Anda di halaman 1dari 8

Global Sensory Impairment in Older Adults in the United States

Camil Correia, BS,*a Kevin J. Lopez, BS,†a Kristen E. Wroblewski, MS,‡ Megan Huisingh-Scheetz,
MD, MPH,§¶ David W. Kern, PhD,** Rachel C. Chen, BS,* L P. Schumm, MA,‡¶
William Dale, MD, PhD,§¶ Martha K. McClintock, PhD,¶**†† and Jayant M. Pinto, MD†¶

Key words: sensory function; aging; hearing; vision;

OBJECTIVES: To determine whether there may be a com- smell; taste; touch
mon mechanism resulting in global sensory impairment of
the five classical senses (vision, smell, hearing, touch, and
taste) in older adults.
DESIGN: Representative, population-based study.
SETTING: National Social Life, Health, and Aging
PARTICIPANTS: Community-dwelling U.S. adults aged
57 to 85.
A ging has long been associated with decline in sensory
function, a critical component of the health and qual-
ity of life of older people.1 Prior work has demonstrated
MEASUREMENTS: The frequency with which impair- that olfactory loss is associated with cognitive decline,
ment co-occurred across the five senses was estimated as highlighting its importance as an early warning sign of
an integrated measure of sensory aging. It was hypothe- neurological decline with its attendant morbidity and com-
sized that multisensory deficits would be common and promised physical function,2–4 and that it strongly and
reflect global sensory impairment that would largely independently predicts all-cause 5-year mortality.2,5 These
explain the effects of age, sex, and race on sensory dys- and other data are consistent with the idea that sensory
function. function is a critical component of health and life itself.
RESULTS: Two-thirds of subjects had two or more sen- Data from studies of single sensory deficits support
sory deficits, 27% had just one, and 6% had none. this concept. For example, vision impairment is correlated
Seventy-four percent had impairment in taste, 70% in with depression, poor quality of life, cognitive decline, and
touch, 22% in smell, 20% in corrected vision, and 18% in mortality.6–8 Hearing loss is associated with slower gait
corrected hearing. Older adults, men, African Americans, speed9 (a marker of physical decline), poor cognition, and
and Hispanics had greater multisensory impairment (all mortality.10 Like smell, taste has been associated with
P < .01). Global sensory impairment largely accounted for nutritional compromise11 and in patient mortality,12 sug-
the effects of age, sex, and race on the likelihood of gesting that chemosensory function is critical. Tactile dis-
impairment in each of the five senses. crimination declines with age13 due to the cumulative
CONCLUSION: Multisensory impairment is prevalent in effects of decreased nerve conduction velocity,14 decreased
older U.S. adults. These data support the concept of a com- density of Meissner’s and Pacinian corpuscles, and gray
mon process that underlies sensory aging across the five matter changes in the central nervous system and is also
senses. Clinicians assessing individuals with a sensory deficit associated with cognitive decline.15
should consider further evaluation for additional co-occur- Individual sensory impairments are common. The
ring sensory deficits. J Am Geriatr Soc 64:306–313, 2016. prevalence of hearing loss (33%) and vision impairment
(18%) is high in older adults (≥70).16,17 Similarly, deficits
in smell (24%)18,19 and taste (up to ~61%) are widely
From the *Pritzker School of Medicine; †Section of Otolaryngology—
Head and Neck Surgery; ‡Department of Public Health Sciences;
prevalent in adults aged 70 and older.20 Impairment in
Section of Geriatric Medicine and Palliative Care; ¶Center on touch is noted in adults as young as 55.21 These sensory
Demography and Economics of Aging; **Department of Comparative losses have a major effect on how older adults live and
Human Development; and ††Institute for Mind and Biology, University of function, often with profound consequences. As the popu-
Chicago, Chicago, Illinois.
lation ages, these burdens will grow.
These authors contributed equally to this work. Despite these important consequences, little is known
Address correspondence to Jayant M. Pinto, 1035, Section of OHNS, about the prevalence of deficits in multiple sensory sys-
5841 S. Maryland Avenue, Chicago, IL 60637. E-mail: jpinto@surgery. tems, their combined effect, or common mechanisms that
drive the underlying biology. Although some studies have
DOI: 10.1111/jgs.13955

JAGS 64:306–313, 2016

© 2016, Copyright the Authors
Journal compilation © 2016, The American Geriatrics Society 0002-8614/16/$15.00

measured the prevalence of concurrent poor vision and population of community-dwelling adults born between
hearing22 (dual sensory impairment), to the knowledge of 1920 and 1947 (aged 57–85) in the United States.34,35 The
the authors of the current study, none has measured other NORC and University of Chicago institutional review
senses in an elderly population. The importance of consid- boards approved this study, and all respondents provided
ering simultaneous impairments is clear from studies of written, informed consent.
vision and hearing, which have shown that dual losses NSHAP used a modular study design. All respondents
interfere synergistically with independent function, presage were administered a core interview and provided a stan-
cognitive decline, and signal greater mortality.23–25 dard set of biomeasures in their homes, including olfaction
The close connection between these various sensory and gustation testing and hearing assessment. Half the
deficits, cognitive decline, and even death suggest the pos- respondents were also randomized to receive vision and
sibility that global sensory decline, define here as a com- touch testing.30 The analytical sample included the 2,968
mon physiological process underlying deterioration of the respondents who had data on two or more of the five
classical senses, is an early indicator of neurodegenera- senses (Table 1). Of the 1,506 respondents eligible to
tion,2,12,20,24 with attendant poor social and health out- receive all sensory modules, 1,301 (86%) had complete
comes.1 Additionally, frequent associations between health data.30 Race and Hispanic ethnicity were measured using
outcomes and each of the different five classical senses the standard National Institutes of Health items, as
may reflect common mechanisms underlying the effects of reported previously,36 and respondents were coded as
aging on these systems. These could include peripheral white (non-Hispanic), black, Hispanic (excluding those
nerve dysfunction, changes in sensory integration at the who self-identified as black), and other. Respondents were
central level,26 lack of regenerative capacity,27 and sec- asked whether they had received a physician’s diagnosis of
ondary metabolic effects (e.g., consequences of atheroscle- diabetes mellitus, stroke, heart failure, hypertension, or
rosis or lipidemia).28 Finally, because dual sensory myocardial infarction and were asked to rate their overall
impairment has been shown to have greater effects on physical health using the standard categories excellent,
function than single deficits,24 one would expect multisen- very good, good, fair, and poor. Data missing for one or
sory impairment (impairment of more than two senses) to more sensory measures because of the study design were,
cause even more detrimental health effects. Despite this, by definition, missing completely at random and therefore
to the knowledge of the authors, no study has examined did not introduce any bias into the analysis (only a loss of
multiple sensory impairments in a national population precision, relative to a design in which all respondents
sample. were administered all items). Although there was some
To address multisensory impairment, data were ana- item nonresponse because of respondent refusal (or
lyzed from the National Social Life, Health, and Aging responses of don’t know) for each sensory dysfunction
Project (NSHAP), a longitudinal population-based study of item, this nonresponse was in general low, limiting the
adults aged 57–85 that collected extensive health and magnitude of any potential bias.
social measures through in-home interviews and respon-
dent-administered questionnaires.29 Sensory function was
assessed in all five classical senses,30 and respondents were Sensory Function
asked about their physical and mental health, medication
use, cognition, and health behaviors. NSHAP and sec- Vision
ondary analyses of these data have provided insights into a Participants wore their usual glasses or contact lenses, and
number of aspects of aging.31–33 NSHAP offers a unique corrected distance visual acuity was assessed under home
opportunity to examine sensory function broadly. lighting conditions using a Snellen chart test with a stan-
The prevalence of multisensory impairment was esti- dardized protocol.30 Corrected vision was chosen to deter-
mated, and a model of global sensory impairment was mine the actual functional level respondents experienced in
developed based on the interrelationships between mea- daily life, consistent with prior benchmark studies.37 Cate-
sures of all five senses. It was hypothesized that multisen- gories for visual acuity corresponded to those required for
sory deficits would be common in older adults, more a driver’s license (good = 20/40 or better, fair =20/50 to
prevalent in men and minorities, and occur more often 20/63, poor = worse than 20/63).
with increasing age. The concept of global sensory impair-
ment is introduced, a process that it was hypothesized Touch
would largely account for the effects of age, sex, and race
on the likelihood of impairment in each of the five senses. Tactile sensitivity was assessed using a 2-point discrimina-
tion test on the index finger of the dominant hand with
eyes closed.30 Three 2-point tests were conducted at inter-
METHODS point distances of 12, 8, and 4 mm and a single point after
the 12-mm test. A 4-mm threshold (good) was defined as
NSHAP Study Design correctly identifying two points at all three distances plus
the single point test; an 8-mm threshold (fair) was cor-
Respondents rectly identifying two points at 12 and 8 mm plus the sin-
In 2005–06, interviewers from the National Opinion gle point test but not 4 mm, and a 12-mm threshold
Research Center (NORC) conducted in-home interviews (poor) was correctly identifying two points only at 12 mm
with 3,005 community-dwelling older adults (1,454 men, plus the single point test. All other response patterns were
1,551 women) identified using a probability sample of the considered nondiscriminating and also categorized as poor.

tried, unable to do; refused; or don’t know were counted

Table 1. Demographic and Health Characteristics of
as incorrect. Two to four errors was categorized as poor
the U.S. Population of Home–Dwelling Older Adults
Based on the National Social Life, Health and Aging (ageusic), one error as fair (hypogeusic), and no errors as
Project (2005–06) (N = 2,968) good (normogeusic).

Characteristic % Hearing
Age The field interviewer assessed respondents’ conversational
57–64 41.4 hearing during the interview afterward on a 5-point scale
65–74 34.9 (1 = practically deaf, 5 = normal hearing).30 Scores of 4 or
75–85 23.7 5 were categorized as good, 3 as fair, and 1 or 2 as poor.
Male 48.6 Respondents who chose to wear hearing aids during the
Race and ethnicity (n = 2,956)
White 80.7
interview were permitted to do so but not required. Time
Black 9.9 and resource constraints of the omnibus survey precluded
Hispanic (non black) 6.9 the use of additional psychophysical measures of hearing
Other 2.5 (e.g., audiometry).
Self-rated physical health (n = 2,957)
Poor 6.8
Fair 17.9 Statistical Analysis
Good 29.6 Estimates of the prevalence of impairment (defined as hav-
Very good 32.6
ing fair or poor function) for each sense and of the distri-
Excellent 13.1
Comorbid diseases bution of the total number of impairments among the U.S.
Hypertension 53.9 national population of home-dwelling older adults (aged
Diabetes mellitus 19.9 57–85) were obtained by using the sampling weights pro-
Heart attack 11.7 vided with the dataset to account for differential probabili-
Heart failure 8.3 ties of selection and nonresponse, as previously
Stroke 8.2 described.35 Estimates of the population prevalence of sev-
Sensory function (good/fair/poor) eral comorbid diseases and of the distribution of self-rated
Hearing (n = 2,968) 82.0/12.8/5.3
physical health and the demographic variables age, sex,
Visiona (n = 1,417) 80.3/13.6/6.1
Smell (n = 2,939) 77.8/18.8/3.5 and race and ethnicity are also presented.
Toucha (n = 1,464) 30.4/37.7/31.8 Ordinal probit regression was used to model the rela-
Taste (n = 2,735) 26.0/25.8/48.2 tionship between sensory dysfunction (good, fair, poor)
Number of impairmentsb (n = 1,301) and age, sex, and race and ethnicity for each of the five
0 5.9 senses individually (Figure 1A). Ordinal probit regression
1 27.6 is a straightforward extension of the probit regression
2 38.1 model, with the (standard) normal distribution presumed
3 20.3
to underlie the response being split according to multiple
4 6.8
5 1.3 cutpoints (≤1 than the number of observed categories of
the outcome) instead of just one. Thus, the coefficients for
Estimates were weighted using the sample weights distributed with the the covariates have an interpretation identical to that of
dataset to yield population estimates of prevalence. those from probit regression—namely, as the change in
Measures administered to a randomly selected 50% of respondents. standard units of the underlying normal variable associated
An impairment was defined as fair or poor function; 1,301 respondents with a one-unit change in the covariate. A generalized sin-
had data on all five senses.
gle-factor measurement model (Figure 2B) was then fit to
the five observed sensory dysfunction measures, assuming
Smell a single latent variable (with variance equal to 1) capturing
global sensory impairment, which predicts each of the five
Olfaction was evaluated using a validated 5-item odor
sensory dysfunction measures through an ordinal probit
identification test with felt tip pens,38 as previously
regression.39 The proportion of variance in its underlying
described.36 A single odorant was presented, and respon-
distribution (as specified by the ordinal probit regression
dents were instructed to select one of four word or picture
model) explained by global sensory impairment was calcu-
choices, with refusals coded as incorrect. Four or five
lated for each dysfunction measure. This model was then
errors was considered poor (anosmic), two or three errors
expanded by specifying global sensory impairment to be a
as fair (hyposmic), and one or no errors as good (nor-
function of age (in decades), sex, and race and ethnicity
(white, black, Hispanic (nonblack), other) (Figure 2C).
Finally, this structural equation model (SEM) was aug-
mented by adding direct effects of the demographic covari-
Gustation was evaluated using four filter paper strips30 ates on each of the sensory dysfunction measures one at a
that were applied to the tongue in the following order: time (Figure 2D). Indirect effects of age on each dysfunc-
sour, bitter, sweet, and salty, with a sip of water between tion (through its effect on global sensory impairment) were
each application.30 Respondents were asked to describe calculated and compared with the direct effects of age.40
the taste using the same four descriptors. Responses of For all analyses, two-sided P ≤ .05 was considered

Age Gender


57-64 men
65-74 Fair




smell men
75-85 women


75-85 women

impairments taste


impairments taste
75-85 women

57-64 4-5
# of

# of
65-74 3

75-85 2

0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1
Estimated Prevalence Among Older U.S. Adults
Figure 1. Prevalence of sensory impairments for each of the five senses in community-dwelling U.S. older adults according to age
group and sex.

statistically significant. All analyses were performed using (Table 1). Sixty-seven percent of older adults had two or
Stata release 13.1 (StataCorp LP, College Station, TX). more sensory deficits, with two impairments being the
most common (38%) (Figure 1). These deficits were corre-
lated; for example, 34% had none or one, and 8% had
four or five, compared with the 28% and 6%, respectively,
that would be expected under the null hypothesis of inde-
Prevalence of Individual Sensory Impairments
pendence of the senses. Sixty-five percent had substantial
impairment (poor functioning) in at least one sense and
22% in two or more.
Taste impairment was the most prevalent sensory deficit,
with 74% of respondents having an impaired sense of taste
Global Sensory Impairment
(26% fair, 48% poor) (Table 1). Also prevalent was touch
impairment, estimated to be fair in 38% of older adults Each of the sensory outcomes was associated with a single
and poor in 32%. Fourteen percent had fair corrected dis- common factor (Table 3, Single-Factor Model; Figure 2B),
tance vision (20/50 to 20/63), and another 6% had poor with the strongest associations for vision and smell, fol-
corrected distance vision (20/80 or worse). Nineteen per- lowed by hearing. This factor explains a significant pro-
cent had a fair sense of smell, and 3% had a poor sense of portion of the variation in the underlying distributions of
smell. Thirteen percent had fair corrected hearing, and 5% the individual sensory deficits (hearing, 0.15; vision, 0.33;
had poor corrected hearing. smell, 0.30; touch, 0.08; taste 0.05). The effects of a  1-
standard deviation change in this factor on the actual
probability of each deficit are illustrated in Figure S2.
Association with Age, Sex, and Race and Ethnicity
Global sensory impairment (the common factor) was
Older people had worse function in all five senses, with strongly associated with age, sex, and race and ethnicity
the largest differences being for hearing, vision, and smell (Table 3, Structural Equation Model; Figure 2C). Consis-
(Table 2A, Figure 1). Men had worse hearing, smell, and tent with the individual results for each sense reported
taste but better corrected vision than women. Blacks and above, older age was associated with greater global sen-
Hispanics had worse sensory function than whites on all sory impairment, which was also higher for men than for
measures except for hearing, for which there was no evi- women and for blacks and Hispanics than whites.
dence of racial or ethnic differences, and for taste, for To test the hypothesis that global sensory impairment
which blacks had worse function than whites, but Hispan- accounts for much of the association between age and
ics had better function (Figure S1). individual sensory deficits and to examine the fit of the
model, direct effects of the demographic covariates on
each of the sensory deficits were added one at a time to
Prevalence of Co-Occurring Sensory Impairments
the structural equation model (Table 2B; Figure 2D). For
Sensory deficits were widely prevalent in older U.S. adults, vision, smell, and touch, the effect of age on global sen-
with 94% demonstrating at least one sensory deficit sory impairment explained most (if not all) of the

Figure 2. Visual representation of the analytical models: (A) Overall effects of age, sex, and race and ethnicity on each sensory
dysfunction without global sensory impairment, using smell as an example (Table 2). (B) Effects of global sensory impairment on
each of the five sensory dysfunction measures (Table 3, Single-Factor Model). (C) Effects of age, sex, and race and ethnicity on
global sensory impairment and through global sensory impairment on each sensory dysfunction measure (Table 3, Structural
Equation Model). (D) Direct effects of demographic variables on sensory dysfunction controlling for global sensory impairment,
using the effect of age on smell as an example (Table 2).

association with age, as judged by the fact that the direct sample of older adults and emphasizes the broad and
effects of age on these senses were not statistically signifi- prevalent sensory burden that this growing segment of the
cant. Only for hearing and taste were the direct effects of population faces.
age significant. For hearing, the direct effect of age was Prior studies have established that 6% of older adults
0.28 (slightly larger than the indirect effect through global have impaired vision and hearing.22 The current study
sensory impairment of 0.22), and for taste, the estimated results suggest that these same adults may also have addi-
direct effect of 0.21 reflects the fact that the association tional sensory impairments. Across all five senses, 38% of
between age and taste dysfunction is the weakest of all the older adults had two impairments and 28% had three or
senses (as noted above). Thus, global sensory impairment more. Twenty-two percent had substantial impairment
explained most of the association between age and the (poor function) in two or more sensory modalities, repre-
individual sensory dysfunction outcomes. senting a significant burden. Other recent studies of mul-
tiple sensory impairments support these findings and
suggest important associations with function and quality
of life with carefully measured sensory measures.41–43
Multisensory loss is remarkably common in older U.S. These studies did not focus exclusively on older adults;
adults and seems to be driven by a common underlying addressed representative populations; included touch; or
process. To the knowledge of the authors, this population- in one case,43 used objective measures. These and other
based study is the first to examine the full spectrum of sen- factors may explain variability in findings between
sory loss across the five classical senses in a representative studies.

Table 2. Estimated Effects of Age, Sex, and Race and Table 3. Effect of Global Sensory Impairment on Like-
Ethnicity on Individual Sensory Dysfunctions, Unad- lihood of Individual Sensory Dysfunctions and Its Asso-
justed and Adjusted for Global Sensory Impairment ciation with Age, Sex, and Race and Ethnicity
Hearing Vision Smell Touch Taste Structural Equation
Single-Factor Model Model
Variable Coefficient
Variable Coefficient P-Value Coefficient P-Value
Regression models showing the associations between demographic
characteristics and sensory dysfunctiona Sensory dysfunctiona
Age (per decade) 0.48e 0.41e 0.47e 0.21e 0.09d Hearing 0.43 <.001 0.40 <.001
Female 0.39 e
0.04 0.44e Vision 0.71 <.001 0.36 <.001
Race and ethnicity (reference white) Smell 0.65 <.001 0.48 <.001
Black 0.01 0.47e 0.53e 0.36e 0.16d Touch 0.29 <.001 0.19 <.001
Hispanic 0.06 0.39d 0.17c 0.65e 0.16c Taste 0.24 <.001 0.15 <.001
(nonblack) Demographic characteristicsb
Other 0.08 0.01 0.37c 0.03 0.01 Age 1.12 <.001
Associations between demographic characteristics and sensory (per decade)
dysfunction, holding constant global sensory impairmentb Female 0.67 <.001
Age (per decade) 0.28e 0.48 0.02 0.02 0.21c Race and ethnicity (reference white)
Female 0.30e 1.08d 0.06 0.15 0.32e Black 0.93 <.001
Race and ethnicity (reference white) Hispanic 0.46 .002
Black 0.28c 0.18 0.33d 0.22c 0.09 (nonblack)
Hispanic 0.07 0.13 0.02 0.60e 0.33d Other 0.50 .07
Other 0.04 0.42 0.04 0.11 0.13 Coefficients from ordinal probit regressions of each three-category sen-
sory dysfunction measure on the underlying factor (global sensory impair-
Ordinal probit regression models fit individually to each of the five sen- ment), each indicating the change in the likelihood (on the probit scale) of
sory dysfunctions. Coefficients for each covariate indicate the change in being above a given cutpoint associated with a 1–standard deviation
the likelihood (on the probit scale) of being above a given cutpoint associ- increase in the underlying factor.
ated with a 1-unit increase in the covariate. Coefficients indicate the change in the underlying factor associated with a
Direct effects of age, sex, and race and ethnicity on each of the five sen- 1-unit change in the demographic covariate (residual variance of the
sory dysfunctions, adjusting for global sensory impairment. Obtained by underlying factor is constrained to equal 1).
adding direct effects to the structural equation model in Table 3 separately
for each dysfunction.
P < c.05, d.01, e.001.
deficits (e.g., with hearing or vision loss or both) should
consider evaluation of the other senses, because it is highly
A single underlying factor, which was interpreted as likely that such individuals will have these undiagnosed
global sensory impairment, may explain a significant conditions. Individuals with multisensory impairment may
amount of the variation in each of the sensory dysfunc- be at higher risk of important sequelae such as neurode-
tions. This single factor accounts for much of the associa- generation and complications from falls, burns, food poi-
tion between age and each of the sensory impairments, soning, smoke inhalation, and others. If these other
suggesting a common process of sensory aging. There are conditions are identified, even in the face of limited treat-
several possible mechanisms, shared across the senses, ment options, mitigation through awareness, social inter-
which could link their deficits during aging: neurodegener- vention through family or caregiver support, or other
ation,20,44 secondary effects of common environmental means may be instituted. This burden of multisensory
insults,28,45 underlying genetics such as variation in genes impairment may affect people’s ability to manage social,
involved in nerve maintenance or innate immunity,46 coor- cognitive, and physical stresses, so attention to these con-
dinate cellular senescence, or combinations thereof. cerns is critical.
The concept of global sensory impairment also leads to There are several limitations of these findings. High
new ways of thinking about how other factors such as sex rates of multisensory impairment were found in the gen-
and race and ethnicity may affect sensory function through eral population of older adults living at home, but individ-
this common mechanism. Many studies have found associa- uals in a clinic or institutionalized setting may be at even
tions between sex, race, ethnicity, and individual sensory higher risk. Conversely, one recent study that included
impairment22,36 and have proposed mechanisms to explain objective measures of sensory function showed minimal
these. The results here differ in that they investigate the rela- multisensory impairment in adults younger than 45.42
tionship between these factors across all five senses. For Although corrected vision and hearing were measured, def-
example, other than for corrected vision, women seem to be icits in the home environment were still found, which
better protected from sensory loss than men, highlighting should prompt clinicians to be sensitive to the discrepancy
the prospect of a biological mechanism. The higher preva- between clinic- and home-based assessment of sensory
lence and severity of multisensory impairment in blacks is function and the consequences for care. For example,
especially troubling given the well-documented disparities clinic-based estimates of sensory function, under optimal
in access, treatment, and outcomes that they face. controlled conditions, may minimize the effect on daily life
There are important clinical implications of these data. because they do not account for the real-world experience
Clinicians who see patients with single or dual sensory of people in their own home environments.

In addition, this model does not explain everything

observed in the data, and there are some important
deviations, which are consistent with prior findings in the Some of these data were presented at the Association of
literature. For example, sex reliably predicted sensory dys- Chemoreception Sciences annual meeting, Bonita Springs,
function, with men being worse than women in all senses Florida, April 9, 2014.
except for corrected vision, in which women were worse. We thank Linda J. Waite, PhD, Elbert Huang, MD,
This difference may reflect a sex disparity in obtaining or MPH, Robert M. Naclerio, MD, Fuad M. Baroody, MD,
using adequate corrective lenses. Women may also have and members of the Geriatric Assessment Group and
faster rates of decline in vision than men, perhaps indicat- Olfactory Research Group of NSHAP for useful comments
ing an underlying susceptibility to age-related vision loss. provided freely. We gratefully acknowledge the participa-
This is troubling given previous work demonstrating that, tion of the NSHAP respondents.
in older women, vision impairment is a risk factor for cog- Conflict of Interest: The editor in chief has reviewed
nitive decline.23 Medicare fails to cover eyeglasses or con- the conflict of interest checklist provided by the authors
tact lenses, so this may also reflect lack of financial and has determined that the authors have no financial or
resources of women. any other kind of personal conflicts with this paper.
Finally, the integrated measure of hearing was based on This work was supported by the National Institute on
the interviewer’s assessment during social conversation in Aging (R37 AG030481; AG036762; AG029795; K23
the home. Subsequent work could expand on this study by AG036762), the National Institute of Allergy and Infec-
including audiometry as an objective measure of hearing tious Disease (Chronic Rhinosinusitis Integrative Studies
(e.g., at speech and other frequencies), although its inclusion Program, U01AI106683), the Institute for Translational
is challenging (but not impossible, for screening at least1) in Medicine at the University of Chicago (KL2RR025000),
large field studies without specialized personnel and equip- the McHugh Otolaryngology Research Fund, and the
ment (e.g., audiometer, sound booth). The measure of the American Geriatrics Society. DWK was supported by a
ability to hear conversational speech in the home is a major Mellon Foundation Social Sciences Dissertation-Year Fel-
strength of NSHAP in that the focus is on the typical envi- lowship.
ronment and social context that older adults experience in Author Contributions: Correia: data analysis, impor-
their everyday lives, in contrast to clinic- or hospital-based tant intellectual contributions, drafting the manuscript.
assessments. From this perspective, it is likely that the bur- Lopez: data collected and analysis, important intellectual
den of hearing loss in this population is underestimated. contributions. Wroblewski: statistical analyses, editing,
Similarly, the measures of sensory function could be important intellectual content including data interpreta-
enriched (e.g., the addition of near vision, olfactory sensitiv- tion, drafting figures and tables. Huisingh-Scheetz: study
ity, more precise measures), although it is unlikely that such design, important intellectual contributions, editing. Kern:
enriched measures would alter the main results. important intellectual contributions, data interpretation,
What are the implications of these findings? Because editing. Chen: important intellectual contributions.
of the critical nature of these sensory modalities in daily Schumm: analyses, intellectual input, editing. Dale: intel-
living, the results of multisensory loss may explain, in part, lectual input, editing. McClintock: design and oversight of
why older adults report poorer quality of life and chal- collection of NSHAP biomeasures, data acquisition, impor-
lenges in interacting with the environment and other peo- tant intellectual contributions, study design, data analysis,
ple.16 The current data also prompt further questions editing. Pinto: study concept and supervision, development
about the relationship between global sensory impairment of analysis strategy, data interpretation, editing. JMP had
and physical frailty and the concept of decline across mul- full access to all the data in the study and takes responsi-
tiple domains of physical performance, potentially includ- bility for the integrity of the data and the accuracy of the
ing shared physiological mechanisms.37 Are they data analysis.
independent processes, or does global sensory loss develop Sponsor’s Role: The funders had no role in the design,
simultaneously with physical frailty, worsening its effect methods, subject recruitment, data collections, analysis, or
(e.g., weight loss, falls, decreased physical activity) and preparation of this paper.
even increasing the risk of mortality? Given that older
adults face major changes in sensory perception, creating REFERENCES
deleterious effects on health and function, future work
should characterize global sensory impairment and exam- 1. Fischer ME, Cruickshanks KJ, Klein BEK et al. Multiple sensory impair-
ine the trajectory of its decline in longitudinal studies. Cur- ment and quality of life. Ophthalmic Epidemiol 2009;16:346–353.
2. Pinto JM, Wroblewski KE, Kern DW et al. Olfactory dysfunction predicts
rent measures of physical frailty have proven important in 5-year mortality in older adults. PLoS ONE 2014;9:e107541.
the management of the healthcare needs of older adults 3. Devanand DP, Michaels-Marston KS, Liu X et al. Olfactory deficits in
and an analogous consideration of multisensory impair- patients with mild cognitive impairment predict Alzheimer’s disease at fol-
ment may prove useful. low-up. Am J Psychiatry 2000;157:1399–1405.
4. Wilson RS, Schneider JA, Arnold SE et al. Olfactory identification and inci-
In summary, further examination of multisensory loss dence of mild cognitive impairment in older age. Arch Gen Psychiatry
in older adults will provide answers to these and other 2007;64:802–808.
gaps in knowledge and allow preventative measures to be 5. Wilson RS, Yu L, Bennett DA. Odor identification and mortality in old
designed or therapeutic targets in the underlying biology age. Chem Senses 2011;36:63–67.
6. Wang JJ, Mitchell P, Simpson JM et al. Visual impairment, age-related cat-
that is common to these five sensory modalities to be iden- aract, and mortality. Arch Ophthalmol 2001;119:1186–1190.
tified, with great promise for alleviating the suffering that 7. Rovner BW, Ganguli M. Depression and disability associated with
they cause. impaired vision: The MoVies Project. J Am Geriatr Soc 1998;46:617–619.

8. Harrabi H, Kergoat M-J, Rousseau J et al. Age-related eye disease and cog- 32. Kim J, Waite LJ. Relationship quality and shared activity in marital and
nitive function. Invest Ophthalmol Vis Sci 2015;56:1217–1221. cohabiting dyads in the National Social Life, Health, and Aging Project,
9. Li L, Simonsick EM, Ferrucci L et al. Hearing loss and gait speed among Wave 2. J Gerontol B Psychol Sci Soc Sci 2014;69B(Suppl 2):S64–S74.
older adults in the United States. Gait Posture 2013;38:25–29. 33. Qato DM, Alexander GC, Conti RM et al. Use of prescription and over-
10. Genther DJ, Betz J, Pratt S et al. Association of hearing impairment and the-counter medications and dietary supplements among older adults in the
mortality in older adults. J Gerontol Ser A Biol Sci Med Sci 2014;70A:85– United States. JAMA 2008;300:2867–2878.
90. 34. Suzman R. The National Social Life, Health, and Aging Project: An intro-
11. Schiffman SS, Graham BG. Taste and smell perception affect appetite and duction. J Gerontol B Psychol Sci Soc Sci 2009;64B(Suppl 1):i5–i11.
immunity in the elderly. Eur J Clin Nutr 2000;54:S54–S63. 35. O’Muircheartaigh C, Eckman S, Smith S. Statistical design and estimation
12. Solemdal K, Møinichen-Berstad C, Mowe M et al. Impaired taste and for the national social life, health, and aging project. J Gerontol B Psychol
increased mortality in acutely hospitalized older people. Chem Senses Sci Soc Sci 2009;64B(Suppl 1):i12–i19.
2014;39:263–269. 36. Pinto JM, Schumm LP, Wroblewski KE et al. Racial disparities in olfactory
13. Decorps J, Saumet JL, Sommer P et al. Effect of ageing on tactile transduc- loss among older adults in the United States. J Gerontol A Biol Sci Med Sci
tion processes. Ageing Res Rev 2014;13:90–99. 2014;69A:323–329.
14. Besne I, Descombes C, Breton L. Effect of age and anatomical site on den- 37. Jacobs JM, Hammerman-Rozenberg R, Maaravi Y et al. The impact of
sity of sensory innervation in human epidermis. Arch Dermatol visual impairment on health, function and mortality. Aging Clin Exp Res
2002;138:1445–1450. 2013;17:281–286.
15. Yang J, Ogasa T, Ohta Y et al. Decline of human tactile angle discrimina- 38. Mueller C, Renner B. A new procedure for the short screening of olfactory
tion in patients with mild cognitive impairment and Alzheimer’s disease. J function using five items from the “Sniffin’ Sticks” identification test kit.
Alzheimer’s Dis 2010;22:225–234. Am J Rhinol 2006;20:113–116.
16. Campbell VA, Crews JE, Moriarty DG et al. Surveillance for sensory 39. Skrondal A, Rabe-Hesketh S. Generalized latent variable modeling: Mul-
impairment, activity limitation, and health-related quality of life among tilevel, longitudinal, and structural equation models. CRC Press, 2004 [on-
older adults in the United States, 1993–1997. MMWR CDC Surveill Summ line]. Available at
1999;48:131–156. d=YUpDqCzb-WMC&pgis=1 Accessed May 24, 2015.
17. Crews JE, Campbell VA. Vision impairment and hearing loss among com- 40. MacKinnon DP, Lockwood CM, Brown CH et al. The intermediate end-
munity-dwelling older Americans: Implications for health and functioning. point effect in logistic and probit regression. Clin Trials 2007;4:499–513.
Am J Public Health 2004;94:823–829. 41. Khil L, Wellmann J, Berger K. Impact of combined sensory impairments on
18. Kern DW, Wroblewski KE, Schumm LP et al. Field survey measures of health-related quality of life. Qual Life Res 2015;24:2099–2103.
olfaction: The Olfactory Function Field Exam (OFFE). Field Methods 42. Khil L, Wellmann J, Berger K. Determinants of single and multiple sensory
2014;26:421–434. impairments in an urban population. Otolaryngol Head Neck Surg
19. Murphy C, Schubert CR, Cruickshanks KJ et al. Prevalence of olfactory 2015;153:364–371.
impairment in older adults. JAMA 2002;288:2307–2312. 43. Michikawa T, Nishiwaki Y, Takebayashi T. Are you conscious of any age-
20. Welge-L€ ussen A. Ageing, neurodegeneration, and olfactory and gustatory related taste impairment? Prevalence of and factors associated with taste
loss. B-ENT 2009;5(Suppl 13):129–132. impairment in Japan. J Am Geriatr Soc 2010;58:981–983.
21. Wickremaratchi MM, Llewelyn JG. Effects of ageing on touch. Postgrad 44. Steinbach S, Hundt W, Vaitl A et al. Taste in mild cognitive impairment
Med J 2006;82:301–304. and Alzheimer’s disease. J Neurol 2010;257:238–246.
22. Schneider J, Gopinath B, McMahon C et al. Prevalence and 5-year inci- 45. Chinta SJ, Lieu CA, Demaria M et al. Environmental stress, ageing and
dence of dual sensory impairment in an older Australian population. Ann glial cell senescence: A novel mechanistic link to Parkinson’s disease? J
Epidemiol 2012;22:295–301. Intern Med 2013;273:429–436.
23. Lin MY, Gutierrez PR, Stone KL et al. Vision impairment and combined 46. Chaum E, Winborn CS, Bhattacharya S. Genomic regulation of senescence
vision and hearing impairment predict cognitive and functional decline in and innate immunity signaling in the retinal pigment epithelium. Mamm
older women. J Am Geriatr Soc 2004;52:1996–2002. Genome 2015;26:210–221.
24. Gopinath B, Schneider J, McMahon CM et al. Dual sensory impairment in
older adults increases the risk of mortality: A population-based study. PLoS
ONE 2013;8:e55054.
25. Loprinzi P, Smit E, Lin F. Accelerometer-assessed physical activity and SUPPORTING INFORMATION
objectively determined dual sensory impairment in US adults. Mayo Clin
2013;88:690–696. Additional Supporting Information may be found in the
26. Bertolini G, Wicki A, Baumann CR et al. Impaired tilt perception in online version of this article:
Parkinson’s disease: A central vestibular integration failure. PLoS ONE
27. Kerezoudi E, Thomas PK. Influence of age on regeneration in the periph- Figure S1. Prevalence of sensory impairments for each
eral nervous system. Gerontology 1999;45:301–306. of the five senses in community-dwelling U.S. older adults
28. Schubert CR, Cruickshanks KJ, Fischer ME et al. Carotid intima media according to race and ethnicity.
thickness, atherosclerosis, and 5-year decline in odor identification: The
Figure S2. Effects of a change in global sensory
Beaver Dam Offspring Study. J Gerontol A Biol Sci Med Sci
2015;70A:879–884. impairment (1 standard deviation) on each of the five
29. Smith S, Jaszczak A. Instrument development, study design implementa- sensory dysfunction measures (Table 3, Single-Factor
tion, and survey conduct for the National Social Life. Health, and Aging Model).
Project J Gerontol B Psychol Sci Soc Sci 2009;64B(Suppl 1):i20–i29.
Please note: Wiley-Blackwell is not responsible for the
30. Schumm LP, McClintock M, Williams S et al. Assessment of sensory func-
tion in the National Social Life, Health, and Aging Project. J Gerontol B content, accuracy, errors, or functionality of any support-
Psychol Sci Soc Sci 2009;64(Suppl 1):i76–i85. ing materials supplied by the authors. Any queries (other
31. Lindau ST, Schumm LP, Laumann EO et al. A study of sexuality and than missing material) should be directed to the corre-
health among older adults in the United States. N Engl J Med
sponding author for the article.