors for shingles include
age, being non-Hispanic White, female, having morbidities, or a compromised immune systems
(Thomas & Hall, 2004; Kawai & Yawn, 2017). Psychological stress has been posited as a potential
risk factor for shingles due to stress-induced adverse physiological responses, including immune
dysfunction (Kawai & Yawn, 2017).
Stressful Events and Chronic Stress as Potential Triggers for Shingles
Stress is an inherently complex concept, but the timescale involved in the conceptualization
and measurement of psychological stress suggests that there are two main types of stress: event-
based and chronic stress (Epel et al., 2018). Stress carries mental and physiological health risks
14

(Epel et al., 2019), which is believed to increase risks for shingles through altered immune function
(Segerstrom & Miller, 2004).
Event-based stress is time-limited and occurs when a person encounters undesired,
unscheduled, or uncontrollable life events (Wheaton et al., 2013; Epel et al., 2018). For example,
spousal death is one of the most stressful and common life events in older adults. Following the
traumatic loss of a loved one, the surviving spouse can experience emotional crises such as sorrow,
numbness, and even anger. High acute stress level following stressful events evokes dysregulation
of the hypothalamus adrenal axis and secretion of cortisol, and immunological imbalance
(Michaud et al., 2008; Segerstrom & Miller, 2004). Studies have found evidence of decreased
inflammatory response and reduced T-lymphocytes and natural killer cells among bereaved
spouses a few months after spousal death (Buckley et al., 2009, 2012). These adverse physiological
responses that involve declined immune function are believed to play a role in the reactivation of
dormant VZV in the development of shingles.
Older adults can experience stressful events in various life domains (Turner et al., 1995),
including financial or job loss, relocation, and the onset of spousal disability. Older adults are
vulnerable to financial shock or involuntary job loss due to limited working opportunities and
financial insecurity in old age. For example, during the latest economic recession, older persons
experienced significant wealth loss and suffered from poor mental health (Hawkley et al., 2020;
Zhao & Burge., 2020; Wilkinson et al., 2016). A residential move can be challenging for older
adults because it breaks existing relationships with their home and neighborhood and can change
independence and quality of life (Calvo et al., 2009). Spouses often become primary caregivers
when their partners develop a disability and taking a new role that requires physical and emotional
sacrifice can be stressful (Schulz & Sherwood., 2008).
15

In contrast to event-based stress, chronic stress has no clear start and endpoint (Wheaton
et al., 2013; Epel et al., 2018). Individuals who experience recurrent and uncontrollable stress over
extended period can feel overwhelmed, have difficulty functioning normally, and report mental
exhaustion such as burn-out symptoms (Burton et al., 2003; Epel et al., 2018). Chronic stress can
result from adverse life events if their sequelae and effects continue (Pearlin, 1981). For example,
onset of spousal disability can cause chronic caregiving problems over time. Older adults can
experience chronic stress in a range of life domains: health, housing, finance, and relationships
(Brown et al., 2018). Research suggests that chronic stress produces sustained high levels of stress
hormones and low-grade chronic inflammation (Dhabhar, 2014; Seiler et al., 2020). Continued
physiological strain contributes to less effective immune function needed for controlling viral
dormancy (Seiler et al., 2020). For instance, evidence suggests that caregivers were more
susceptible to experience disease from a reactivated virus (Segerstrom & Miller, 2004; Glaser et
al., 1994).
Although growing research suggests that stress may play a significant role in causing
shingles, we found several limitations in stress conceptualization and measurement in the previous
literature. First, prior research has not captured the total stress burden by primarily focusing on
event-based stressors. For example, most studies have relied on either a count of events perceived
to be stressful (Schmader et al., 1998a, b; 1990) or a checklist of stressful events (Marin et al.,
2016; Lasserre et al., 2012; Harpaz et al., 2014), with limited attention to chronic stress. Even
when chronic stress is considered, the source of chronic stress is often limited to an interpersonal
domain (e.g., a lack of social support) (Schmader et al., 1998a, b). Without attention to multiple
types and sources of stressors that comprise the total stress burden (Wheaton, 1994; Turner et al.,
1995), our understanding of the role of stress on shingles is incomplete.
16

Second, the time window between the reported stress and shingles is often long or not clear,
leading to questions on whether the impact of stress exposure on shingles is captured in a timely
manner. A stressful event is considered an immediate threat that lasts for a short-term period (Epel
et al., 2018), but distant and retrospective reporting of stress in prior studies in prior work has not
been sensitive to capturing the acute impact of event-based stressors. For example, no association
between event-based stressors and incident shingles with as long as a 3-year follow-up period was
observed in a prospective population-based studies (Schmader et al., 1998a, b). Case-control
studies have found harmful effects of self-reported stress three to six months prior to the onset of
shingles, but findings are not free from recall bias as experience of shingles can possibly affect the
reporting of stress (Marin et al., 2016, Schmader et al., 1990; Lasserre et al., 2012). In terms of
chronic stressors, null effects of interpersonal problems on shingles have been reported without
revealing how long the difficulties persisted (Schmader et al., 1998a, b). The stress literature
suggests that problems should have had lasted for six months or more to be considered as chronic
stressors (Epel et al., 2019), therefore missing information on the duration of stressful problems
leads to inconclusive results on chronic stress.
Third, much empirical work has not appropriately included factors that may affect stress
level or shingles risk, which can influence the true association between stress and shingles onset.
Stress literature emphasizes that stress is a process that interacts with individual and social context.
How stress is framed or presented can determine the meaning of stress and produce differential
health outcomes (Pearlin et al., 1981; McLeod, 2012). For example, assets can serve as a useful
stress-buffering resource when an individual loses a job. Prior studies examined stressful
experiences ignoring one's socioeconomic context (Schmader et al., 1990; Marin et al., 2016;
Harpaz et al., 2014), which potentially neglects the process of how stress exposure is internalized
17

in a context of an individual’s resources. Furthermore, factors that affect shingles risks, such as
vaccination, have been overlooked in investigating the association between stress and shingles
despite vaccine availability during the study period (Lasserre et al., 2012; Harpaz et al., 2014).
Lastly, previously reported evidence of the significant association between stress and
shingles is derived from a regional sample or non-US populations (Schmadser et al., 1998a, b;
1990; Marin et al., 2016; Lasserre et al., 2012; Harpaz et al., 2014), which is not generalizable to
the entire older Americans.
Current Study
To address gaps in the literature, this study advances conceptual and methodological
approaches in examining the relationship between stress and shingles by incorporating both
stressful events and chronic stress with a timely sensitive measurement window. Using a nationally
representative sample of older adults, we explore a hypothesis that both stressful events and
chronic stressors are associated with incident shingles, even after controlling for sociodemographic
and health factors and usage of preventive medicine for shingles.
Methods
Data
The data come from the 2006-2016 Health and Retirement Study (HRS), a biannual
nationally representative longitudinal survey of U.S. adults older than 50. The longitudinal design
of the survey allows for tracking major life events, including loss of family, exit from work,
financial loss, residential moves, and the onset of functioning loss and disability. Since 2008, HRS
has asked a series of shingles questions, including a history of shingles, pain, vaccination, and
healthcare-seeking behavior.
18

HRS collects psychological and lifestyle data using a self-administered questionnaire
(SAQ) which is left with participants to return by mail after a face-to-face interview. Two-random
rotating half-samples complete this survey every four years. One-half of participants initially
reported chronic stressors in 2006, with the follow-up data in 2010 and 2014. The other half of
participants first provided the data on chronic stress in 2012 and follow-up data in 2016.
For incidence analysis, we used person-intervals data based on two consecutive waves
where individuals aged 50 and older are at risk of getting new shingles in two-year intervals: 2006-
2008, 2008-2010, 2010-2012, 2012-2014, 2014-2016. We included intervals if respondents who
were alive andhad not had shingles by the beginning of the interval have information available on
shingles event during the interval reported at the end of the interval. There were 29,111
observations with shingles and stress data available. We dropped 384 observations with missing
data on covariates. The final sample size for analyses was 28,727 person-interval observations
from 11,376 persons. Compared to the analytic data, intervals with missing information were more
likely to be drawn from non-Hispanic blacks (6.3% vs 10.6%, p=0.000) and non-shingles vaccine
recipients (83.9% vs 91.4%, p=0.000), but they were similar in age, gender, number of chronic
diseases.
Measures
Incidence of Shingles
The incidence of shingles for each two-year interval over ten-year period was defined by
the response to a question, “Since the previous wave/in the last two years, have you had
shingles/have you ever had shingles?” at each wave from 2008 to 2016. The incidence of shingles
reflects reporting having the disease during the interval after reporting not having had the condition
19

at the beginning of the interval. For example, if a respondent who reported in 2008 “have not ever
had shingles” reported “have had shingles” at the 2010 interview, we consider shingles incident
during the 2008-2010 interval. Because the onward question on shingles status started in 2008, we
relied on reported age at the first shingles among those with a history of shingles to identify
shingles status in 2006-2008. Shingles incidence in the 2006-2008 interval reflects the reporting
of the first shingles in the age at the time of 2-year interval among those who had not had shingles
by 2006 interview. This retrospective and prospective approach in identifying incident shingles is
based on the same assumption that at the beginning of an interval not having had the condition.
Incidence of shingles in our study reflects not only includes the first occurrence of shingles but
also recurrent cases because we include intervals as long as individuals not having had the disease
reported at the beginning of intervals (time t) and at risk of having a new shingle in two-year
intervals (t+2)
Chronic Stressors
We created a summative chronic stressor measure by adding respondents’ reported chronic
problems in eight life domains. Participants were asked whether any of the following were current
and ongoing problems that had lasted 12 months or longer; (1) respondent’s health problems, (2)
physical and emotional problems of a spouse or child, (3) problems with alcohol or drug use in a
family member, (4) difficulties at work, (5) financial strains, (6) housing problems, (7) problems
in a close relationship, (8) providing help to at least one sick, limited, or frail family member or
friend. The summative chronic problems measure ranges from 0 to 8, with a higher value
representing more chronic stressors.
20

Due to the 4-year follow-up period for HRS psychosocial data collection, chronic stress
data were not available in some years. We estimate chronic stressors in missing years as the
average value of the preceding and following interviews. For example, chronic stress in 2008 is
estimated to be the mean of reported chronic stress in 2006 and 2010. With this approach, we are
able to assign ongoing chronic stressors at each wave for the two sub-samples. Specifically, one-
half sample has ongoing stressors reported in 2006, 2010, and 2014, which are interpolated to
provide chronic stressors in 2008 and 2012. The other half sample has ongoing stressors reported
in 2012 and 2016, which are interpolated into chronic stressors in 2014.
Stressful Life Events
We counted the number of stressful events indicated by the following five life transitions:
spousal loss, involuntary job loss, residential moves, negative wealth shock, and the spousal onset
of disability. The occurrence of stressful life events over the 2-year interval is tracked by
comparing the status reported at the beginning and the end of the interval or by reports of events
between waves. The experience of a stressful event is indicated by a dichotomous variable coded
1 for those who experienced a stressful life event and 0 for those who did not experience an event.
Spousal loss occurs with divorce or widowhood among the married. Involuntary job loss refers to
losing a job due to business closure or lay-off. Adverse wealth shocks are defined as a loss of 75%
or more of total wealth among those who have had positive wealth (Pool et al., 2017). A residential
move is defined as a change in main residence. The onset of a spousal functioning difficulty
includes the development of at least one difficulty in activities of daily living (ADL) (i.e. walking
across a room, dressing, bathing, eating, getting into bed, and toileting) among those with no ADL
limitation at the beginning of the interval. The summative measure of stressful events ranges from
0 to 5, with a higher value representing more event-based stressors.
21

Covariates
We include time-invariant sociodemographic factors. Gender is indicated as female or male
(reference). Race/ethnicity is categorized as non-Hispanic whites (reference), non-Hispanic blacks,
Hispanics, and others. Education is measured with years of education. We included several time-
varying factors. Age as years and number of self-reported physician-diagnosed health conditions,
including hypertension, diabetes, cancer, heart diseases, lung disease, arthritis, stroke, were
updated at baseline of each interval. As the shingles vaccine has been available during the study
period, we identified vaccination status (not-vaccinated vs vaccinated) at the baseline for each
interval.
Statistical Methods
We use a discrete-time logistic regression modeling approach to assess the influences of
stressors on the development of shingles. For the analysis, we created person-interval record files
from 2006 through 2016 where anyone at risk of shingles incidence at two-year intervals can
contribute to observations. While person-level analysis often encounters selection bias due to loss
to follow-up over a long period (Howe et al., 2016), the person-interval analysis enables to use of
all possible two consecutive waves to estimate the relationship between stress and shingles. To
capture the stress effect within the proximate timeline of shingles, we associated the occurrence of
a stressful event with the incidence of shingles in the same interval. In addition, we associated
chronic problems reported at the beginning of each interval and the incidence of shingles in the
interval over two years.
In the regression modeling, we first associated shingles onset with the two summative
indices of chronic and event-based stressors with basic controls for age and gender. Then, we add
22

factors that are known to be associated with stress and shingles, such as race/ethnicity, education,
vaccination, and comorbid conditions. All analyses were weighted using sample weights to
produce estimates representative of the older U.S. population. Since multiple observations could
come from the same individual, we employed a multilevel model adjusted for standard errors.
Results
Table 1 presents the weighted descriptive statistics for the 28,727 person-intervals in HRS
2006-2016. The mean age in the data was 67.5 years (range: 50 – 104). Women made up more
than half of the observations (55.4%). The majority of observations were from whites (86.2%)
followed by blacks (6.3%), Hispanics (5.1%), and others (2.3%). The mean years of education was
13.4 years (range 0 – 17) and the number of chronic diseases was 1.9 (range 0 – 7). On average,
16% of observations had a history of vaccination.
When looking at the chronic stressors, the mean number experienced in an interval was 2.4
(range 0 – 8). Chronic stressors are prevalent in old age, with 89% of intervals exposed to a chronic
stressor. The most common stressors were health (69.1%), physical or emotional problems in
spouse/children (45.5%), and finances (43.7%). A housing problem was the least commonly
reported problem (14.5%). The mean number of stressful events in the 2-year intervals was 0.19.
Stressful late-life events happened in 17.3% of the observations. The most common stressful event
was negative wealth shock (6.0%), followed by a residential move (4.5%). The least common
event was job loss, reported by 2.5%.
Figure 1 shows the weighted shingles incidence rate by stress measures. There were 800
incident shingles cases in 28,727 person-intervals. The overall incidence rate in those aged 50 and
older is 26.2 per 1,000 person-intervals (95% CI: 24.3, 28.0). When comparing the mean of
23

reported chronic stressors and incident shingles over two years, the incident shingles rate was
higher in intervals with more reported stressors problems. The lowest rate of new shingles cases
was identified among intervals with zero to two chronic stressors (24.4 per 1,000 per person-
intervals), and the highest rate was among observations with more than five chronic problems
(31.0 per 1,000 per person-intervals). A similar trend occurred for stressful life events, with the
lowest shingles incidence rate linked to zero stressful event (25.6 per 1,000 person-intervals) and
the highest shingles incidence rate related to two or more events (39.2 per 1,000 person-intervals).
Table 2 presents the results of the discrete logistic regression models predicting incident
shingles by measures of stress. Model 1 shows the associations between the two summative indices
of stress and incident shingles with controls for age and gender. Findings indicate that having more
chronic stressors was significantly associated with a 7% increased risk of shingles onset (OR=1.07,
95% CI:1.02, 1.12), whereas experiencing more stressful events was not associated with incident
shingles (OR=1.09, 95% CI:0.92, 1.30). Model 2 added race/ethnicity, education level, comorbid
health conditions, and vaccination. The significant association of chronic stressors on shingles
remained even after adjusting for covariates (OR=1.06, 95% CI:1.005, 1.11). Our result suggests
that experiencing chronic stressors, but not stressful events, predicts incident shingles even when
we take into account other factors that might affect shingles onset.
Discussion
This is the first study to examine the associations between reported stress and shingles
onset by incorporating both stressful events and chronic stressors from multiple life domains in a
nationally representative sample of U.S. older adults. We found that problems reflecting chronic
stress were associated with an increased risk of shingles even after controlling for
sociodemographic and health factors and vaccination. The impact of chronic stressors and onset
24

of shingles was equivalent to three additional years of age. Considering age as the most well-
established risk factor for shingles (Thomas & Hall, 2004; Kawai & Yawn, 2017), chronic stress
appears to be a critical pathway in which older adults can develop shingles. While event-based
stressors have been largely employed in shingles literature (Schmadser et al., 1998a, b; 1990;
Marin et al., 2016; Lasserre et al., 2012; Harpaz et al., 2014), the results of stressful events were
mixed. With a prospective and timely-sensitive stress assessment window, our findings suggest
that stressful events were not significantly associated with shingles onset. This paper is innovative
in that we demonstrate differential effects of event-based and chronic stressors on the incidence of
shingles in reflecting the unique timeframe of stressors.
Our study found that the shingles incidence rate in those aged 50 and older is 26.2 per 1,000
person-intervals (95% CI: 24.3, 28.0). Since the interval in our study covers two years, the rate is
translated to 13.1 per 1,000 person-years. The incidence rate in our data is slightly higher than 8.5-
9.9 per 1,000 person-year reported from persons aged 50 and older in health administrative data
Johnson et al., 2015; Tseung et al., 2020). The gap may be attributable to differential health care
access and use behaviors and definition of incident shingles. We confirmed other risk factors for
shingles, such as age, being female or being non-Hispanic white, and having a higher number of
chronic conditions, which were reported in previous studies (Thomas & Hall, 2004; Kawai &
Yawn, 2017). The null effect of vaccination on the incidence of shingles in our study may be
attributable to two factors. This study covers ten years of shingles vaccine availability and the time
since vaccination has increased and the effectiveness of the vaccine waned. In addition, some
individuals received the vaccine after having had a shingles episode.
Experiencing chronic stressors in later life is common. Older adults, on average, had two
chronic problems in eight focal life domains and exposure to health, family, and finance-related
25

problems the most. Prior studies found no associations between chronic stressors and shingles by
measuring a large array of social support indicators, including perceived support, social network,
and social interaction (Schmader et al., 1998a, b). In addition to the limited scope of chronic
stressors, lack of information on the duration of interpersonal problems were barriers to
generalizing findings (Schmader et al., 1998a, b). To overcome the limited scope and unclear
duration of chronic stressors in previous studies, this study identified chronic problems that have
lasted a year or longer from a wide arrange of life domains and found adverse effects of chronic
stress on the incidence of shingles. Evidence of chronic stressors as a risk factor for shingles may
alerts older adults’ health and psychological well-being which has been deteriorated in a time of
COVID-19 pandemic.
We found no significant association between stressful life events and shingles by exploring
a range of stressful events, including spousal loss, involuntary job loss, residential moves, negative
wealth shock, and the spousal onset of disability. Null effects of stressful events on shingles have
been previously reported in some (Schmader et al., 1998 a, b; Harpaz et al., 2015), but not all
studies (Schmader et al., 1990; Lasserre et al., 2012; Marin et al., 2016). A recent study found no
significant associations between a spousal death or catastrophic health event and incident shingles
with up to 4 months of follow-up (Harpaz et al., 2014). Other studies that relied on a standardized
list of late-life events (e.g., Geriatric scale of recent life events) also found no relationship between
the experience of adverse events and shingles with as long as 3-years follow-up period (Schmader
et al.,1998 a, b). On the other hand, a few studies relying on retrospective reporting of stress found
adverse effects of stressful events on shingles (Mairin et al., 2016; Lasserre et al., 2012; Schmader,
1990). For example, a case-control study found that those who developed shingles had events
perceived as stressful more frequently than those without shingles in the two months before
26

shingles onset (Schmader et al., 1990). However, caution is needed in interpreting the results
because self-identified prior stressful events may be subject to recall bias (Raphael, 1987). Because
distant or retrospective stress assessment window challenges to measure the short-term effect of
stressful events on shingles onset (Epel et al., 2019), our study prospectively predicted incidence
of shingles in the same interval where stressful events occurred. Based on time-appropriate stress
exposure, our results demonstrate that there is no significant relationship between stressful events
and shingles.
Differential effects of event-based and chronic stress in developing shingles may be better
understood within the stress process framework (Pearlin et al., 1981; McLeod, 2012). The model
suggests that individuals negotiate and internalize the meaning of stress. Therefore, the same stress
exposure can produce heterogeneous health outcomes depending on how individuals evaluate and
cope with stress (Pearlin et al., 1981; Wheaton et al., 2013). In other words, if an individual’s
capacity and resources are greater than the demands of a stressful situation, a person perceives
stress to be a challenge but can maintain normal functioning in the face of adversity (Blascovich
& Mendes, 2000). On the contrary, stress can be a constant threat if the required efforts overwhelm
an individual’s capacity and resources (Blascovich & Mendes, 2000). While a previous study
found null effects of spousal death and catastrophic illness on the incidence of shingles, stressful
life events were found to increase mental health visits (Harpaz et al., 2014). These findings imply
that stressful events contribute to poor psychological well-being, but older adults may have a
greater capability to counteract negative stimuli that affect shingles risks.
Older adults maintain various psychological resources, such as perceived support,
experience with adversity, knowledge, mastery, which may serve as a reserve capacity to deal with
stressful late-life transitions (Schetter & Dolbier, 2011; Fontes & Neri, 2015). Cumulative
27

psychological resources over the life course may effectively reduce stress levels following
terrifying, traumatic, or shocking life events in old ages. Additionally, older adults are motivated
to maximize positive affect to attain emotionally meaningful goals in later life (Charles &
Carstensen, 2010), which may optimize stress levels. Taken together, it is plausible that older
adults’ reserve psychological capacity and emotion regulation help them better deal with stressful
events and minimalize its adverse effects on shingles.
Although older adults are resilient to stressful transitions and events, it is challenging to
withstand chronic stress because its sources are more likely to be fundamental and constant (Brown
et al., 2018). Enduring and recurrent difficulties often emerge in focal life domains (e.g., health
and finance) and develop over a long-term period which cannot be easily intervened. In reflecting
of social determinants of health framework (Braveman et al., 2011), adverse life conditions where
we live, work, age get under the skin through the mechanism of chronic stress. Our findings shed
light on the importance of older adults’ living conditions that may translate to adverse health
outcomes beyond individual-level stress management.
There are some limitations to this study. First, we assessed the incidence of shingles with
stressful life events occurring within the same 2-year interval. It is possible that some shingles
cases occurred before the onset of stressful events. Because stressful events are acute and short-
term stress, we attempted to capture the proximate impact of stressful events on developing
shingles incidence in a relatively short time frame. Although we have limited information on
temporal ordering, we believe that using 2-year intervals, helps to understand the linkage between
stressful event and shingles that reflects intensity and severity of stressors. Second, this study
measured stress exposures, not stress appraisal. It is possible that people can feel and evaluate
stress events and situations differently, which means we do not have findings on the association
28

between realized stress and the incidence of shingles. Recognizing stress as a dynamic process that
interacts with contextual factors, we controlled for sociodemographic factors. No data is available
for stress appraisal to both stressful events and chronic problems in our data. However, future
research will be needed to evaluate how perceived stress in a wide range of stressors is associated
with the incidence of shingles.
Shingles poses substantial health and economic burden among older adults. As shingles
cases are projected to increase with the growing older population in the future (Kawai et al., 2014),
identifying risk factors for shingles may guide prevention strategies. Although cumulative
evidence suggests a plausible linkage between stress and shingles, stress has not been established
as a risk factor for shingles (Thomas & Hall, 2004; Kawai & Yawn, 2017). The present study
demonstrates that a greater burden of chronic life problems can cause the onset of shingles. Later
life involves unpredictable and inevitable adversities, which can serve as a common pathway to
increase shingles incidence in the older population. Paying special attention to chronic stress as a
trigger of shingles can have profound implications for successful aging, extending additional years
with healthy mental and physical well-being in later life.

29

Bibliography
Blascovich, J., & Mendes, W. B. (2000). Challenge and threat appraisals: The role of affective
cues In Forgas J (Ed.), Feeling and thinking: The role of affect in social cognition (pp. 59–82).
Braveman, P., Egerter, S., & Williams, D. R. (2011). The social determinants of health: coming of
age. Annual review of public health, 32, 381-398.
Brown, L. L., Mitchell, U. A., & Ailshire, J. A. (2018). Disentangling the stress process:
Race/ethnic differences in the exposure and appraisal of chronic stressors among older adults. The
Journals of Gerontology: Series B, 75(3), 650-660.
Buckley, T., Bartrop, R., McKinley, S., Ward, C., Bramwell, M., Roche, D., ... & Tofler, G. (2009).
Prospective study of early bereavement on psychological and behavioural cardiac risk
factors. Internal Medicine Journal, 39(6), 370-378.
Buckley, T., Sunari, D., Marshall, A., Bartrop, R., McKinley, S., & Tofler, G. (2012).
Physiological correlates of bereavement and the impact of bereavement interventions. Dialogues
in clinical neuroscience, 14(2), 129.
Burton, L. C., Zdaniuk, B., Schulz, R., Jackson, S., & Hirsch, C. (2003). Transitions in spousal
caregiving. The Gerontologist, 43(2), 230-241.
Calvo, E., Haverstick, K., & Zhivan, N. A. (2009). Determinants and consequences of moving
decisions for older Americans. CRR WP, 16, 1-38.
Charles, S. T., & Carstensen, L. L. (2010). Social and emotional aging. Annual review of
psychology, 61, 383-409.

30

Dhabhar FS (2014) Effects of stress on immune function: the good, the bad, and the beautiful.
Immunol Res 58:193–210
Epel, E. S., Crosswell, A. D., Mayer, S. E., Prather, A. A., Slavich, G. M., Puterman, E., & Mendes,
W. B. (2018). More than a feeling: A unified view of stress measurement for population science.
Frontiers in neuroendocrinology, 49, 146-169.
Fontes, A. P., & Neri, A. L. (2015). Resilience in aging: literature review. Ciencia & saude coletiva,
20, 1475-1495.
Glaser, R., Pearl, D. K., Kiecolt-Glaser, J. K., & Malarkey, W. B. (1994). Plasma cortisol levels
and reactivation of latent Epstein-Barr virus in response to examination stress.
Psychoneuroendocrinology, 19(8), 765-772.
Harpaz, R., Leung, J. W., Brown, C. J., & Zhou, F. J. (2014). Psychological stress as a trigger for
herpes zoster: might the conventiona l wisdom be wrong?. Clinical Infectious Diseases, 60(5),
781-785.
Harpaz, R., Ortega-Sanchez, I. R., & Seward, J. F. (2008). Prevention of herpes zoster:
recommendations of the Advisory Committee on Immunization Practices (ACIP). Morbidity and
Mortality Weekly Report: Recommendations and Reports, 57(5), 1-30.
Hawkley, L. C., Zheng, B., & Song, X. (2020). Negative financial shock increases loneliness in
older adults, 2006–2016: Reduced effect during the Great Recession (2008–2010). Social Science
& Medicine, 255(C).
Holmes, S. J., Reef, S. E., Hadler, S. C., Williams, W. W., & Wharton, M. (1996). Prevention of
varicella; recommendations of the Advisory Committee on Immunization Practices.

31

Howe, C. J., Cole, S. R., Lau, B., Napravnik, S., & Eron, J. J., Jr (2016). Selection Bias Due to
Loss to Follow Up in Cohort Studies. Epidemiology (Cambridge, Mass.), 27(1), 91–97.
https://doi.org/10.1097/EDE.0000000000000409
Insinga, R. P., Itzler, R. F., Pellissier, J. M., Saddier, P., & Nikas, A. A. (2005). The incidence of
herpes zoster in a United States administrative database. Journal of general internal medicine,
20(8), 748-753.
Johnson, B. H., Palmer, L., Gatwood, J., Lenhart, G., Kawai, K., & Acosta, C. J. (2015). Annual
incidence rates of herpes zoster among an immunocompetent population in the United States.
BMC infectious diseases, 15(1), 1-5.
Kawai, K., & Yawn, B. P. (2017). Risk factors for herpes zoster: a systematic review and meta-
analysis. In Mayo clinic proceedings (Vol. 92, No. 12, pp. 1806-1821). Elsevier.
Kawai, K., Gebremeskel, B. G., & Acosta, C. J. (2014). Systematic review of incidence and
complications of herpes zoster: towards a global perspective. BMJ open, 4(6), e004833.
Lasserre, A., Blaizeau, F., Gorwood, P., Bloch, K., Chauvin, P., Liard, F., ... & Hanslik, T. (2012).
Herpes zoster: family history and psychological stress—case–control study. Journal of clinical
virology, 55(2), 153-157.
Marin, M., Harpaz, R., Zhang, J., Wollan, P. C., Bialek, S. R., & Yawn, B. P. (2016, May). Risk
factors for herpes zoster among adults. In Open forum infectious diseases (Vol. 3, No. 3, p.
ofw119). Oxford University Press.
McLeod, J. D. (2012). The meanings of stress: Expanding the stress process model. Society and
Mental Health, 2(3), 172-186.

32

Michaud K. Matheson K. Kelly O. Anisman H . (2008). Impact of stressors in a natural context on
release of cortisol in healthy adult humans: A meta-analysis. Stress (Amsterdam, Netherlands), 11,
177–197. doi:10.1080/10253890701727874
Oxman, M. N. (2009). Herpes zoster pathogenesis and cell-mediated immunity and
immunosenescence. The Journal of the American Osteopathic Association, 109, S13-S17.
Ozawa, S., Portnoy, A., Getaneh, H., Clark, S., Knoll, M., Bishai, D., ... & Patwardhan, P. D.
(2016). Modeling the economic burden of adult vaccine-preventable diseases in the United States.
Health Affairs, 35(11), 2124-2132.
Pearlin, L. I., Lieberman, M. A., Menaghan, E. G., & Mullan, J. T. (1981). The stress process.
Journal of Health and Social Behavior, 22, 337–356. doi:10.2307/2136676
Pool, L. R., Needham, B. L., Burgard, S. A., Elliott, M. R., & de Leon, C. F. M. (2017). Negative
wealth shock and short-term changes in depressive symptoms and medication adherence among
late middle-aged adults. J Epidemiol Community Health, 71(8), 758-763.
Raphael, K. (1987). Recall bias: a proposal for assessment and control. International journal of
epidemiology, 16(2), 167-170.
Schetter, C. D., & Dolbier, C. (2011). Resilience in the context of chronic stress and health in
adults. Social and Personality Psychology Compass, 5(9), 634-652.
Schmader K, George LK, Burchett BM, Hamilton JD, Pieper CF. (1998a) Race and stress in the
incidence of herpes zoster in older adults. Journal of the American Geriatrics Society.
Aug;46(8):973-7.

33

Schmader K, George LK, Burchett BM, Pieper CF. (1998b) Racial and psychosocial risk factors
for herpes zoster in the elderly. The Journal of infectious diseases. Nov 1;178(Supplement_1):S67-
70.
Schmader, K., Studenski, S., MacMillan, J., Grufferman, S., & Cohen, H. J. (1990). Are stressful
life events risk factors for herpes zoster?. Journal of the American Geriatrics Society, 38(11), 1188-
1194.
Schulz, R., & Sherwood, P. R. (2008). Physical and mental health effects of family caregiving.
Journal of Social Work Education, 44(sup3), 105-113.
Segerstrom, S. C., & Miller, G. E. (2004). Psychological stress and the human immune system: a
meta-analytic study of 30 years of inquiry. Psychological bulletin, 130(4), 601.
Seiler, A., Fagundes, C. P., & Christian, L. M. (2020). The impact of everyday stressors on the
immune system and health. In Stress Challenges and Immunity in Space (pp. 71-92). Springer,
Cham.
Thomas, S. L., & Hall, A. J. (2004). What does epidemiology tell us about risk factors for herpes
zoster?. The Lancet infectious diseases, 4(1), 26-33.
Tseng, H. F., Bruxvoort, K., Ackerson, B., Luo, Y., Tanenbaum, H., Tian, Y., ... & Sy, L. S. (2020).
The epidemiology of herpes zoster in immunocompetent, unvaccinated adults≥ 50 years old:
incidence, complications, hospitalization, mortality, and recurrence. The Journal of infectious
diseases, 222(5), 798-806.
Turner, R. J., Wheaton, B., & Lloyd, D. A. (1995). The epidemiology of social stress. American
sociological review, 104-125.

34

Wheaton, B. (1994). Sampling the stress universe. In Stress and mental health (pp. 77-114).
Springer, Boston, MA.
Wheaton, B., Young, M., Montazer, S., & Stuart-Lahman, K. (2013). Social stress in the twenty-
first century. In Handbook of the sociology of mental health (pp. 299-323). Springer, Dordrecht.
Wilkinson, L. R. (2016). Financial strain and mental health among older adults during the great
recession. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 71(4),
745-754.
Zhao, L., & Burge, G. (2020). Retirement, Unretirement, and Housing Wealth during the Great
Recession. The Journal of Real Estate Finance and Economics, 1-28.

35

Table 1. Descriptive Statistics for 28,727 person-interval observations, Health and Retirement
Study, 2006-2016

Weighted
Mean
Weighted
Percent
Age (50 - 104) 67.5

Female

55.4
Race/Ethnicity

Non-Hispanic Whites

86.2
Non-Hispanic Blacks

6.3
Hispanics

5.1
Others

2.3
Education Years (0 - 17) 13.4

Number of Chronic Diseases (0 - 7) 1.9

Vaccination

16.1
Number of Chronic Stresses (0 - 8) 2.4

Experience of a Chronic Problem

89.0
Health

69.1
Spouse & Children

45.5
Alcohol/Drug

18.5
Work

18.6
Finance

43.7
Housing

14.5
Relationship

25.2
Caregiving

39.7
Number of Stressful Events (0 - 3) 0.19

Experience of a Stressful Event

17.3
Spousal Loss

2.6
Job Loss

2.5
Residential Change

4.5
Spousal ADL Onset

3.6
Wealth Loss

6.0

36

Figure 1. Shingles Incidence Rate by Stress Measures

Person-Interval N= 28,727

37

Table 2. Discrete-time Logistic Regression Models Predicting Risks for Incidence of Shingles
associated with Chronic Problems and Stressful Life Events
Model 1 Model 2
OR 95% CI OR 95% CI
Number of Stressful Events 1.09

(0.92,1.30) 1.08

(0.91,1.29)
Number of Chronic Problems 1.07 ** (1.02,1.12) 1.06 * (1.005,1.11)
Age 1.02 *** (1.01,1.03) 1.02 ** (1.01,1.03)
Female 1.41 *** (1.17,1.69) 1.40 *** (1.17,1.69)
Race/Ethnicity

Non-Hispanic Whites

Non-Hispanic Blacks

0.65 * (0.47,0.92)
Hispanics

1.16

(0.82,1.64)
Others

0.69

(0.34,1.40)
Education Years

0.97 * (0.94,1.00)
Number of Chronic Diseases

1.08 * (1.01,1.16)
Vaccination

1.01

(0.80,1.27)
Person-Interval N 28,727 28,727

38

Chapter 3. Age, Chronic Diseases, CRP: An Integrated Approach for Incidence of Shingles
among Older Americans

Abstract
Age and chronic diseases are considered independent risk factors for shingles among older adults;
yet, chronic inflammation and progression of age-related chronic diseases concurrently emerge in
the aging process. The objective of this study is to examine how health conditions common in old
age affect the risk for incidence of shingles among the older population using a framework that
integrates age groups (50-64, 65-79, 80+) and seven chronic conditions (hypertension, diabetes,
cancer, lung disease, heart disease, stroke, and arthritis) along with chronic inflammation measured
by C-reactive protein (CRP). Drawn from the 2010-2018 Health and Retirement Study, we
restricted our analyses to a pooled sample of 10,914 adults aged 50 and older who had no history
of shingles in 2010/2012 and followed this group over six years (2016/2018) until their first
shingles occurrence, death, or their last interview (25,465 person-interval observations). The
discrete-time logistic regression models revealed that age, chronic disease, and inflammation are
significantly associated with an increased risk of getting shingles. However, chronic diseases
appear to account for age differences in shingles risk in full models containing age, chronic
diseases, and inflammation all. Among chronic diseases, heart diseases and arthritis significantly
predict shingles incidence. These findings demonstrate that older adults’ higher susceptibility to
shingles is mainly driven by chronic inflammation and chronic diseases, not by age alone.

39

Introduction
Shingles is a painful and debilitating disease causing significant public health concerns in
older adults. Shingles occurs when the dormant varicella-zoster virus (VZV) initially acquired
from chickenpox becomes reactivated when cell-mediated immune function weakens, highlighting
the role of immunity in pathology of the disease (Oxman, 2009). The most well-established risk
factor for shingles is age. Research suggests that the risk of getting shingles sharply increases after
50 and continues to increase with age (Kawai et al., 2014).
Immunosenescence, age-related dysfunction and decline of the immune system is believed
to explain the associations between age and shingles (Oxman, 2009). Depletion of the availability
of T cell-associated with age, more precisely the decrease in CD4+ naïve T-cells and an increase
in CD8+ memory cells in old age can result in inefficient immune surveillance of previously
encountered pathogens (Xu et al., 2020; Fulop et al., 2016). Age-associated immunological
changes are accompanied by an elevated level of senescent cells, proinflammatory cytokines, and
antigen stimulation, which contribute to the phenomenon of “inflammaging” characterized by a
low-grade inflammation (Fulop et al., 2016; Feehan et al., 2021). The persistent inflammatory state
contributes to inefficient immunological control of infections by making immune cells functionally
exhausted. (Minato et al., 2020; Xu et al., 2020). Research found that chronic inflammation is
implicated with age-related chronic diseases of atherosclerosis, tumor formation, insulin resistance,
and arthritis. Obesity is heavily involved in progression of chronic diseases by which obese adipose
tissues infiltrated by immune cells become highly inflammatory and develop metabolic syndrome
(Frasca et al., 2017; Rodriguez et al., 2013). A higher frequency of shingles cases has been reported
among older adults with single or concurrent chronic diseases such as cardiovascular disease,
diabetes, cancer, lung diseases, or arthritis (Marra et al., 2020; Joesoef et al., 2012). Taken together,
40

disruption to immunological function becomes apparent in old age, along with age-related
physiological dysregulation and presence of chronic diseases (Barbé-Tuana et al., 2020).
Despite potential biological underpinnings of age, chronic inflammation, and chronic
diseases in shingles onset, no studies have systematically examined the extent to which each factor
is associated with shingles incidence. Most prior studies have investigated age or chronic disease
as independent and discrete pathway to shingles, although a single or multiple chronic diseases
become prevalent in advanced age (Barbé-Tuana et al., 2020; Xu et al., 2020). No population-
based studies of shingles have incorporated a biomarker for chronic inflammation which reflects
a major age-related physiological alteration that lowers immune function (Fulop et al., 2016; 2020).
Lastly, prior empirical studies have relied on specific regional or health claims data using cross-
sectional design (Kawai & Yawn, 2017; Marra et al., 2020), so we lack an understanding of how
age, chronic inflammation, andchronic disease affect the risks for shingles onset in a representative
national sample. It is crucial to expand and incorporate age-related risk factors from multiple health
domains, which may shed light on important pathways to shingles onset among older population.
To address the research gaps, the present study uses a nationally representative sample of
community-dwelling adults aged 50 and older over six years to examine a hypothesis that age,
chronic diseases, and inflammation are significantly associated with increased risks for shingles
incidence; Considering a greater adverse physiological risk linked to advanced age, we
hypothesize that chronic diseases interact with age to increase the association of age with the
incidence of shingles.

41

Methods
Data
We used the Health and Retirement Study (HRS), a longitudinal and nationally
representative study of noninstitutionalized people aged 50 and older in the United States. Since
1992, HRS has collected data every two years on a broad array of measures on health and aging
among older adults, including demographic and socioeconomic factors, chronic diseases,
insurance, and health care utilization. In 2010, the cohort of mid-baby boomers was added into the
survey to maintain representativeness of the U.S. population over age 50. Given that the risks for
shingles incidence substantially increase after age 50, our analyses used data from the 2010-2018
waves. Respondents’ information was derived from the RAND HRS data file, version P.
HRS collects blood-based biomarker data that includes CRP every four years from two
random half-samples of households: the collection was administered to half of the sample in 2010
and 2014 and the other half sample in 2012 and 2016. The analytic sample was drawn from 12,198
individuals aged 50 or older without a history of shingles at the baseline interview, beginning in
2010 or 2012, depending on the wave of CRP collection. As we aim to understand factors
associated with incidence of shingles in two-year person-intervals, we developed files for each
person interval from 2010/2012 through 2016/2018. We eliminated 1,284 members who could not
contribute to the first person-interval due to missing covariates at the beginning of the interval or
shingles status at the end of the interval, resulting in 10,914 sample members. Individuals were
followed through the incidence of shingles, death, or the last interval. Finally, the 10,914 sample
members provided 25,465 person-interval observations.

42

Measures
Incident Shingles
We identified the incidence of shingles for each two-year interval over a 6 year period in
two subsamples: One-half sample members from 2010 through 2016, and the other half sample
from 2012 through 2018. Shingles incidence is defined by reporting getting shingles during the
interval among those who had not had shingles at the beginning of the interval. Shingles was
reported in response to the question, "Since the previous wave/in the last two years, have you had
shingles?" which was asked at each interview. For example, if a respondent reported "yes" to the
question in 2014 following reporting not having had shingles in 2012 or before, we determined
that the respondent experienced incident shingles during the interval between 2012 and 2014. After
a person reports the onset of shingles, the respondent is no longer at risk of incident shingles, and
one’s later intervals are not included in the analysis. The intervals with no shingles status either at
the beginning or end of the interval were treated as missing.
Age and Chronic Diseases
Age at the beginning of each interval is categorized into three groups: 50-64 years, 65-79
years, and 80 years and above. Seven age-related chronic conditions that have been reported to be
related to shingles are included in our analysis (Kawai & Yawn, 2017; Marra et al., 2020): (a)
hypertension; (b) diabetes; (c) cancer except skin cancer; (d) chronic lung disease; (e) a heart attack,
coronary heart disease, angina, congestive heart failure, or other heart problems; (f) stroke or
transient ischemic attack; and (g) arthritis. Sample members report whether a doctor has ever told
them that they had each of the diseases at each interview. As the accumulation of chronic diseases
is common with aging (Yarnall et al., 2017), we summed the number of chronic diseases
43

respondents reported at the beginning of each interval (Range 0 – 7). We use the presence of
individual diseases as well as the number of diseases in the analysis.
Chronic Inflammation
C-reactive protein, a marker of systemic inflammation, is assayed from dried blood spots
(DBS) in HRS. We used the NHANES-equivalent CRP values that account for potential
differences in DBS assays and laboratories (Crimmins et al., 2017). CRP is measured in units of
micrograms per milliliter (μg/mL) and we used log-transformed CRP measure to improve their
distribution (Mitchell & Aneshensel, 2017). We followed the analytic sample members at 2-year
intervals, with CRP collected in 4-year intervals. For years without data on CRP, we used the
average value of CRP obtained in the preceding and the following interview. For example, we
interpolated CRP values measured in 2010 and 2014 into 2012.
Covariates
We include gender (female=1), race/ethnicity classified as non-Hispanic whites (reference),
non-Hispanic blacks, Hispanics, others, and socioeconomic status indicated by years of education
reported at the baseline and assumed to remain constant. We also included three time-varying
variables that were measured at the beginning of each interval and updated. Having a resource to
receive health care is indicated by whether a person is a beneficiary of any kind of insurance (coded
1) or not. Obesity is known to incite chronic low-grade inflammation (Frasca et al., 2017) and we
defined being obese as BMI greater than or equal to 30 kg/m
2
. The shingles vaccines approved in
2006 was available during the study period, but the vaccine’s efficacy wanes over time (Dooling
et al., 2018). We classified sample members into those ever received a shingles vaccine or not at
each interval.
44

Analytic Strategy
First, we examined level of chronic diseases, and inflammation, and accumulation of
shingles onset by age groups. Second, we fit a series of multivariate discrete-logistic regression
models. Model 1 establish the independent effects of age and chronic inflammation on the
incidence of shingles controlled for demographic factors. Model 2 establish the independent effects
of chronic diseases along with chronic inflammation on shingles onset. In Model 3, we
incorporated age, number of chronic diseases, and chronic inflammation simultaneously to
examine their relative importance in causing shingles. In Model 4, we added education, insurance,
vaccination, and obesity to examine whether the associations observed in previous models remain
robust free from other covariates. In Model 5, we added interaction terms between age and the
number of chronic diseases to determine whether the effects of the age groups differ by chronic
diseases. Third, we entered each of the seven chronic diseases into the model (hypertension,
diabetes, cancer, heart disease, lung disease, stroke, arthritis) and applied same regression
approach (Model 1-5) that examined the associations between age, number of chronic diseases,
chronic inflammation, and incidence of shingles. This approach identifies the extent to which each
chronic disease is independently linked incidence of shingles within the context of age, chronic
inflammation, and comorbid condition.
Analyses were conducted using Stata 14. In regression analyses, we associate age, chronic
diseases, chronic inflammation, and covariates at the baseline of each interval with incidence of
shingles over two-year interval in person-interval data. All analyses used sample weights to adjust
for differential sampling probability and produce estimates reflecting a nationally representative
sample of older U.S. adults. Additionally, we accounted for the clustering of interval observations
within the same respondents.
45

Results
Table 1 shows the weighted characteristics of the 10,914 sample members and by age
groups at the baseline. The mean age of the sample was 64.1 years, ranging from 50 to 99. About
59% were aged 50-64, 32% were aged 65-79, and the remaining 9% were aged 80 or older. Women
contributed more than half of the sample (53.2%), and non-Hispanic Whites were a majority
(77.8%). Most respondents were insured (91.1%), and the mean years of education was 13.2 years.
About 12% of respondents reported having had a shingles vaccination as of the baseline interview.
The mean value of logged CRP was 1.3 and the number of chronic diseases was 1.7 in sample
members. The most common chronic conditions were hypertension (53.1%) and arthritis (52.6%),
occurring in half of the respondents. The least prevalent chronic conditions were lung disease
(8.0%) and stroke (6.0%). When looking at the number of chronic diseases by age group, older
adults were more likely to have had a higher number of chronic diseases. The youngest group on
average had 1.4 diseases and the oldest group had 2.5 diseases (P=0.000). Regarding CRP, the
mean level of CRP in adults aged 65-79 was 1.3 which is higher than 1.2 in adults aged 50-64
(P=0.0504). The oldest old age group on average had 1.3 CRP, indicating no linear increase in
CRP across three age groups.
Figure 1 displays the incidence of shingles by age groups between 2010/2012 and
2016/2018. There were 518 incident shingles cases in a total of 25,465 person-interval cases. The
overall incidence of shingles was 19.0 per 1,000 person-intervals (95% CI: 17.3, 20.7). There is a
graded increase in incidence shingles rate by age groups (P=0.006). The incidence rate was 16.4
(95% CI:14.1, 18.7) in adults aged 50-64, 21.3 (95% CI: 18.5, 24.1) in adults aged 65-79, and 25.3
(95% CI: 19.8, 30.8) in adults aged 80 or older.
46

Table 2 presents the results of discrete-time logistic regression models predicting the
incidence of shingles by age, number of chronic diseases, and level of general inflammation.
Model 1 shows that a higher level of CRP (OR=1.20, 95% CI: 1.06, 1.36) is significantly
associated with increased risk of incidence of shingles. In addition, the model establishes age
differences in the risk for shingles incidence adjusted for chronic inflammation, gender, and
race/ethnicity. Those in the age group 65-79 (OR=1.32, 95% CI:1.05, 1.67), and 80 or older
(OR=1.56, 95% CI:1.16, 2.09) had higher shingles risk relative to those in the age 50-64 bracket.
Model 2 shows significant associations between number of chronic diseases and incidence of
shingles (OR=1.17, 95% CI: 1.08, 1.27) along with inflammation when controlled for
demographic factors. When incorporating age, chronic diseases, and inflammation simultaneously
in Model 3, more chronic diseases (OR=1.14, 95% CI:1.05, 1.24) and inflammation (OR=1.15,
95% CI:1.02, 1.31) remain significant predictors of incident shingles. Chronic diseases account
for the age differences initially observed; age 65-79 (OR=1.19, 95% CI:0.94, 1.52) and 80 or older
(OR=1.34, 95% CI:0.99, 1.82) as age became no longer significantly associated with shingles after
chronic diseases are taken into account. The significant effects of chronic diseases and
inflammation on shingles onset observed in the integrated Model 3 remain even after adding other
covariates (socioeconomic status, vaccination, insurance, and obesity) in Model 4. The interaction
terms for age and the number of chronic diseases in Model 5 provide no evidence that chronic
diseases interact with age in developing incident shingles.
To further our understanding of the models in Table 2, we replaced the number of chronic
diseases with indicators of the presence of the seven individual diseases (hypertension, diabetes,
heart disease, lung disease, cancer, stroke, arthritis) in Table 3. Models 1 and 2 establish
independent age, chronic disease, inflammation differentials in explaining risks for incidence of
47

shingles controlled for demographic factors. In Model 2, higher risk of incident shingles was
observed among those who had heart disease (OR=1.41, 95% CI:1.11, 1.80) and arthritis
(OR=1.38, 95% CI:1.09, 1.76). After incorporating age and seven diseases into Model 3, we found
no significant effects of advanced age (65 and 79, OR=1.17, 95% CI:0.91, 1.49 ; age 80 or older,
OR=1.29, 95% CI: 0.94, 1.77) associated with incident shingles. However, two diseases were
significantly associated with an increased risk of shingles incidence: 37 % increase in relative risk
among those with heart disease (OR=1.37, 95% CI:1.08, 1.75) and 34% increase for those with
arthritis (OR=1.34, 95% CI:1.04, 1.71). In Model 4, we found the higher risk for shingles among
patients with heart disease or arthritis observed in model 3 remains with additional control factors
(socioeconomic status, vaccination, insurance, and obesity). Lastly, we added interaction terms for
age groups and the number of diseases in Model 5. While we found a significant effect of the
number of chronic diseases (OR=1.39, 95% CI: 1.07, 1.80) in developing shingles, neither any
individual disease nor interaction terms of age and number of diseases was associated with incident
shingles.
Discussion
Anyone who has been contracted chickenpox can get shingles, and older adults are more
susceptible to getting the disease than their younger counterparts (Kawai et al., 2014). Evidence
establishing independent associations for age or chronic disease linked to shingles reinforces the
view that aging conditions may significantly affect shingles risks, yet the relative significance of
age and chronic disease in predicting shingles onset among older adults has not been systematically
investigated. Our study takes a novel empirical approach in incorporating age and chronic diseases
with an indicator of chronic inflammation to examine shingles onset.
48

There are several important findings from this work. First, we found a pattern of higher
concentration of chronic diseases and inflammation with age, which indicates links between age,
inflammation, and chronic diseases. Although we found a graded increase in the number of
diseases with age across three age groups (50-64, 65-79, 80+), the highest level of CRP was not
found in the oldest group. A prior study reported a decreased level of CRP after age 75 among
community-dwelling men aged 60 and older (Roediger et al., 2019), which may be attributable to
selective mortality in this health disadvantaged group.
Second, we calculated the shingles incidence rate in a nationally representative sample of
adults aged 50 and older from 2010 to 2018. The overall shingles incidence rate is 19.0 per 1,000
person-intervals (95% CI: 17.3, 20.7), with a higher rate in older age groups. Since the interval in
our study covers two years, the rate is translated to 9.5 per 1,000 person-years, which is similar to
9.9/1,000 person-years reported in adults aged 50 and older from private medical claims data
(Tseng et al., 2020). The national estimate of shingles incidence rate among older adults in our
study will provide foundational data for shingles surveillance in older Americans.
Third, we found that chronic inflammation measured by CRP is a significant biological
predictor of the incidence of shingles. A systematic low-grade inflammatory state is one of the
major physiological hallmarks of aging (Xu et al., 2020). A higher level of CRP has been linked
to weak immune functioning as indicated by herpes zoster vaccine responses in a prior study
(Verschoor et al., 2017). However, our study is the first to find that chronic inflammation is a
critical pathway to developing shingles in a population-based study.
Building on the significant role of inflammation, our study investigated the importance of age and
chronic diseases as risk factors for shingles onset. While the significant association between age
and incidence of shingles was identified in a model with an absence of chronic diseases; more
49

diseases remain a significant determinant of shingles but not age when we integrated age and
chronic diseases. Previous research of multimorbidity and shingles found adults aged 20-64 with
a higher burden of chronic disease were at heightened risk for shingles using cross-sectional data
(Joesoef et al., 2012). However, the relative contribution of multiple chronic diseases and age in
explaining the incidence of shingles were not reported. Our finding highlights that having chronic
diseases is the factor that makes older adults susceptible to shingles which have been masked by
age. To better understand the effect of individual chronic diseases on the incidence of shingles, we
discomposed chronic diseases into hypertension, diabetes, cancer, lung disease, heart disease,
stroke, and arthritis. We found that heart disease and arthritis were significant risk factors for the
incidence of shingles when controlled for age, chronic inflammation, and covariates. The
associations between these two diseases and shingles have been previously recognized (Smitten et
al., 2007; Veetil et al., 2013; McDonald et al., 2009; Wu et al., 2015), but in cross-sectional designs,
comorbid conditions inhibit the interpretation of disease as an etiological agent for the incidence
of shingles. Our study based on longitudinal data of community-dwelling older adults
demonstrates that heart disease and arthritis are significant risk factors for shingles among the
older population. Lastly, we found that obese older adults are less likely to get shingles. This
finding may look counterintuitive given the linkage between obesity and chronic inflammation,
yet, a paradoxical protective effects of obesity on infection related health outcomes have been well
established (Singanayagam et al., 2013).
Fourth, we found no significant effects of interaction between age and chronic diseases on
the incidence of shingles. When factors such as chronic diseases challenge aging immune function,
adaptive and maladaptive responses concurrently occur not only in the immune system but also in
other physiological systems (Fulop et al., 2020). Complex crosstalk and interaction between organs,
50

metabolism, immunity, and nervous system manifest in age-related chronic diseases (Oishi &
Manabe, 2020), which may explain why we could not delineate age-chronic disease interactive
pathway into shingles onset per se.
Our study includes some limitations. First, we chose seven diseases (hypertension, diabetes,
cancer, heart disease, lung disease, stroke, and arthritis) as age-related chronic diseases, but did
not include less common diseases for which data were not collected such as lupus and renal
diseases which have been reported to be associated with shingles (Marra et al., 2020). However,
the chronic diseases in our study are the most prevalent conditions in old age. Given our research
interest in exploring the interplay between chronic diseases with age in the onset of shingles, the
omission of relatively rare diseases is not likely to substantially change our findings. Second, our
study investigated chronic diseases without information on disease severity and management
because we did not have data on the complete information on diseases profile. Finally, the presence
of chronic diseases is self-reports of doctors’ diagnoses, which may increase measurement error.
However, the reported chronic diseases in HRS data have been validated are appropriately
indicating population health (Wallace & Herzog, 1995).
Conclusion
The main empirical finding of this study is that chronic diseases are important determinants
of incident shingles along with chronic inflammation. These findings challenge a dominant
consensus that age is the fundamental cause of older adults’ higher risk of getting shingles. The
field of geroscience highlights the importance of understanding multidimensional risk factors that
are interconnected for aging-related health outcomes (Franceschi et al., 2018). Our study is a
frontier to investigate common health features associated with age-related to each other as possible
pathways for older adults’ getting shingles.
51

Older Americans are living into old age with more burden of multiple chronic diseases than
ever before (Salive, 2012). Another cost of increasing chronic diseases in the aging population will
be an increasing risk for shingles if vaccines are not acquired (Wang et al., 2020). The increasing
trend of incidence of shingles over decades in the U.S. has become a public health concern (Kawai
et al., 2014). To our knowledge, this study is the first to examine the associations between age,
chronic diseases, chronic inflammation, and incidence of shingles in the nationally representative
older population in a longitudinal setting. Our findings underscore the importance of targeting
chronic diseases management and control as strategies for protecting older adults against shingles.

52

Bibliography
Barbé-Tuana, F., Funchal, G., Schmitz, C. R. R., Maurmann, R. M., & Bauer, M. E. (2020, August).
The interplay between immunosenescence and age-related diseases. In Seminars in
Immunopathology (pp. 1-13). Springer Berlin Heidelberg.
Crimmins, E., Faul, J., Kim, J., & Weir, D. (2017). Documentation of blood-based biomarkers in
the 2014 Health and Retirement Study. Ann Arbor, MI: University of Michigan, Survey Research
Center.
Dooling, K. L., Guo, A., Patel, M., Lee, G. M., Moore, K., Belongia, E. A., & Harpaz, R. (2018).
Recommendations of the Advisory Committee on Immunization Practices for use of herpes zoster
vaccines. Morbidity and Mortality Weekly Report, 67(3), 103.
Feehan, J., Tripodi, N., & Apostolopoulos, V. (2021). The twilight of the immune system: The
impact of immunosenescence in aging. Maturitas.
Frasca, D., Blomberg, B. B., & Paganelli, R. (2017). Aging, obesity, and inflammatory age-related
diseases. Frontiers in immunology, 8, 1745.
Franceschi, C., Garagnani, P., Parini, P., Giuliani, C., & Santoro, A. (2018). Inflammaging: a new
immune–metabolic viewpoint for age-related diseases. Nature Reviews Endocrinology, 14(10),
576-590.
Fülöp, T., Dupuis, G., Witkowski, J. M., & Larbi, A. (2016). The role of immunosenescence in
the development of age-related diseases. Revista de investigacion clinica, 68(2), 84-91.
53

Fulop, T., Larbi, A., Dupuis, G., Le Page, A., Frost, E. H., Cohen, A. A., ... & Franceschi, C.
(2018). Immunosenescence and inflamm-aging as two sides of the same coin: friends or foes?.
Frontiers in immunology, 8, 1960.
Fulop, T., Larbi, A., Hirokawa, K., Cohen, A. A., & Witkowski, J. M. (2020, September).
Immunosenescence is both functional/adaptive and dysfunctional/maladaptive. In Seminars in
Immunopathology (pp. 1-16). Springer Berlin Heidelberg.
Harpaz, R., Ortega-Sanchez, I. R., & Seward, J. F. (2008). Prevention of herpes zoster:
recommendations of the Advisory Committee on Immunization Practices (ACIP). Morbidity and
Mortality Weekly Report: Recommendations and Reports, 57(5), 1-30.
Joesoef, R. M., Harpaz, R., Leung, J., & Bialek, S. R. (2012, October). Chronic medical conditions
as risk factors for herpes zoster. In Mayo Clinic Proceedings (Vol. 87, No. 10, pp. 961-967).
Elsevier.
Kawai, K., & Yawn, B. P. (2017, December). Risk factors for herpes zoster: a systematic review
and meta-analysis. In Mayo clinic proceedings (Vol. 92, No. 12, pp. 1806-1821). Elsevier.
Kawai, K., Gebremeskel, B. G., & Acosta, C. J. (2014). Systematic review of incidence and
complications of herpes zoster: towards a global perspective. BMJ open, 4(6).
Marra, F., Parhar, K., Huang, B., & Vadlamudi, N. (2020, January). Risk factors for herpes zoster
infection: a meta-analysis. In Open forum infectious diseases (Vol. 7, No. 1, p. ofaa005). U.S.:
Oxford University Press.
54

McDonald, J. R., Zeringue, A. L., Caplan, L., Ranganathan, P., Xian, H., Burroughs, T. E., ... &
Eisen, S. A. (2009). Herpes zoster risk factors in a national cohort of veterans with rheumatoid
arthritis. Clinical infectious diseases, 48(10), 1364-1371.
Minato, N., Hattori, M., & Hamazaki, Y. (2020). Physiology and pathology of T-cell aging.
International immunology, 32(4), 223-231.
Mitchell, U. A., & Aneshensel, C. S. (2017). Social Inequalities in Inflammation: Age Variations
in Older Persons. Journal of aging and health, 29(5), 769–787.
https://doi.org/10.1177/0898264316645546
Oxman, M. N. (2009). Herpes zoster pathogenesis and cell-mediated immunity and
immunosenescence. The Journal of the American Osteopathic Association, 109(6_suppl_2), S13-
S17.
Oishi, Y., & Manabe, I. (2020). Organ system crosstalk in cardiometabolic disease in the age of
multimorbidity. Frontiers in cardiovascular medicine, 7, 64.
Rodríguez-Hernández, H., Simental-Mendía, L. E., Rodríguez-Ramírez, G., & Reyes-Romero, M.
A. (2013). Obesity and inflammation: epidemiology, risk factors, and markers of inflammation.
International journal of endocrinology, 2013.
Salive, M. E. (2012). Multimorbidity in older adults. Epidemiologic reviews, 35(1), 75-83.
Smitten, A. L., Choi, H. K., Hochberg, M. C., Suissa, S., Simon, T. A., Testa, M. A., & Chan, K.
A. (2007). The risk of herpes zoster in patients with rheumatoid arthritis in the United States and
the United Kingdom. Arthritis Care & Research, 57(8), 1431-1438.
55

Singanayagam, A., Singanayagam, A., & Chalmers, J. D. (2013). Obesity is associated with
improved survival in community-acquired pneumonia. European Respiratory Journal, 42(1), 180-
187.
Tseng, H. F., Bruxvoort, K., Ackerson, B., Luo, Y., Tanenbaum, H., Tian, Y., ... & Sy, L. S. (2020).
The epidemiology of herpes zoster in immunocompetent, unvaccinated adults≥ 50 years old:
incidence, complications, hospitalization, mortality, and recurrence. The Journal of infectious
diseases, 222(5), 798-806.
Veetil, B. M. A., Myasoedova, E., Matteson, E. L., Gabriel, S. E., Green, A. B., & Crowson, C. S.
(2013). Incidence and time trends of herpes zoster in rheumatoid arthritis: a population‐based
cohort study. Arthritis care & research, 65(6), 854-861.
Verschoor, C. P., Lelic, A., Parsons, R., Evelegh, C., Bramson, J. L., Johnstone, J., ... & Bowdish,
D. M. (2017). Serum C-reactive protein and congestive heart failure as significant predictors of
herpes zoster vaccine response in elderly nursing home residents. The Journal of infectious
diseases, 216(2), 191-197.
Wang, B., Li, R., Lu, Z., & Huang, Y. (2020). Does comorbidity increase the risk of patients with
COVID-19: evidence from meta-analysis. Aging (Albany NY), 12(7), 6049.
Wallace RB, Herzog AR. Overview of the health measures in the Health and Retirement Study. J
Hum Res. 1995;30:S84–S107.
Wu, P. H., Lin, Y. T., Lin, C. Y., Huang, M. Y., Chang, W. C., & Chang, W. P. (2015). A
nationwide population-based cohort study to identify the correlation between heart failure and the
subsequent risk of herpes zoster. BMC infectious diseases, 15(1), 1-9.
56

Xu, W., Wong, G., Hwang, Y. Y., & Larbi, A. (2020, November). The untwining of
immunosenescence and aging. In Seminars in Immunopathology (pp. 1-14). Springer Berlin
Heidelberg.
Yarnall, A. J., Sayer, A. A., Clegg, A., Rockwood, K., Parker, S., & Hindle, J. V. (2017). New
horizons in multimorbidity in older adults. Age and ageing, 46(6), 882-888.

57

Table 1. Descriptive Statistics for the Full Sample and by Age groups at Baseline in the 2010/2012
Health and Retirement Study
Full Sample Age 50-64 Age 65-79 Age 80 +
Min Max Mean (SD)/% Mean (SD) Mean (SD) Mean (SD)
Age 50 99 64.1 (9.7)

Female 53.2%

Race/Ethnicity

Non-Hispanic Whites 77.8%

Non-Hispanic Blacks 10.3%

Hispanics 8.6%

Others 3.2%

Years of Education 0 17 13.2 (2.9)

Logged CRP 0.1 5.6 1.25 (0.8) 1.2 (0.8) 1.3 (0.8) 1.3 (0.8)
Number of Chronic Diseases 0 7 1.7 (1.3) 1.4 (1.2) 2.2 (1.3) 2.5 (1.3)
Single Chronic Disease

Hypertension 53.1%

Diabetes 20.0%

Cancer 12.2%

Lung Disease 8.0%

Heart Disease 19.3%

Stroke 6.0%

Arthritis 52.6%

Insurance 91.1%

Obese 36.7%

Vaccination 12.3%

N 10,914 5,572 4,126 1,216

58

Figure 1. Incidence Shingles Rate by Age Groups Between 2010/2012 and 2016/2018

0
5
10
15
20
25
30
35
Overall Age 50-64 Age 65-79 Age 80+
Incidence Shingles Rate
Per 1,000 person-Interval
59

Table 2. Discrete-time Logistic Regression Models Examining the Associations between Age, Number of Chronic Diseases, CRP, and Incidence of
Shingles, 2010-2018 HRS
M1 M2 M3 M4 M5
OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI
Age

50-64 (ref)

65-79 1.32 * (1.05,1.67)

1.19

(0.94,1.52) 1.03

(0.80,1.34) 1.24

(0.79,1.95)
80+ 1.56 ** (1.16,2.09)

1.34

(0.99,1.82) 1.06

(0.77,1.46) 1.13

(0.63,2.03)
Logged CRP 1.20 ** (1.06,1.36) 1.14 * (1.01,1.30) 1.15 * (1.02,1.31) 1.23 ** (1.08,1.40) 1.23 ** (1.08,1.40)
Number of Chronic Diseases

1.17 *** (1.08,1.27) 1.14 ** (1.05,1.24) 1.17 *** (1.07,1.27) 1.21 ** (1.06,1.39)
Interaction Terms

Age 50-64 x # of Diseases

Age 65-79 x # of Diseases

0.92

(0.77,1.09)
Age 80 + x # of Diseases

0.96

(0.78,1.18)
Female 1.48 *** (1.18,1.86) 1.50 *** (1.19,1.88) 1.49 *** (1.18,1.87) 1.48 *** (1.18,1.86) 1.48 *** (1.17,1.86)
Race/Ethnicity

Non-Hispanic Whites (ref)

Non-Hispanic Blacks 0.95

(0.68,1.32) 0.87

(0.63,1.21) 0.90

(0.65,1.25) 0.92

(0.66,1.29) 0.92

(0.66,1.28)
Hispanics 1.21

(0.87,1.68) 1.16

(0.84,1.61) 1.19

(0.86,1.66) 1.19

(0.81,1.74) 1.19

(0.81,1.74)
Others 1.60

(0.91,2.82) 1.55

(0.88,2.73) 1.60

(0.91,2.83) 1.60

(0.91,2.82) 1.59

(0.90,2.81)
Education

0.98

(0.94,1.01) 0.98

(0.94,1.01)
Vaccination

1.15

(0.88,1.51) 1.15

(0.88,1.50)
Insurance

1.92 * (1.07,3.44) 1.91 * (1.06,3.42)
Obesity

0.66 *** (0.52,0.84) 0.65 *** (0.51,0.83)
N 25465 25465 25465 25465 25465
60

Table 3. Discrete-time Logistic Regression Models Examining the Associations between Age, Discrete Chronic Diseases, CRP, and Incidence of
Shingles, 2010-2018 HRS

* p<.05; ** p<.01; *** p<.001
M1 M2 M3 M4 M5
OR 95% CI OR 95% CI OR 95% CI OR 95% CI OR 95% CI
Age

50-64 (reference)

65-79 1.32 * (1.05,1.67)

1.17

(0.91,1.49) 1.02

(0.79,1.33) 1.24

(0.78,1.96)
80+ 1.56 ** (1.16,2.09)

1.29

(0.94,1.77) 1.04

(0.75,1.44) 1.12

(0.62,2.03)
Logged CRP 1.20 ** (1.06,1.36) 1.15 * (1.01,1.31) 1.16 * (1.02,1.32) 1.23 ** (1.08,1.40) 1.23 ** (1.08,1.40)
Individual Chronic Diseases

Hypertension

0.98

(0.77,1.23) 0.95

(0.76,1.20) 1.00

(0.79,1.26) 0.75

(0.53,1.05)
Diabetes

1.02

(0.79,1.31) 1.02

(0.79,1.30) 1.07

(0.83,1.39) 0.81

(0.57,1.15)
Cancer

1.32

(0.99,1.76) 1.28

(0.95,1.71) 1.28

(0.95,1.72) 0.97

(0.65,1.44)
Lung Disease

1.09

(0.77,1.54) 1.09

(0.77,1.54) 1.08

(0.76,1.53) 0.81

(0.52,1.26)
Heart Disease

1.41 ** (1.11,1.80) 1.37 * (1.08,1.75) 1.36 * (1.07,1.74) 1.03

(0.72,1.47)
Stroke

1.01

(0.68,1.49) 0.99

(0.66,1.47) 0.98

(0.66,1.46) 0.74

(0.47,1.18)
Arthritis

1.38 ** (1.09,1.76) 1.34 * (1.04,1.71) 1.34 * (1.05,1.72) 1

(1.00,1.00)
Number of Chronic Diseases

1.39 * (1.07,1.80)
Interaction Terms

50-64 (ref) x #Diseases

65-79 x #Diseases

0.91

(0.76,1.09)
80+ x #Diseases

0.96

(0.77,1.18)
Female 1.48 *** (1.18,1.86) 1.48 *** (1.17,1.86) 1.47 ** (1.16,1.85) 1.46 ** (1.16,1.85) 1.46 ** (1.16,1.85)
Race

Non-Hispanic Whites

Non-Hispanic Blacks 0.95

(0.68,1.32) 0.93

(0.67,1.29) 0.95

(0.68,1.32) 0.96

(0.69,1.35) 0.96

(0.68,1.34)
Hispanics 1.21

(0.87,1.68) 1.24

(0.89,1.72) 1.27

(0.91,1.76) 1.24

(0.84,1.81) 1.23

(0.84,1.80)
Others 1.6

(0.91,2.82) 1.64

(0.93,2.90) 1.68

(0.95,2.98) 1.67

(0.94,2.95) 1.66

(0.94,2.94)
Education

0.97

(0.94,1.01) 0.97

(0.94,1.01)
Vaccination

1.14

(0.87,1.49) 1.13

(0.86,1.48)
Insurance

1.89 * (1.06,3.39) 1.88 * (1.05,3.36)
Obesity 0.68 ** (0.53,0.86) 0.67 ** (0.53,0.86)
N 25465 25465 25465 25465 25465
61

Appendix
Table 1. Weighted Mean of CRP and Number of Chronic Diseases by Age Groups at Baseline (2010/2012)

NHANES CRP P value
Logged
NHANES CRP P value
Chronic
diseases P value
Age 50-64 (1) 4.07 0.03
(1)vs(2)
1.24 0.0504
(1)vs(2)
1.35 0.00
(1)vs(2)

Age 65-79 (2) 4.53 0.83
(2)vs(3)
1.28 0.36
(2)vs(3)
2.17 0.00
(2)vs(3)

Age 80 + (3) 4.46 0.22
(1)vs(3)
1.25 0.70
(1)vs(3)
2.47 0.00
(1)vs(3)

62

Chapter 4. Shingles Vaccine Uptake among Older Adults: Identifying Early, Late, and
Non-adopters

Abstract
Objectives: We examined the characteristics of early, late, and non-adopters of the shingles
vaccine over the first ten years of vaccine availability.
Methods: Using data from the Health and Retirement Study for those 62 and older in 2008, we
examined characteristics linked to vaccination from 2006 through 2016 based on multinomial
logistic regression.
Results: About two-thirds (64.6%) of those eligible for the shingles vaccine when it was introduced
remained unvaccinated after ten years; 15.2% were early vaccine adopters (2006-2010), and 20.2%
were late adopters (2011-2016). Early adopters were more likely than late adopters to be older,
non-Hispanic White, socioeconomically advantaged and conscientious, recipients of the flu
vaccine, and residents of an area with a higher shingles immunization rate.
Conclusion: Shingles vaccine uptake is unequally distributed by sociodemographic status and
geography. Efforts to improve and expedite uptake of new preventative measures should target
specific populations and specific geographic areas.

63

Introduction
Reducing the burden of vaccine-preventable diseases through immunization is a public
health goal. Shingles affects one in three Americans in their lifetime by causing a painful rash
which can lead to poor social and emotional well-being, difficulties in physical functioning,
prolonged medical care, and hospitalization (Harpaz et al., 2008; Katz et al., 2004). Shingles
prevalence and severity can be reduced by receipt of a shingles vaccine. Zostavax, the vaccine
approved in 2006, was estimated to halve the risk of contracting the disease (Oxman et al., 2005)
and Shingrix, a two-dose vaccine approved in 2017, is even more effective (90%) in preventing
shingles (Lal et al., 2015). Despite evidence of the effectiveness of these vaccines, a large
proportion of older persons have yet to be vaccinated for shingles. In 2016, estimates from the
National Health Interview Survey indicate that two-thirds (67%) of adults aged 60 and older
remained unvaccinated ten years after the vaccine first became available (Dooling et al., 2018).
There is growing interest in how to accelerate the adoption of the newer shingles vaccine
as well as other vaccines because the speed of adoption determines how fast the burden of
preventable diseases can be reduced. Influenza vaccine was first introduced in 1930s and started
being widely available in 1945, but it took almost 5 decades for the vaccination rate to reach 50%
(NCHS, 2018). On the other hand, pneumococcal vaccine was adopted more quickly than flu
vaccine. In the United States, pneumococcal vaccine was first licensed in 1977 and in 2017, 69%
of older adults ages 65+ had ever received a pneumococcal vaccine (Gierke et al. 2021).
Given the slow dissemination of earlier vaccines targeted to the older population,
understanding the time frame involved with shingles vaccine acceptance should help focus new
vaccination strategy. Some individuals may be aware of medical developments and be ready to
immediately adopt preventive measures; others may wait until their health care provider makes a
64

suggestion to get a vaccine; yet others may only become convinced by having personal experience
with people who have gotten the disease or been vaccinated. The issues associated with vaccination
acquisition result in three distinct groups: early adopters who quickly receive the vaccine, late
adopters who are slow to react so get vaccinated after a significant delay, and those who are never
vaccinated. Because delaying the shingles vaccine presents an unnecessary risk of disease,
understanding these three distinct groups for a recent vaccine roll-out could aide in implementing
intervention strategies.
Previous research on shingles vaccination has primarily compared the characteristics of
vaccine recipients and non-recipients at one-time point (Lu et al., 2009; 2017; Lee et al., 2013;
Tseng et al., 2012; Zhang et al., 2012, 2017; Langan et al., 2013; Shen et al., 2019; Vogelsang et
al., 2021).

These studies have found greater vaccine uptake among those who have advantaged
socioeconomic backgrounds, good health, frequent healthcare encounters and experience with
immunization, know someone with shingles, and who live in the West Central region of the United
States. While prior studies relied on cross-sectional data to identify factors that were associated
with the acceptance of the vaccine at specific time points, they were not able to determine
differences in uptake over time. Given the lack of understanding on how people adopt vaccine
over time, not only whether people acquire vaccine, but also when they do – early or late – becomes
important for improving further shingles prevention. Considering shingles-related costs are $782
million annually (Ozawa et al., 2016), early shingles vaccination can significantly decrease
potential economic and societal costs resulting from shingles.
Despite the importance of understanding the characteristics of early and late vaccine
recipients, to our knowledge, no studies to date have examined the timing of vaccination of any
kind among the older population. The objective of the current study is to identify the characteristics
65

of early, late, and non-adopters of the shingles vaccine over the ten years after vaccine approval in
a nationally representative longitudinal sample of older Americans.
Early shingles vaccination requires prior awareness of the existence and effectiveness of
the vaccine because it was a newly recommended vaccine for older adults following the
pneumococcal vaccine (Lu et al., 2017). In addition, there was a national shortage in supply of
shingles vaccine between 2008 and 2011(Hechter et al., 2017), highlighting the importance of
access to the information on vaccine availability. Shingles vaccine was the most expensive vaccine
recommended for older adults with limited coverage in the early phase of the vaccine release,
which involved financial burden (GAO, 2011; Hurley et al., 2010). There were public health
efforts to lower older adults’ out-of-pocket cost for paying the vaccine through the Medicare
prescription drug insurance in 2008, Affordable Care Act in 2010, and CMS-initiated a new
vaccine tier in 2011 (GAO, 2011; Hechter et al., 2017; Williams et al., 2016), but older adults
financial responsibility remains, with average $54 copayment (Yan et al., 2018). Therefore, we
expect that those with higher socioeconomic status will have higher medical literacy, greater
financial resources, and insurance coverage and be more likely to get vaccinated earlier.
Previous literature has demonstrated some role for personality traits, particularly
conscientiousness, as an important influence on the utilization of preventive medical services
(Armon & Toker, 2013; Schmid et al., 2017). We also expect that those who have had experience
with other vaccines such as flu and pneumonia will be more likely to get vaccinated early. In
addition to an individual’s own resources and experience, the decision to get vaccinated often
depends on external influences such as recommendations from peers (Sato & Takasaki, 2016; Rao
et al., 2007) and immunization practice in the community (Romley et al., 2016).

Therefore, we
expect that those with frequent social interactions and those living in an area with higher shingles
66

vaccination rates will be more likely to get vaccinated earlier as they are more likely to be exposed
to up-to-date health information and positive anecdotes regarding vaccination.
Methods
Data
We used data from the 2006-2016 Health and Retirement Study (HRS), an ongoing,
biennial, nationally representative longitudinal survey of persons over age 50. The sample is based
on a multi-stage areal probability sampling design with an oversampling of African Americans
and Hispanics. The study collects extensive information on demographics, health, health care
utilization, availability of insurance, etc. HRS has collected information on shingles vaccination
in the core interview at each wave since 2008. In addition to information from the core survey,
measures of social interaction and personality were obtained from the HRS Psychosocial and
Lifestyle self-administered questionnaire which is given to a rotating half of the sample every two
years, e.g., one half in 2006 and the other half in 2008 and repeated after four years (e.g., in 2010
and 2012). Respondent information was obtained from the RAND HRS data file, version P. A
measure of the level of shingles vaccination in a person’s area of residence was developed from
the HRS sample based on the zip code/hospital referral region (HRR) crosswalk file from
Contextual Data resource linked to HRS (Ailshire & Young, 2018).
Sample
When the shingles vaccine first became publicly available in 2006, it was recommended
for people age 60 and older. For this reason, we limited the sample to those who were eligible to
receive the vaccine when it was first introduced, or the 12,568 adults aged 62 and older at the 2008
HRS interview. We eliminated 83 individuals who did not report their shingles vaccination
experience, 124 respondents whose timing of vaccination was unknown due to skipping the
67

vaccine question for re-interviewees in 2014 or who reported a date before the vaccine was
available, and 2,361 respondents who had missing data on social interactions, personality, or other
covariates. We linked each person to the estimated level of vaccination in the hospital referral
region where he/she lived. There were 66 respondents for whom HRR vaccination rates could not
be calculated. We do not include sample members who could not be geocoded to a hospital referral
region due to missing data on their geographic residence (n=40) or who were the only resident in
the HRS sample in a hospital referral region (n=26). The final analytic sample was 9,934.
Compared to the analytic sample, those with missing information (N=2,634) were more likely to
be older (75.6 vs. 72.5, p<.001) and less likely to be non-Hispanic White (72.4% vs. 84.3%,
p<.001), but they were similar in gender. Those who were missing from the analysis and who
provided information on vaccination (N=2,510) had a lower vaccination rate (16.2% vs. 35.4%,
p<.001) than those who were in the analysis.
Measures
Vaccine Adopters
We classified each respondent into three categories: non-adopters who did not get
vaccinated by 2016, early adopters who were vaccinated in the four years from 2006 to 2010, and
late adopters who were vaccinated between 2011 and 2016. Information on receipt and timing of
vaccination was obtained from two survey questions: “Have you ever had the shingles vaccine?”
asked in 2008; the same question was asked of those who had not reported the vaccine in previous
waves in 2010, 2012, and 2016. If the respondent reported receiving a vaccine, the year at
vaccination was asked in the 2008 and 2010 surveys.
Individual-level Characteristics Affecting Vaccination
68

Individual characteristics included demographic and socioeconomic factors, having
prescription drug insurance, perceived health status, having had a flu vaccine, level of social
interactions with family/children and friends, and conscientiousness. Baseline demographic
information included age, sex, race/ethnicity (categorized as non-Hispanic White, non-Hispanic
Black, Hispanic, and other), and marital status with being married as the reference compared to
divorced/separated, widowed or never married. Two measures of socioeconomic status were
included: logged household income and highest completed level of education (less than high
school (reference), high school, and some college or higher). Having a source of coverage for the
shingles vaccination was categorized into having a prescription drug insurance plan (reference) or
not. Poor perceived health was indexed as fair or poor self-rated health compared to good or better
(reference). An indicator of the use of preventive health care as well as experience with vaccines
was whether respondents had ever received the flu vaccine (reference). Conscientiousness was
measured as the average self-rating on five items (organized, responsible, hardworking, careless
(reverse-coded), and thorough) using a 4-point rating scale (1=not at all, 2=a little, 3=some, 4= a
lot). The scale ranged from 1 to 4, with a higher value indicating being more conscientious. A scale
of social interaction with family was based on the average frequency (1=less than once a year or
never, 2=once or twice per year, 3=every few months, 4=once or twice per month, 5=once or twice
per week, 6=three or more times per week) of three modes of contacts (written or email, telephone,
and in-person contact) between the respondents and their children and other family members. The
scale ranged from 1 to 6, with a higher value representing more frequent interaction. Similarly, a
scale of social interaction with friends ranging from 1-6 is the average frequency of the three modes
of social contact described above between the respondent and friends. All individual-level
characteristics were derived from data collected in 2008 except for information on social
69

interactions and personality traits which was collected for half the sample in 2006 and the other
half in 2008, and 2010/2012 if missing from 2006/2008.
Areal-level Vaccination
We calculated the shingles vaccination rates in the hospital referral region of residence
based on HRS respondents’ reports of vaccination for the areas to indicate both collective medical
practice and exposure to vaccinated persons in the area. A hospital referral region (HRR) is
designed to reflect a regional health care market; the U.S. is divided into 306 hospital referral
regions. Health care quality, access, and cost significantly differ across hospital referral regions
(Goodman et al., 2010). Our sample lived in 245 HRRs out of the 306 HRRs in the United States.
On average, 104 sample members lived in each HRR, with only 2.5% of the sample living in an
HRR with five or fewer sample members. The vaccination prevalence rate in an HRR was
computed as the number of survey respondents vaccinated through 2016 divided by the number of
respondents living in the HRR. In order to avoid a potential bidirectionality issue, we excluded the
individual respondent from both the numerator and denominator when calculating the rate for an
HRR. For analysis, the HRR were divided into quartiles by the vaccination rate in sample members,
with the upper quartile indicating the highest vaccination rate.
Statistical Analysis
We first examined descriptive profiles of early, late and non-adopters, indicating
differences between early and late adopters and non-adopters as well as differences between early
and late adopters. Then we used multinomial logistic regression models to investigate the factors
associated with early and late adopters relative to those who did not get vaccinated and to each
other. Analyses were conducted using Stata, version 16. In all analyses, results were weighted to
represent the national population.
70

Results
Table 1 presents weighted characteristics for the full sample and by vaccine adopter
category. Non-adopters accounted for about two-thirds (64.6%) of the sample; early vaccine
adopters (15.2%) and late vaccine adopters (20.2%) made up about a third of the sample. This
indicates that the majority of the sample remained unvaccinated after ten years of vaccine
availability.
When looking at the sample characteristics by vaccine adopter category, early adopters
were older than late adopters (72.3 vs. 70.2, p<.001) but younger than non-adopters (73.2, p<.001).
While both early and late adopters were less likely to perceive their health as poor than non-
adopters, late adopters considered themselves in poor health more than early adopters (18.8% vs.
15.6%, p=0.03). Previous vaccine experience was higher among early adopters (89.6%) than late
adopters (80.0%) (p<.001). More than one-third of early adopters (36.2%) lived in the area with
the highest shingles vaccination rate, compared with 26.1% in late adopters and 24.7% of non-
adopters.
Table 2 displays the multinomial logistic regression model results examining the
characteristics of early and late adopters compared to those who were not vaccinated (Columns 1
and 2) and to each other (Column 3). Advanced age was associated with a decreased relative
likelihood of early (RRR=0.98, 95% CI: 0.97, 0.99) and late vaccination (RRR=0.94, 95% CI:
0.93, 0.95) over non-vaccination. A higher level of education and income were attributes of early
and late vaccine adopters relative to non-vaccinators. Individuals covered by prescription drug
plans were 1.8 times more likely to get vaccinated late than those without prescription insurance.
Perceived poor health was related to about a 40% decrease in the relative likelihood of early or
late vaccination. Experience of flu vaccination was linked to five-fold and three-fold increased
71

odds of being early and late adopters, respectively. Those who more frequently engaged with
friends were more likely to receive the vaccine early (RRR=1.16, 95% CI: 1.07, 1.25) and late
(RRR=1.08, 95% CI: 1.01, 1.15). Those living in regions above the lowest quartile of shingles
immunization rate generally had a greater likelihood of receiving the vaccine – both early and late
- than their counterparts living in areas in the lowest quartile of shingles vaccination.
Column 3 in Table 2 shows the differences between early adopters and late adopters. Those
with advanced age were more likely to receive the vaccine early (RRR=1.04, 95% CI: 1.03, 1.05).
The relative likelihood of early vaccination over late vaccination is 60% lower for non-Hispanic
Blacks than for non-Hispanic Whites. Socioeconomic factors played a role in early vaccination.
For example, among the vaccinated, those who completed some college or higher education were
1.6 times as likely to receive the vaccine early compared to those who completed less than high
school. In addition, having a higher household income was associated with a 14% increased
relative probability of early vaccination over late vaccination. Experience of other immunization
significantly influenced the timing of vaccination. Those who have received a flu vaccine were
almost two times more likely to receive the shingles vaccine early. A more conscientiousness
personality was associated with an increased relative likelihood of early vaccination relative to late
vaccination. Collective vaccination practice in the area of residence was a critical promoter for
one’s early vaccination. Those living in the region with the highest level of shingles immunization
were 1.7 times more likely to receive the vaccine early than their counterparts living in the region
with the lowest level of the immunization rate.

72

Discussion
During the first ten years the shingles vaccine was available, vaccine uptake rate remained
low at 35.4% in our national sample of persons who were 60 and older when the vaccine became
available. The shingles vaccination rate in our sample is similar to the 33% national immunization
rate observed among adults 60 and older, reported by the CDC in 2016 (Dooling et al., 2018).
While quite similar, our study showed a slightly higher rate partly because our sample includes the
institutionalized population while the estimates provided by CDC used the non-institutionalized
population from the National Health Interview Survey.
With the longitudinal sample, our study findings add to the literature by identifying factors
associated with the acceptance and timing of the shingles vaccine. As shown in the previous studies,
vaccine adoption was facilitated by individual characteristics such as a higher level of education
and income and flu vaccination experience. Our findings are consistent with previous research
regarding geographic disparities in shingles vaccination, with relatively higher rates in the West
and Midwest (Vogelsang et al., 2021; Terlizzi & Black, 2020). While categorization of geography
and related medical practices differs across studies where we relied on HRRs that often cross
boundaries of geographic divisions, those who resided in an area with higher levels of shingles
vaccination were more likely to receive shingles vaccine.
However, in contrast to other studies showing a positive association of shingles vaccination
with advanced age (Lu et al., 2017; Vogelsang et al., 2021), we found that vaccination was lower
among those with older age and perceived poor health. The discrepancy in age effects on
vaccination may reflect differences in sample ages and study periods. Our sample members were
aged 60 and older in 2006 and reported vaccination status in prospective interviews making them
73

older than other samples who were aged 60 and older and reported ever receiving a vaccine at later
years in the vaccine release (Lu et al., 2017; Vogelsang et al., 2021).
It is likely that older people with multiple chronic conditions and who take multiple
medications may be more cautious about potential side effects and see less benefit to the vaccine.
Given the increased risk of getting shingles with age (Insinga et al., 2005), lower vaccination with
increasing age is of concern. Public education that addresses the importance of vaccine in
preventing shingles in old age would be important to encouraging shingles vaccine. Given the
positive influences of frequent social interaction and higher areal level of vaccination on vaccine
acceptance, dissemination of more accurate information about shingles and the vaccine through
social and local networks may increase the vaccine uptake of older people.
In addition to characterizing shingles vaccine adopters and non-adopters, our study also
demonstrated that early adopters were significantly different from late adopters. While older age
was negatively related to vaccine adoption relative to non-adoption, among adopters, older persons
were more likely to be early vaccine adopters relative to late adopters. This may suggest that older
people may delay their decision to get the vaccine, but once they see the benefits of vaccine, they
may take a vaccine early. On the other hand, many of the other characteristics of vaccine adopters
compared to non-adopters characterized early vs. late adopters. Our study showed that higher
socioeconomic status was strongly correlated with adoption of the vaccination (Lu et al., 2009;
Tseng et al., 2012; Zhang et al., 2012; Langan et al., 2013; Vogelsang et al., 2021), but
socioeconomic resources also impacted the timing of adoption within the vaccine adopter group.
Those who received the vaccine early had greater socioeconomic resources and may have been
better positioned to take a new and effective preventive medicine when it was first introduced (Lu
et al., 2008). Unequal uptake of the vaccine in the early years after its release adds to existing
74

disparities in health. According to fundamental cause theory, high SES groups can take advantage
of a broad range of resources to access a new medical technology when it is introduced (Phelan et
al., 2005). In fact, those with higher SES were better able to acquire the vaccine even when the
supply was limited. Further, those from higher-income households were among the first to be
vaccinated prior to the policies that aim to remove copayment of the vaccine.
Although socioeconomic barriers seemed to play a pivotal role in vaccine uptake during
the early window of availability, an individual’s social interactions and personality, as well as area
of residence were key drivers of early vaccine uptake. Those with a more conscientious personality
and whose care had included other vaccines were more likely to get the injection early on. Those
who were more alert to information about their health matters may have found information
indicating the effectiveness of the shingles vaccination in preventing shingles and thus got
immunized soon after the vaccine became publicly available. We also found that living in a region
with the highest shingles vaccination rates induced individuals to become early adopters. The
associations between higher areal vaccination rates and one’s greater propensity to get vaccinated
early may be a reflection of an individual’s desire to conform to peers’ health care behaviors (Sato
et al., 2016; Rao et al., 2007) or the result of a disparity in medical practices across hospital referral
regions (Molitor et al., 2018).

While better perceived health was related to vaccine uptake whether early or late, it did not
differentiate early and late vaccine takers. That is, while both early and late vaccine adopters
perceived their health better than non-adopters did, early and late vaccine takers were not different
from each other in their perception of health. This may suggest that while poorer perceived health
may decrease people’s decision to take a shingles vaccine, it doesn’t determine the timing of
vaccine uptake.
75

We believe this study is the first to analyze the factors that differentiate the acceptance and
the timing of the shingles vaccine uptake among older adults with longitudinal national data.
However, there are a couple of limitations to note. The experience and year of vaccination were
self-reported in our data. Recall bias may affect the classification of shingles vaccine adopter
categories and our ability to predict factors associated with the timing of the vaccine uptake
accurately. However, given the consistency found between self-reported adult vaccination uptake
and health administrative records (King et al., 2018), we do not expect that recall bias substantially
altered the conclusion from this study. Also, our independent variables are time invariant, thus
they are not sensitive to potentially time-varying factors such as income and insurance coverage
upon the vaccine uptake or over time. In addition, our sample was limited to those aged 60 and
over when the first shingles vaccine was introduced because shingles vaccine was not
recommended for younger people. Recently, the Advisory Committee on Immunization Practices
updated recommendations of usage of a new shingles vaccine for broader age groups including
those aged 50 to 60. Future studies are needed on inoculation choices in younger groups to better
counter non-adoption and delay of vaccine uptake in these groups.
Public Health Implications
Unequal uptake of the shingles vaccine by race/ethnicity, socioeconomic status, and area
of residence in the early period of the vaccine release indicates social inequity in the prevention of
shingles. Race/ethnic minorities and those with disadvantaged socioeconomic status have fewer
opportunities and access to the vaccination, particularly timely vaccination. The new shingles
vaccine (Shingrix) approved in 2017 is still in the initial phase of dissemination and is expensive
without insurance coverage. Efforts to address financial barriers and medical literacy among those
with disadvantaged socioeconomic status and those living in medically underserved areas would
76

be important in order to increase the vaccination rate promptly and thus reduce the occurrence of
shingles. The more affordable the shingles vaccine is and the sooner the public are aware of the
benefits and get vaccinated, the more likely our society will save the medical and social costs
associated with shingles.
77

Bibliography
Ailshire J, Matt Y. (2018). Contextual Data Resource: Dartmouth Atlas of Health Care. Version
1.0. LosAngeles, CA: USC/UCLA Center on Biodemography and Population Health.
Armon, G., & Toker, S. (2013). The role of personality in predicting repeat participation in
periodic health screening. Journal of personality, 81(5), 452-464.
Dooling, K. L., Guo, A., Patel, M., Lee, G. M., Moore, K., Belongia, E. A., & Harpaz, R. (2018).
Recommendations of the Advisory Committee on Immunization Practices for use of herpes zoster
vaccines. Morbidity and Mortality Weekly Report, 67(3), 103.
Gierke A. Patricia Wodi, and Miwako Kobayashi. (2021). Pneumococcal Disease. Epidemiology
and Prevention of Vaccine-Preventable Diseases. The Pink Book:13th Edition.
TGoodman, D. C., Brownlee, S., Chang, C., & Fisher, E. S. (2010). Regional and racial variation
in primary care and the quality of care among Medicare beneficiaries. Lebanon, NH: Dartmouth
Atlas of Health Care.
Harpaz, R., Ortega-Sanchez, I. R., & Seward, J. F. (2008). Prevention of herpes zoster:
recommendations of the Advisory Committee on Immunization Practices (ACIP). Morbidity and
Mortality Weekly Report: Recommendations and Reports, 57(5), 1-30.
Hechter, R. C., Qian, L., Yan, S., Luo, Y., Krishnarajah, G., & Tseng, H. F. (2017). Impact of the
change of copay policy in Medicare Part D on zoster vaccine uptake among Medicare beneficiaries
in a managed care organization. BMC health services research, 17(1), 1-8.
Hurley, L. P., Lindley, M. C., Harpaz, R., Stokley, S., Daley, M. F., Crane, L. A., ... & Kempe, A.
(2010). Barriers to the use of herpes zoster vaccine. Annals of internal medicine, 152(9), 555-560.
78

Insinga, R. P., Itzler, R. F., Pellissier, J. M., Saddier, P., & Nikas, A. A. (2005). The incidence of
herpes zoster in a United States administrative database. Journal of general internal medicine,
20(8), 748-753.
Katz, J., Cooper, E. M., Walther, R. R., Sweeney, E. W., & Dworkin, R. H. (2004). Acute pain in
herpes zoster and its impact on health-related quality of life. Clinical Infectious Diseases, 39(3),
342-348.
King, J. P., McLean, H. Q., & Belongia, E. A. (2018). Validation of self‐reported influenza
vaccination in the current and prior season. Influenza and other respiratory viruses, 12(6), 808-813.
Lal, H., Cunningham, A. L., Godeaux, O., Chlibek, R., Diez-Domingo, J., Hwang, S. J., ... &
Heineman, T. C. (2015). Efficacy of an adjuvanted herpes zoster subunit vaccine in older adults.
New England Journal of Medicine, 372(22), 2087-2096.
Langan, S. M., Smeeth, L., Margolis, D. J., & Thomas, S. L. (2013). Herpes zoster vaccine
effectiveness against incident herpes zoster and post-herpetic neuralgia in an older US population:
a cohort study. PLoS medicine, 10(4), e1001420.
Lee, T. J., Hayes, S., Cummings, D. M., Cao, Q., Carpenter, K., Heim, L., & Edwards, H. (2013).
Herpes zoster knowledge, prevalence, and vaccination rate by race. The Journal of the American
Board of Family Medicine, 26(1), 45-51.
Lu, P. J., Euler, G. L., Jumaan, A. O., & Harpaz, R. (2009). Herpes zoster vaccination among
adults aged 60 years or older in the United States, 2007: uptake of the first new vaccine to target
seniors. Vaccine, 27(6), 882-887. Lu, P. J., O’Halloran, A., Williams, W. W., & Harpaz, R. (2017).
National Center for Health Statistics 2018
79

Lu, P. J., O’Halloran, A., Williams, W. W., & Harpaz, R. (2017). National and state-specific
shingles vaccination among adults aged≥ 60 years. American journal of preventive medicine, 52(3),
362-372.
Molitor, D. (2018). The evolution of physician practice styles: evidence from cardiologist
migration. American Economic Journal: Economic Policy, 10(1), 326-56.National Center for
Health Statistics. 2018. United States
Oxman, M. N., Levin, M. J., Johnson, G. R., Schmader, K. E., Straus, S. E., Gelb, L. D., ... &
Silber, J. L. (2005). A vaccine to prevent herpes zoster and postherpetic neuralgia in older adults.
New England Journal of Medicine, 352(22), 2271-2284.
Ozawa, S., Portnoy, A., Getaneh, H., Clark, S., Knoll, M., Bishai, D., ... & Patwardhan, P. D.
(2016). Modeling the economic burden of adult vaccine-preventable diseases in the United States.
Health Affairs, 35(11), 2124-2132.
Phelan, J. C., & Link, B. G. (2005). Controlling disease and creating disparities: a fundamental
cause perspective. The Journals of Gerontology Series B: Psychological Sciences and Social
Sciences, 60(Special_Issue_2), S27-S33.
Rao, N., Möbius, M. M., & Rosenblat, T. (2007). Social networks and vaccination decisions (No.
07-12). Working Papers.
Romley, J., Goutam, P., & Sood, N. (2016). National survey indicates that individual vaccination
decisions respond positively to community vaccination rates. PloS one, 11(11), e0166858.
Sato, R., & Takasaki, Y. (2016). Peer effects on vaccination: Experimental evidence from rural
Nigeria. CIRJE Discussion Paper, CIRJE-F-1002.
80

Schmid, P., Rauber, D., Betsch, C., Lidolt, G., & Denker, M. L. (2017). Barriers of influenza
vaccination intention and behavior–a systematic review of influenza vaccine hesitancy, 2005–2016.
PloS one, 12(1), e0170550.
Shen, A. K., Warnock, R., Selna, W., MaCurdy, T. E., Chu, S., & Kelman, J. A. (2019).
Vaccination among Medicare-fee-for service beneficiaries: Characteristics and predictors of
vaccine receipt, 2014–2017. Vaccine, 37(9), 1194-1201.
Terlizzi, E. P., & Black, L. I. (2020). Shingles vaccination among adults aged 60 and over: United
States, 2018.
Tseng, H. F., Chi, M., Smith, N., Marcy, S. M., Sy, L. S., & Jacobsen, S. J. (2012). Herpes zoster
vaccine and the incidence of recurrent herpes zoster in an immunocompetent elderly population.
The Journal of infectious diseases, 206(2), 190-196.
US Government Accountability Office. (2011). Medicare: many factors, including administrative
challenges, affect access to Part D vaccinations.
Vogelsang, E. M., & Polonijo, A. N. (2021). Social Determinants of Shingles Vaccination in the
United States. The Journals of Gerontology: Series B.
Williams, W. W., Lu, P. J., O’Halloran, A., Kim, D. K., Grohskopf, L. A., Pilishvili, T., ... &
Bridges, C. B. (2016). Surveillance of vaccination coverage among adult populations—United
States, 2014. Morbidity and Mortality Weekly Report: Surveillance Summaries, 65(1), 1-36.
Yan, S., DerSarkissian, M., Bhak, R. H., Lefebvre, P., Duh, M. S., & Krishnarajah, G. (2018).
Relationship between patient copayments in Medicare Part D and vaccination claim status for
81

herpes zoster and tetanus-diphtheria-acellular pertussis. Current medical research and opinion,
34(7), 1261-1269.
Zhang, D., Johnson, K., Newransky, C., & Acosta, C. J. (2017). Herpes zoster vaccine coverage
in older adults in the US, 2007–2013. American journal of preventive medicine, 52(1), e17-e23.
Zhang, J., Xie, F., Delzell, E., Chen, L., Winthrop, K. L., Lewis, J. D., ... & Curtis, J. R. (2012).
Association between vaccination for herpes zoster and risk of herpes zoster infection among older
patients with selected immune-mediated diseases. Jama, 308(1), 43-49.

82

Table 1. Descriptive Statistics for the Full Sample and by Vaccine Adopter Category, Health and Retirement Study, 2006-2016

Full Sample
(1) Non
Adopter
(2) Early
Adopter
(3) Late
Adopter
p for (1)
and (2)
p for (1)
and (3)
p for (2)
and (3) (64.6%) (15.2%) (20.2%)
%/Mean (SD) %/Mean (SD) %/Mean (SD) %/Mean (SD)
Individual-Level Variable

Age, years 72.5 (7.9) 73.2 (8.2) 72.3 (7.3) 70.2 (6.7) 0.00 0.00 0.00
Female 56.3% 56.2% 58.9% 54.7% 0.10 0.31 0.03
Race/Ethnicity

Non-Hispanic Whites 84.3% 81.6% 92.9% 86.1% 0.00 0.00 0.00
Non-Hispanic Blacks 7.6% 9.4% 2.2% 6.2% 0.00 0.00 0.00
Hispanics 6.0% 6.8% 2.8% 5.5% 0.00 0.06 0.00
Other 2.2% 2.2% 2.1% 2.2% 0.94 0.99 0.96
Married 60.9% 57.5% 67.3% 67.0% 0.00 0.00 0.86
Education

Less than High School 23.1% 27.7% 11.1% 17.7% 0.00 0.00 0.00
High School 33.1% 34.1% 27.7% 33.9% 0.00 0.88 0.00
Some College or above 43.7% 38.2% 61.2% 48.4% 0.00 0.00 0.00
Logged Household Income 10.6 (1.0) 10.4 (0.9) 10.9 (1.0) 10.7 (0.9) 0.00 0.00 0.00
Has Prescription Drug Plans 93.9% 92.6% 96.3% 96.2% 0.00 0.00 0.84
Self-reported Poor Health 26.3% 31.1% 15.6% 18.8% 0.00 0.00 0.03
Received Flu Vaccine 72.1% 65.5% 89.6% 80.0% 0.00 0.00 0.00
Conscientious Personality 3.4 (0.5) 3.3 (0.5) 3.5 (0.4) 3.4 (0.5) 0.00 0.00 0.00
Social Interaction with Family/Children 3.7 (0.9) 3.6 (0.9) 3.8 (0.9) 3.7 (0.9) 0.00 0.00 0.02
Social Interaction with Friends 3.7 (1.1) 3.6 (1.1) 4.0 (1.0) 3.8 (1.0) 0.00 0.00 0.00
Area-Level Variable

Shingles Vaccination Rate in HRR

1st quartile (0% - 26.3%) 23.2% 27.0% 13.4% 18.5% 0.00 0.00 0.00
2nd quartile (26.7% - 33.3%) 24.9% 24.9% 23.8% 25.8% 0.45 0.46 0.24
3rd quartile (33.7% - 39.8%) 25.1% 23.4% 26.5% 29.6% 0.03 0.00 0.09
4th quartile (40% - 100%) 26.7% 24.7% 36.2% 26.1% 0.00 0.30 0.00
N 9,934 6,586 1,445 1,903

83

Table 2. Multinomial Logistic Regression Models Predicting Early Vaccine Adopters and Late
Vaccine Adopters

Early adopters vs Late adopters vs Early adopters vs
Non-adopters (ref) Non-adopters (ref) Late adopters (ref)
RRR 95% CI RRR 95% CI RRR 95% CI
Individual-Level Variable

Age, years 0.98 (0.97,0.99) 0.94 (0.93,0.95) 1.04 (1.03,1.05)
Female 1.15 (0.99,1.34) 0.98 (0.86,1.12) 1.17 (0.99,1.39)
Race/Ethnicity

Non-Hispanic Whites

Non-Hispanic Blacks 0.36 (0.25,0.52) 0.82 (0.66,1.02) 0.44 (0.30,0.65)
Hispanics 0.70 (0.45,1.07) 1.09 (0.83,1.43) 0.64 (0.40,1.01)
Other 1.03 (0.63,1.71) 1.04 (0.66,1.62) 1.00 (0.57,1.75)
Married 1.04 (0.88,1.24) 1.02 (0.88,1.19) 1.02 (0.83,1.24)
Education

Less than High School

High School 1.30 (1.04,1.62) 1.18 (0.99,1.40) 1.10 (0.86,1.42)
College or above 2.00 (1.61,2.50) 1.26 (1.05,1.50) 1.59 (1.24,2.05)
Logged Household Income 1.27 (1.14,1.41) 1.12 (1.03,1.21) 1.14 (1.01,1.28)
Has Prescription Drug Plan 1.42 (0.98,2.07) 1.77 (1.26,2.47) 0.81 (0.50,1.28)
Self-reported Poor Health 0.59 (0.49,0.71) 0.62 (0.53,0.73) 0.95 (0.76,1.18)
Received Flu Vaccine 4.68 (3.73,5.88) 2.57 (2.20,3.01) 1.82 (1.41,2.35)
Conscientious Personality 1.43 (1.21,1.69) 1.12 (0.98,1.29) 1.27 (1.05,1.53)
Social Interaction with
Family/Children 1.10 (1.02,1.20) 1.06 (0.99,1.14) 1.04 (0.95,1.14)
Social Interaction with Friends 1.16 (1.07,1.25) 1.08 (1.01,1.15) 1.07 (0.98,1.17)
Area-Level Variable

Quartile of Shingles Vaccination Rate

1st quartile

2nd quartile 1.90 (1.52,2.37) 1.48 (1.23,1.77) 1.28 (0.99,1.66)
3rd quartile 1.96 (1.57,2.45) 1.70 (1.42,2.04) 1.15 (0.90,1.48)
4th quartile 2.28 (1.84,2.82) 1.36 (1.13,1.63) 1.68 (1.31,2.15)
N 9,934
84

Chapter 5. Conclusion
The three chapters in this dissertation significantly contribute to our understanding of
various pathways to the incidence of shingles and factors affecting the acceptance and timing of
shingles vaccination among older Americans. This work will have implications for the public
health agendas that aim to control risk factors for diseases and promote usage of effective health
care for improving health of the older population.
The first study examines the associations between reported stress and shingles incidence
by incorporating event-based and chronic stress from multiple life domains. We found that
problems reflecting chronic stress were associated with an increased risk of shingles even after
adjustments of socio-demographic, health, and vaccination factors. However, stressful events were
not associated with shingles onset. The findings suggest that event-based and chronic stressors
operate differently in the development of shingles. The second study examines the independent
and interdependent role of age, chronic diseases, and chronic inflammation in causing shingles.
We found that chronic diseases account for age differentials in shingles risk in a full model that
includes age, chronic disease, and chronic inflammation all, indicating that older adults’ higher
susceptibility to shingles is mainly driven by chronic inflammation and chronic diseases, not age
alone. The third study investigates the determinants of the acceptance and timing of shingles
vaccination. Findings indicate that early adopters were more likely than late adopters to be older,
non-Hispanic white, socioeconomically advantaged, conscientious, recipients of flu vaccine, and
residents of an area with a higher shingles immunization rate. This study provides the first
empirical evidence showing that factors associated with early adopters are significantly different
from the factors associated with late adopters.
85

Given the substantial burden of shingles in the older population and health care system,
this work addresses multifaceted factors that may affect the variability of shingles experience in
the older population. First of all, this work revisits the documented potential risk factors of shingles
in prior studies and determined its temporal directionality using a longitudinal framework. For
example, several chronic diseases were considered risk factors of shingles as shingles frequently
occur among older adults with chronic diseases in cross-sectional studies (Kawai & Yawn, 2017;
Marra et al., 2020). We found that heart disease and arthritis are associated with an increased risk
of incidence of shingles when age, inflammation, and other covariates are controlled in a panel
data. Similarly, prior case-control studies suggest the possible associations between stress and
shingles (Marin et al., 2016, Schmader et al., 1990; Lasserre et al., 2012). Because shingles itself
is a painful and stressful experience, having the disease can distort their retrospective reporting of
perceived stress before the onset of shingles (Raphael, 1987). To avoid potential recall bias from
reporting stress experience, we relied on sample members who did not have shingles at the baseline
interview and followed up their shingles status prospectively. Therefore, the evidence of the null
effect of event-based stress and positive effect of chronic stress on incidence of shingles in our
study warrants the directionality between stress and shingles. Secondly, we explored complex
mechanisms of the risk factors of incidence of shingles by exploring disentangled or intertwined
pathways. In the first chapter where we examine the association between stress and incidence of
shingles, we disentangled stressors into stressful events and chronic problems that arise from
multiple life domains. This disentangling process of stress is important to understand various
sources of stress exposure in older populations that may differentially shape shingles risks (Epel
et al., 2018). In the efforts of identifying risk factors of shingles, the second chapter integrates age
and chronic diseases, which were often examined as single pathways (Kawai & Yawn, 2017;
86

Marra et al., 2020). The integrative approach for age, chronic diseases, and chronic inflammation
reflects the overlap and intertwined physiological and morbid process underpinning
immunosenescence. Thirdly, this work not only explores risk factors of incidence of shingles but
also investigates the usage of a new medical technology that lowers shingles risks. Although the
shingles vaccine is effective in preventing the disease, a lower vaccination rate in the older
population puts older adults at risk of shingles. Achieving a high level of vaccination rate in the
early phase of the vaccine dissemination is important to influence and motivate the remaining
unvaccinated population (Deshpandé et al., 2021). Despite the recognition of vaccination as a time-
sensitive goal, previous empirical studies that relied on cross-sectional data have been unable to
explore the factors affecting the uptake of vaccine over time (Lu et al., 2007; 2017, Lee et al., 2013;
Tseng et al., 2012; Zhang et al., 2012; 2017, Langan et al., 2013; Shen et al., 2019; Vogelsang et
al., 2021). The differential characteristics of early and late vaccine recipients serve as important
scientific resource that establishes a tailored strategy to reach out to different target groups of
vaccination. Finally, an additional strength of this work is that it uses a nationally representative
sample of older adults, which generalizes our findings to community-living older Americans.
Additionally, the data allows to estimate the national burden of shingles, which addresses
limitations in previous estimates for shingles incidence derived from unvaccinated members of
regional health care organization (Tseng et al., 2020). The incidence rate for shingles is 26.2 per
1,000 person-intervals from a national sample of unvaccinated and vaccinated adults aged 50 and
older over two years, which will serve as foundational data for surveillance on trends and changes
in shingles incidence among older Americans.
To sum up, this dissertation takes a multidimensional approach to assess factors that affect
older adults' likelihood of getting shingles. This work identifies a wide range of risk factors of
87

shingles, including psychological stress, age, chronic diseases, and chronic inflammation.
Importantly, we explored the usage of new medical technology for preventing shingles available
to older adults. Considering the conditions older adults live and age are the sources of risk factors
for shingles (e.g., stress and chronic diseases), researchers should pay attention to the fundamental
cause of shingles risk factors within the framework of social determinants of health (Braveman et
al., 2011). Advanced medical technology such as vaccines helps older adults to maintain healthy
later in life, but its usage is low (Gierke et al. 2021). Strategies to promote faster and broader usage
of effective health care measures should consider addressing barriers in late or non-adoption,
which would have a substantial public health impact in the older population.
Future Directions
I propose following suggestions for future studies. First, although this dissertation
demonstrates the adverse role of chronic stressors in incidence of shingles, the present work could
not fully reflect the stress process due to data limitation. Stress literature emphasizes that
individuals can react differently to stressful situations, highlighting the importance of stress
appraisal in health outcomes (Perlin et al., 1981; McLeod, 2012). While we include
sociodemographic and health factors that may affect the meaning of stressors, the present work
focuses on exposures and does not include the measures of how stressful situations are perceived
to be. Older adults’ psychological resilience toward life adversity has been well-established (Kim
et al., 2021). Therefore, including stress appraisal in stressful events and chronic problems will
advance our understanding on stress and health in the older population. Second, this work provides
important evidence that age and age-related health conditions measured by chronic diseases and
inflammation are significant contributors to shingles onset, yet, the inclusion of more
comprehensive biological markers reflecting in the aging process has not been possible. Age-
88

related changes in immune function have been suggested to be the main pathway to shingles in
older adults (Oxman et al., 2009). To understand why older adults are more disproportionately
affected by shingles than their younger counterparts, future work is needed to incorporate
indicators of biological mechanisms that are related to immunosenescence at the time of shingles
incidence. Third, this dissertation focuses on the factors associated with the timing and acceptance
of the initial shingles vaccine approved in 2006 among adults aged 60 and older. A newer vaccine
has been recommended for adults aged 50 and older since 2017, regardless of vaccination status
in the past. A newer vaccine is more effective in preventing the disease, but possible barriers to
the vaccine uptake include a two-dose schedule and limited coverage in those who are yet eligible
to Medicare. Examining vaccine uptake for a newer shingles vaccine stratified by early, late, and
non-adopter of the first vaccine will likely present a dynamics of vaccination behaviors in older
adults who encounter updated vaccines over time.

89

Bibliography
Braveman, P., Egerter, S., & Williams, D. R. (2011). The social determinants of health: coming of
age. Annual review of public health, 32, 381-398.
Deshpandé R., Mintz O., & Currim I (2021). How Influencers, Celebrities, and FOMO Can Win
Over Vaccine Skeptics. Harvard Business School. https://hbswk.hbs.edu/item/how-influencers-
celebrities-and-fomo-can-win-over-vaccine-skeptics#commentsAnchor
Epel, E. S., Crosswell, A. D., Mayer, S. E., Prather, A. A., Slavich, G. M., Puterman, E., & Mendes,
W. B. (2018). More than a feeling: A unified view of stress measurement for population science.
Frontiers in neuroendocrinology, 49, 146-169.
Gierke R., Patricia W., Kobayashi M., (2015). Pneumococcal Disease. Epidemiology and
Prevention of Vaccine-Preventable Diseases.
Kawai, K., & Yawn, B. P. (2017, December). Risk factors for herpes zoster: a systematic review
and meta-analysis. In Mayo clinic proceedings (Vol. 92, No. 12, pp. 1806-1821). Elsevier.
Kim, E. S., Tkatch, R., Martin, D., MacLeod, S., Sandy, L., & Yeh, C. (2021). Resilient Aging:
Psychological Well-Being and Social Well-Being as Targets for the Promotion of Healthy
Aging. Gerontology and Geriatric Medicine, 7, 23337214211002951.
Langan, S. M., Smeeth, L., Margolis, D. J., & Thomas, S. L. (2013). Herpes zoster vaccine
effectiveness against incident herpes zoster and post-herpetic neuralgia in an older US population:
a cohort study. PLoS medicine, 10(4), e1001420.
90

Lasserre, A., Blaizeau, F., Gorwood, P., Bloch, K., Chauvin, P., Liard, F., ... & Hanslik, T. (2012).
Herpes zoster: family history and psychological stress—case–control study. Journal of clinical
virology, 55(2), 153-157.
Lee, T. J., Hayes, S., Cummings, D. M., Cao, Q., Carpenter, K., Heim, L., & Edwards, H. (2013).
Herpes zoster knowledge, prevalence, and vaccination rate by race. The Journal of the American
Board of Family Medicine, 26(1), 45-51.
Lu, P. J., Euler, G. L., Jumaan, A. O., & Harpaz, R. (2009). Herpes zoster vaccination among
adults aged 60 years or older in the United States, 2007: uptake of the first new vaccine to target
seniors. Vaccine, 27(6), 882-887.Lu, P. J., O’Halloran, A., Williams, W. W., & Harpaz, R. (2017).
National and state-specific shingles vaccination among adults aged≥ 60 years. American journal
of preventive medicine, 52(3), 362-372. Marin, M., Harpaz, R., Zhang, J., Wollan, P. C., Bialek,
S. R., & Yawn, B. P. (2016, May). Risk factors for herpes zoster among adults. In Open forum
infectious diseases (Vol. 3, No. 3, p. ofw119). Oxford University Press.
Marra, F., Parhar, K., Huang, B., & Vadlamudi, N. (2020, January). Risk factors for herpes zoster
infection: a meta-analysis. In Open forum infectious diseases (Vol. 7, No. 1, p. ofaa005). US:
Oxford University Press.
McLeod, J. D. (2012). The meanings of stress: Expanding the stress process model. Society and
Mental Health, 2(3), 172-186.
Oxman, M. N. (2009). Herpes zoster pathogenesis and cell-mediated immunity and
immunosenescence. The Journal of the American Osteopathic Association, 109(6_suppl_2), S13-
S17.
91

Pearlin, L. I., Lieberman, M. A., Menaghan, E. G., & Mullan, J. T. (1981). The stress process.
Journal of Health and Social Behavior, 22, 337–356. doi:10.2307/2136676
Raphael, K. (1987). Recall bias: a proposal for assessment and control. International journal of
epidemiology, 16(2), 167-170.
Schmader, K., Studenski, S., MacMillan, J., Grufferman, S., & Cohen, H. J. (1990). Are stressful
life events risk factors for herpes zoster?. Journal of the American Geriatrics Society, 38(11), 1188-
1194.
Shen, A. K., Warnock, R., Selna, W., MaCurdy, T. E., Chu, S., & Kelman, J. A. (2019).
Vaccination among Medicare-fee-for service beneficiaries: Characteristics and predictors of
vaccine receipt, 2014–2017. Vaccine, 37(9), 1194-1201.
Tseng, H. F., Chi, M., Smith, N., Marcy, S. M., Sy, L. S., & Jacobsen, S. J. (2012). Herpes zoster
vaccine and the incidence of recurrent herpes zoster in an immunocompetent elderly population.
The Journal of infectious diseases, 206(2), 190-196.
Tseng, H. F., Bruxvoort, K., Ackerson, B., Luo, Y., Tanenbaum, H., Tian, Y., ... & Sy, L. S. (2020).
The epidemiology of herpes zoster in immunocompetent, unvaccinated adults≥ 50 years old:
incidence, complications, hospitalization, mortality, and recurrence. The Journal of infectious
diseases, 222(5), 798-806.
Vogelsang, E. M., & Polonijo, A. N. (2021). Social Determinants of Shingles Vaccination in the
United States. The Journals of Gerontology: Series B.
Zhang, D., Johnson, K., Newransky, C., & Acosta, C. J. (2017). Herpes zoster vaccine coverage
in older adults in the US, 2007–2013. American journal of preventive medicine, 52(1), e17-e23.
92

Zhang, J., Xie, F., Delzell, E., Chen, L., Winthrop, K. L., Lewis, J. D., ... & Curtis, J. R. (2012).
Association between vaccination for herpes zoster and risk of herpes zoster infection among older
patients with selected immune-mediated diseases. Jama, 308(1), 43-49.

Abstract (if available)

Abstract This dissertation examines multifaceted factors affecting older adults’ shingles experience that can be prevented and controlled by identifying risk factors for incidence of shingles and factors affecting the timing of shingles vaccine uptake. ❧ A growing literature has linked psychological and biomedical conditions in old age to shingles, offering potential mechanisms through which older adults experience the disease. However, prior studies at the population level fail to appropriately conceptualize and measure psychological stress, leading to inconsistent findings. Indeed, previous research overlooks that age and age-related health conditions are highly interrelated, providing a limited insights on the relative importance of a set of biological and morbid conditions associated with age on incidence of shingles. Therefore, it remains unclear how a broad range of age-related psychological, biological, and morbid characteristics is associated with the incidence of shingles. Further, this dissertation recognizes that the dichotomous view on shingles vaccine uptake or not in previous literature has limitations in examining the heterogeneity within vaccine recipients. Understanding factors affecting early or late vaccine uptake is critical to tackling slow dissemination of shingles vaccination in older adults. ❧ This dissertation provides important epidemiological and public health findings in a nationally representative sample of older adults, which significantly contribute to shingles literature. The first study examines the associations between reported stress and shingles incidence by incorporating event-based and chronic stress from multiple life domains. Research suggests that problems reflecting chronic stress were associated with the incidence of shingles, whereas stressful events were not associated with shingles onset. The results demonstrate that psychological stress triggers shingles, and types of stress differentially operate in shingles risks. The second study integrates age and age-related health conditions, including chronic diseases and chronic inflammation, to observe its relative contribution to the incidence of shingles. We found that age, chronic disease, and chronic inflammation are significantly associated with increased risk of shingles. However, chronic diseases account for age differentials in shingles risk in the model that combined age, chronic disease, and chronic inflammation. The findings imply that age-related physiological and morbid conditions are the main drivers of increased risk for shingles incidence in older adults, which challenges the interpretation of age as a strong risk factor for shingles. The third study examines how individual and areal-level factors affect the acceptance and timing of shingles vaccination over ten years. Findings suggest significant differences between early and late vaccine recipients in terms of socio-demographic, health care, personality, and geography factors. ❧ This dissertation lays an empirical foundation for preventing and controlling factors that affect older adults’ likelihood of getting shingles. The adverse effect of chronic life problems on the incidence of shingles in this work calls for a better understanding of the conditions where older adults live and age, in addition to stress management. Considering chronic disease and chronic inflammation are common aging conditions, health promotion has the potential to retard shingles onset among older adults. Additionally, this research uncovers the disparity in early shingles vaccine uptake by race/ethnicity, socioeconomic factors, and geography, which should be acknowledged in vaccination promotion strategies for a newer shingles vaccine.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Social determinants of physiological health and mortality in China

PDF

Mechanisms of stress effects on learning and decision making in younger and older adults

PDF

A biodemographic approach to understanding sociodemographic disparities in kidney functioning on three dimensions: individual, population, and cross-national

PDF

Cognitive health and self-rated memory in later life: linkages to race/ethnicity, multimorbidity, and survival

PDF

Is stress exposure enough? Race/ethnic differences in the exposure and appraisal of chronic stressors among older adults

PDF

Self-perceptions of Aging in the Context of Neighborhood and Their Interplay in Late-life Cognitive Health

PDF

Statistical algorithms for examining gene and environmental influences on human aging

PDF

Family caregiver well-being: generational differences and coping strategies

PDF

Biomarkers of age-related health changes: associations with health outcomes and disparities

PDF

Internet communication use, psychological functioning and social connectedness at older ages

PDF

Age integration in late life: sociodemographic & psychosocial correlates of intergenerational-only, peer-only, and age-integrated social networks

PDF

Eliciting perspectives on palliative care: outpatient visits, advance care planning, and the impact of COVID-19

PDF

Exploring geography and multi-level factors associated with disparities in HPV vaccination among Latino adolescents in Los Angeles County

PDF

Palliative care: breaking the knowledge barrier among community-based older adults

PDF

Cardiovascular disease risk factors and cognitive function

PDF

Estimating survival in the face of pain: evidence from the health and retirement study

PDF

The effects of a fasting mimicking diet (FMD) on mouse models of Alzheimer's and Parkinson's disease

PDF

How emotional arousal influences memory and learning in younger and older adults

PDF

TLR4-mediated innate immune response and neuroinflammation: focus on APOE and obesity

PDF

Genetic and environmental risk factors for childhood cancer

Asset Metadata

Creator Kang, Hyewon (author)

Core Title Three essays on modifiable determinants of shingles: risk factors for shingles incidence and factors affecting timing of vaccine uptake

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Gerontology

Publication Date 12/03/2021

Defense Date 09/20/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Age,chronic diseases,chronic inflammation,OAI-PMH Harvest,shingles,Stress,vaccine

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Ailshire, Jennifer (committee member), Ho, Jessica (committee member)

Creator Email hyewonka@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC17895966

Unique identifier UC17895966

Legacy Identifier etd-KangHyewon-10272

Document Type Dissertation

Format application/pdf (imt)

Rights Kang, Hyewon

Type texts

Source 20211208-wayne-usctheses-batch-902-nissen (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu