Food-related choking deaths among the elderly

Abstract

During 2007-2010 in the USA, 2214 deaths among people aged ≥65 were attributed to choking on food. The death rate for this cause is higher among the elderly than among any other age group. Using data from the US National Vital Statistics System, we examined the relationship between food suffocation and other causes of death listed on the death certificate. Among decedents aged ≥65, the three most common additional conditions listed on the death certificate were heart disease, dementia and diabetes. However, after estimating the expected joint frequency of other causes based on the overall distribution of all causes of death, we find that three causes-dementia (including Alzheimer's disease), Parkinson's disease and pneumonitis-are most strongly associated with deaths from choking on food among older people.

Full text

Food-related choking deaths among the elderly
Ellen Kramarow,
1
Margaret Warner,
2
Li-Hui Chen
1
▸Additional material is
published online only. To view
please visit the journal online
(http://dx.doi.org/10.1136/
injuryprev-2013-040795).
1
Office of Analysis and
Epidemiology, National Center
for Health Statistics, Centers
for Disease Control and
Prevention, Hyattsville,
Maryland, USA
2
Division of Vital Statistics,
National Center for Health
Statistics, Centers for Disease
Control and Prevention
Hyattsville, Hyattsville,
Maryland, USA
Correspondence to
Dr Ellen Kramarow, Office of
Analysis and Epidemiology,
National Center for Health
Statistics, Centers for Disease
Control and Prevention, 3311
Toledo Road, Hyattsville, MD
20782, USA; ekramarow@cdc.
gov
Received 26 February 2013
Revised 15 July 2013
Accepted 23 July 2013
Published Online First
3 September 2013
To cite: Kramarow E,
Warner M, Chen L-H. Inj
Prev 2014;20:200–203.
ABSTRACT
During 2007–2010 in the USA, 2214 deaths among
people aged ≥65 were attributed to choking on food.
The death rate for this cause is higher among the elderly
than among any other age group. Using data from the
US National Vital Statistics System, we examined the
relationship between food suffocation and other causes
of death listed on the death certificate. Among
decedents aged ≥65, the three most common additional
conditions listed on the death certificate were heart
disease, dementia and diabetes. However, after
estimating the expected joint frequency of other causes
based on the overall distribution of all causes of death,
we find that three causes—dementia (including
Alzheimer’s disease), Parkinson’s disease and
pneumonitis—are most strongly associated with deaths
from choking on food among older people.
INTRODUCTION
Unintentional injuries are the ninth leading cause
of death among people aged ≥65 in the USA.
1
Within the injury category, suffocation ranks third
and is often overlooked given the attention paid to
the two leading causes of injury deaths in the
elderly—falls and motor vehicle traffic deaths.
From 2007 to 2010 in the USA, 2214 deaths
among people aged ≥65 were attributed to
food-related suffocation, an average of >500
deaths a year. The death rate for food-related suffo-
cation is nearly seven times higher among people
aged ≥65 than among children aged 1–4.
Food-related suffocation (choking on food)
accounts for about 17% of all elderly suffocation
deaths. Older people often have chronic conditions
that involve difficulties with chewing and swallow-
ing which may be related to the risks of choking on
food. This study uses information on all causes of
death recorded on death certificates collected by
the US National Vital Statistics System (NVSS) to
examine the association between food suffocation
and other causes of death. While suffocating on
food results in only a small proportion of all deaths
among older people, these deaths may be prevent-
able.
2–4
DATA AND METHODS
The NVSS mortality data are based on information
from all resident death certificates filed in the 50
US states and the District of Columbia. More than
99% of deaths occurring in the USA are believed to
be registered. The data are not subject to sampling
error but may be affected by random variation.
15
Death certificates are the basis for most national
statistics on death in the USA. On the death certifi-
cate, the certifier lists the chain of events leading to
death along with other conditions that contributed
to the death. For deaths due to injuries, the certifier
describes how the injury occurred. The certifier must
be a physician treating the decedent or a medical-
legal officer who is either a coroner or medical exam-
iner, depending on the state and county in which the
decedent died. Differences are found across states as
to who can certify deaths and under what circum-
stances a medical-legal officer must be involved in the
certification. However, in general a medical-legal
officer will certify deaths when they occur suddenly
and without warning, when the decedent was not
being treated by a physician, or when the death was
unattended.
6
The conditions, diseases and injuries
recorded on the death certificate are processed in
accordance with the International Classification of
Diseases, Tenth Revision (ICD-10) and constitute the
multiple-cause-of-death data. These data can be used
to describe associations among conditions listed on
the death certificate.
78
For the purpose of national mortality statistics,
one condition is selected as the underlying cause
based on how the information is reported on the
death certificate and using international selection
rules. According to the US cause-of-death classifica-
tion instructions, the underlying cause is defined as
“the disease or injury which initiated the train of
morbid events leading directly or indirectly to
death or the circumstances of the accident or vio-
lence which produced the fatal injury.”The instruc-
tions make clear that “a death often results from
the combined effect of two or more conditions”,
which may be independent of each other or may be
‘causally related.’The causal sequence of condi-
tions listed on the death certificate is designed to
represent the progression leading to death. The
underlying cause selection rules rely on the order
in which the certifier lists the causes unless the
sequence is considered illogical.
9
In most cases of food suffocation deaths, it
should be considered the underlying cause of
death. For the elderly, particularly among those
with many chronic conditions, the causal sequence
may be less clear. The certifier may feel elderly
people were too frail to eat and that their disease
played a major role in the death. Many combina-
tions of conditions and injuries are considered
improbable and edited during the selection of an
underlying cause, but the cause-of-death selection
rules do permit the underlying cause of death to be
a disease that affects the ability to swallow when
the co-occurring cause is food suffocation.
9
In this paper, we first examine deaths among
people aged ≥65 years in 2007–2010 that have an
underlying cause of food suffocation (ICD-10 code
of W79, ‘inhalation and ingestion of food causing
obstruction of respiratory tract’). We combine
4 years of data to obtain numbers large enough for
reliable calculations and calculate death rates per
100 000 population by age and sex.
200 Kramarow E, et al.Inj Prev 2014;20:200–203. doi:10.1136/injuryprev-2013-040795
Brief report
group.bmj.com on June 10, 2015 - Published by http://injuryprevention.bmj.com/Downloaded from

The distribution of deaths by place of injury is presented. We
then examine the multiple causes of death, calculating the per-
centage of deaths with food choking as an underlying cause that
also had a comorbid condition reported on the death certificate.
We use a strength-of-association measure to describe the rela-
tionship between food choking and other causes of death.
When calculating the strength of association measure, we allow
for the food choking death to be written anywhere on the death
certificate, not only as the underlying cause. The measure is the
ratio of observed deaths with both causes to the expected joint
frequency of deaths with the same causes, assuming the causes
were independent.
10–12
The expected number of deaths is calculated as:
(Number of deaths with suffocation)
(Number of deaths with selected cause)
Total number of deaths from all causes
The ratio is calculated as:
Observed numbers of deaths with both causes
Expected numbers of deaths with both causes
A ratio >1 indicates a positive association between the two
causes—that is, the two causes occurred together more fre-
quently than expected based on the distribution of the causes in
the population. A ratio of ≤1 denotes no positive relationship.
CIs are estimated based on the SE of a standardised mortality
ratio (observed/expected deaths).
RESULTS
Figure 1 shows the death rates for food suffocation by dece-
dent’s age and sex for 2007–2010. Death rates are significantly
higher among male than female subjects at ages 1–4 and ages
≥25. Although the overall death rates are low, the rates are
higher among those aged <5 than among those aged 5–44.
Rates increase steadily after age 5, with the highest death rates
for food suffocation found among the oldest population (4.5
deaths per 100 000 for men aged ≥85). The pattern of increas-
ing death rates shown in figure 1 significantly fits a log-linear
model calculated using the Joinpoint regression programme
13
(see the online supplementary appendix for rates and SEs).
Table 1 shows the place of injury for food choking deaths
among this population. Most food choking events resulting in
death among older people occurred in the home (38%), but
one-quarter took place in a residential institution such as a
nursing home. About 9% occurred in trade and service areas,
which would include restaurants. Previous studies have noted a
variation in the percentage of injury deaths lacking information
on location of injury.
14
The percentage of food choking deaths
recorded without a specified place of injury (19%) is somewhat
lower than the overall percentage with unspecified location for
all non-transportation unintentional injuries among the same
age group (25%).
Figure 2 shows the percentage distribution of diseases listed
as associated causes for deaths with food suffocation as the
underlying cause of death. Decedents can have more than one
of the causes listed as an associated cause. Among people aged
≥65, heart disease is the most common condition listed among
the multiple causes when the underlying cause of death is food
suffocation: 32% of the food choking deaths among older
people also had heart disease listed as an associated cause.
(When cardiac arrest is excluded as a cause, heart disease is
mentioned in 23% of food suffocation deaths.) Other diseases
Figure 1 Death rates for accidental
suffocation caused by inhalation or
ingestion of food, by age and sex:
USA, 2007–2010. Source: CDC/NCHS/
National Vital Statistics System, 2007–
2010. Note: Rates shown are for food
suffocation as an underlying cause of
death.
Table 1 Place of injury for deaths from accidental suffocation
caused by inhalation or ingestion of food among decedents aged
≥65: USA, 2007–2010
Place of injury Number %
Home 831 37.5
Residential institution (eg, nursing home) 542 24.5
School, other institution (eg, hospital) 102 4.6
Sports areas 1 0.0
Street and highway 13 0.6
Trade and service area (eg, restaurant) 207 9.3
Industrial/construction areas 0 0.0
Farm 4 0.2
Other 89 4.0
Unspecified 425 19.2
Total 2214 100.0
Source: CDC/NCHS/National Vital Statistics System, 2007–2010.
Note: Numbers shown are for food suffocation as an underlying cause of death.
Kramarow E, et al.Inj Prev 2014;20:200–203. doi:10.1136/injuryprev-2013-040795 201
Brief report
group.bmj.com on June 10, 2015 - Published by http://injuryprevention.bmj.com/Downloaded from

were mentioned much less frequently: Alzheimer’s disease and
other dementias (11%); diabetes (7%) ; stroke (<2%); and
Parkinson’s disease (<0.5%). Pneumonitis was listed in nearly
3% of the food suffocation deaths.
Figure 3 presents the strength of association results and shows
no association between food choking and heart disease. The
large percentage of food choking deaths with heart disease as an
associated cause (seen in figure 2), therefore, is not more than
would be expected given the prevalence of heart disease as a
cause of death among this age group. We also found no positive
association of food choking with cardiac arrest alone (data not
shown). Figure 3 does show a positive relationship between
food choking and Parkinson’s disease, pneumonitis and demen-
tia (including Alzheimer’s disease).
DISCUSSION
Suffocation on food is a preventable cause of death
2–4
; yet,
among people aged ≥65 in the USA, about 500 deaths a year
are attributed to this event as an underlying cause of death with
at least another 300 deaths listing food suffocation as an asso-
ciated cause of death. Also, more than 2000 elderly deaths each
year are attributed to suffocation on ‘other objects.’Some of
these deaths may also involve food. The rules embedded in the
cause-of-death classification system instruct that phrases on the
death certificate of ‘aspiration’and ‘choked’without additional
information default to the ICD-10 code of W80, ‘inhalation
and ingestion of other objects.’In addition, an unknown
number of food choking deaths may be attributed to other
causes (eg, cardiac arrest).
15 16
There is a positive association between food suffocation and
pneumonitis; however, it is difficult to discern from the death
certificate whether the food choking incident was the precipitat-
ing cause of the pneumonitis or resulted from the disease.
Pneumonitis is a form of lung inflammation or pneumonia
caused by aspiration of food, liquids, or other objects and is
among the 15 leading causes of death for people aged ≥65. It is
often seen as a cause of death among very old, frail people,
especially those in long-term care.
17
Sorting out the causal
sequence of food choking and other conditions when there may
be a time lag between onset of disease and the food choking
incident (eg, in patients with stroke) can be a challenge during
death investigation.
18
Our study provides an overall picture of the scope of the
problem of food suffocation deaths among the elderly; yet, it
suffers from some limitations of the data. Differences across juris-
dictions and in the qualifications of certifiers may affect reporting
of causes of death.
19
We rely on the certifier and the ICD-10
coding rules and guidelines for the selection of the underlying
cause of death; consequently, misclassifications of food-related suf-
focation as an underlying cause of death is a possibility. Our
measure of association accounts for all listed causes, irrespective of
order, which helps to minimise any bias of recording an incorrect
causal sequence on the death certificate. In addition, omissions of
contributing causes of death on the death certificate might affect
our results. Although the rates are low, deaths among the elderly
with food suffocation as an underlying cause of death have higher
rates of autopsy (25%) than deaths among people of the same age
attributed to all unintentional injuries (17%) or deaths from all
Figure 2 Percentage distribution of
associated causes of death for
accidental suffocation caused by
inhalation or ingestion of food as an
underlying cause of death for
decedents aged ≥65: USA, 2007–
2010. Source: CDC/NCHS/National Vital
Statistics System, 2007–2010.
Figure 3 Strength of association
between accidental suffocation caused
by inhalation or ingestion of food and
selected causes of death among
decedents aged ≥65: USA, 2007–
2010. *95% CI. Note: See text for
definition of strength of association.
Values greater than one indicate a
positive association. Associations do
not take into consideration which
cause was selected as the underlying
cause of death. CIs are estimated
based on the SE of a standardised
mortality ratio (observed/expected
deaths). Source: CDC/NCHS/National
Vital Statistics System, 2007–2010.
202 Kramarow E, et al.Inj Prev 2014;20:200–203. doi:10.1136/injuryprev-2013-040795
Brief report
group.bmj.com on June 10, 2015 - Published by http://injuryprevention.bmj.com/Downloaded from

causes among people aged ≥65 (2%). Future research on this topic
should take into account the order of causes listed on the death cer-
tificate and attempt to link mortality data to other databases where
more information on the circumstances of death can be analysed.
Information about the risk of choking on food among infants
and toddlers is widespread.
3
Although the risks are acknowl-
edged for those in long-term care, the dangers of food suffoca-
tion among older people are not as well known in the general
population.
20
Eating semi-solid foods such as peanut butter,
having poor dentition and consuming alcohol and medication
such as sedatives and antipsychotic drugs are all risk factors for
food suffocation among older adults.
242122
Despite the aware-
ness of choking risks in medical and long-term care situations,
one-quarter of the food choking events among older decedents
occurred in a residential institution such as a nursing home. A
better understanding of the relationship between food suffoca-
tion and associated chronic diseases might assist in developing
prevention strategies.
What is already known on this subject
▸More food-related choking deaths occur annually among the
elderly than among young children and infants.
▸Some elderly people are at high risk for choking because of
complications of chronic diseases.
▸Interventions are available to prevent choking on food
among the elderly.
What this study adds
▸Rates for male subjects are higher than those for female
subjects in many age groups.
▸Contributing causes of death for food choking include
Parkinson’s disease, dementia (including Alzheimer’s
disease) and pneumonitis.
▸About one-quarter of all food choking events among the
elderly occurred in nursing homes or other residential
institutions.
Acknowledgements Judi Wong, former Association of Schools of Public Health
intern, provided assistance in an earlier version of this paper. Thanks to Holly
Hedegaard for comments and suggestions.
Contributors MW conceived the idea for the study; EK, MW and L-HC designed
the study and were responsible for the data analysis and interpretation of results. EK
prepared the initial draft of the manuscript and produced the tables and graphs. All
authors provided input to the revisions and final manuscript.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1 National Center for Health Statistics. Health, United States, 2012: with special
feature on emergency care. Hyattsville, MD, 2013.
2 Berzlanovich AM, Fazeny-Dorner B, Waldhoer T, et al. Foreign body asphyxia: a
preventable cause of death in the elderly. Am J Prev Med 2005;28:65–9.
3 Committee on Injury, Violence and Poison Prevention. Policy statement—prevention
of choking among children. Pediatrics 2010;125:601–7.
4 Manitoba Health. Preventing Suffocation and Choking Injuries in Manitoba. http://
www.gov.mb.ca/healthyliving/hlp/docs/injury/injuries_suffocation.pdf (accessed Jun
2013).
5 Murphy SL, Xu J, Kochanek KD. Deaths: final data for 2010. Natl Vital Stat Rep
2013;61:1–168.
6Medical examiners’and Coroners’handbook on death registration and fetal death
reporting. Hyattsville, MD: National Center for Health Statistics, 2003. http://www.
cdc.gov/nchs/data/misc/hb_me.pdf (accessed August 2013).
7 Anderson RN. Coding and classifying causes of death: trends and international
differences. In: Rogers RG, Crimmins EM. eds. New York: Springer, 2011:467–89.
8 World Health Organization. International statistical classification of diseases and
related health problems, tenth revision: volume II, instruction manual. Geneva:
World Health Organization, 1992.
9Vital statistics, instructions for classifying the underlying cause-of-death, ICD-10.
NCHS Instruction Manual; Part 2a. Hyattsville, MD: National Center for Health
Statistics, 2011. http://www.cdc.gov/nchs/data/dvs/2a2011.pdf (accessed August
30th)
10 Israel RA, Rosenberg HM, Curtin LM. Analytical potential for multiple
cause-of-death data. Am J Epidemiol 1986;124:161–79.
11 Australian Bureau of Statistics. Multiple Cause of Death Analysis, 1997–2001,
Appendix V. Method of calculating the observed to expected ratio as a measure of
the strength of association. Cat. no. 3319.0.55.001–2003. 2003. http://www.abs.
gov.au/AUSSTATS/abs@.nsf/Lookup/3319.0.55.001Main+Features11997-2001?
OpenDocument
12 Gorina Y, Lentzer H. Multiple causes of death in old age. Aging Trends No. 9.
Hyattsville, MD. National Center for Health Statistics, 2008.
13 Joinpoint Regression Program, Version 3.4.3. Statistical Research and Applications
Branch, National Cancer Institute. April 2010.
14 Minino AM, Anderson RN, Fingerhut LA, et al. Deaths: injuries, 2002. Natl Vital
Stat Rep 2006;54:1–125.
15 Vital Statistics. Instructions for classifying the multiple causes of death, ICD-10.
NCHS Instruction Manual; Part 2b. Hyattsville, MD: National Center for Health
Statistics, 2011.
16 Wong SC, Tariq SM. Cardiac arrest following foreign-body aspiration. Respir Care
2011;56:527–9.
17 Fogata ML, Naik PN, Goldberg TH. Causes of death in the very old [letter]. JAm
Geriatr Soc 1996;44:1412–13.
18 Chang CY, Cheng TJ, Lin CY, et al. Reporting of aspiration pneumonia or choking
as a cause of death in patients who died with stroke. Stroke 2013;44:1182–5.
19 Breiding MJ, Wiersema B. Variability of undetermined manner of death classification
in the US. Inj Prev 2006;12(Suppl II):ii49–54.
20 Tanner DC. Lessons from nursing home dysphagia malpractice litigation. J Gerontol
Nurs 2010;36:41–6.
21 Dolkas L, Stanley C, Smith AM, et al. Deaths Associated with Choking in San
Diego. J Forensic Sci 2007;52:176–9.
22 Nikolic S, Zivkovic V, Babic D, et al. Laryngeal choking on food and acute ethanol
intoxication in adults-an autopsy study. J Forensic Sci 2011;56:128–31.
Kramarow E, et al.Inj Prev 2014;20:200–203. doi:10.1136/injuryprev-2013-040795 203
Brief report
group.bmj.com on June 10, 2015 - Published by http://injuryprevention.bmj.com/Downloaded from

elderly
Food-related choking deaths among the
Ellen Kramarow, Margaret Warner and Li-Hui Chen
doi: 10.1136/injuryprev-2013-040795
2014 20: 200-203 originally published online September 3, 2013Inj Prev
http://injuryprevention.bmj.com/content/20/3/200
Updated information and services can be found at:
These include:
Material
Supplementary
2013-040795.DC1.html
http://injuryprevention.bmj.com/content/suppl/2013/09/03/injuryprev-
Supplementary material can be found at:
References #BIBLhttp://injuryprevention.bmj.com/content/20/3/200
This article cites 12 articles, 5 of which you can access for free at:
service
Email alerting box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Notes
http://group.bmj.com/group/rights-licensing/permissions
To request permissions go to:
http://journals.bmj.com/cgi/reprintform
To order reprints go to:
http://group.bmj.com/subscribe/
To subscribe to BMJ go to:
group.bmj.com on June 10, 2015 - Published by http://injuryprevention.bmj.com/Downloaded from