Fat Embolism Syndrome

PAUL D. STEIN, MD; ABDO Y. YAEKOUB, MD; FADI MATTA, MD;

MICHAEL KLEEREKOPER, MD

ABSTRACT: Objectives: To assess the incidence and risk

factors for fat embolism syndrome. Materials and Methods: Data from the National Hospital Discharge Survey
(NHDS) were analyzed using International Classification
of Diseases, Ninth Revision, Clinical Modification (ICD9-CM) codes. Results: From 1979 through 2005 among
928,324,000 patients discharged from short-stay hospitals in the United States, 41,000 (0.004%) had fat embolism syndrome. Among 21,538,000 patients with an
isolated fracture of the femur (any site), tibia, fibula,
pelvis, ribs, humerus, radius, or ulna, 25,000 (0.12%)
developed fat embolism syndrome. Patients with multiple fractures of the femur (excluding neck) more often
had fat embolism syndrome than those with isolated
fractures (1.29% versus 0.54%). The incidence of fat
embolism syndrome was lower with isolated fractures of

the tibia or fibula (0.30%) and even lower with isolated

fractures of the neck of the femur (0.06%). The incidence of fat embolism was too low to calculate with
isolated fractures of the pelvis, ribs, humerus, radius, or
ulna. Nonorthopedic conditions rarely, if ever, were
accompanied by fat embolism syndrome. The fat embolism syndrome was more frequent in men (relative risk
5.71). Children, aged 0 to 9 years rarely had fat embolism syndrome. The fat embolism syndrome most commonly affected patients aged 10 to 39 years. Conclusions: The incidence of the fat embolism syndrome
depends on the bone involved, whether fractures are
isolated or multiple, the age of the patient and the
gender. It rarely occurs as a result of medical conditions.
KEY INDEXING TERMS: Fat embolism syndrome; Fractures. [Am J Med Sci 2008;336(6):472477.]

alyzed data from the National Hospital Discharge

Survey (NHDS).

at embolism syndrome is a serious manifestation of fat embolism that involves progressive

respiratory insufficiency, deteriorating mental status, and petechial rash1 usually occurring within 72
hours of injury.2 Mortality from fat embolism syndrome in case series of 7 to 100 patients ranged from
6% to 29%,311 and a series of 208 patients with fat
embolism syndrome collected from 1963 to 1983
showed a mortality of 54%.12 Early fixation and
modern critical care should help minimize the impact of fat embolism.13 The reported incidence of fat
embolism syndrome in patients with fractures
ranges widely, from 0%6,14 to 35%.6 In the larger
reported case series, however, that evaluated somewhat over 3000 patients to nearly 17,000 patients
with fractures or patients hospitalized with injuries,
the incidence of fat embolism syndrome ranged from
0.3% to 1.3%.4,5,12,15,16 In view of the importance of
fat embolism syndrome, and sparse data on the
incidence and categories of patients at risk, we anFrom the Department of Research, St. Joseph Mercy Oakland
Hospital (PDS, AYY, FM), Pontiac, Michigan; Wayne State University School of Medicine (PDS), Detroit, Michigan; and the Department of Medicine, St. Joseph Mercy Hospital (MK), Ann Arbor,
Michigan.
Submitted February 7, 2008; accepted in revised form March 7,
2008.
Correspondence: Paul D. Stein, MD, Saint Joseph Mercy Oakland, 44405 Woodward Avenue, Pontiac, MI 48341-5023 (E-mail:
steinp@trinity-health.org).

472

Materials and Methods

Data Sources
The number of patients discharged from short-stay nonfederal
hospitals throughout the United States with a diagnostic code for
fat embolism from 1979 through 2005 was obtained from the
NHDS.17 Among these patients, the number with fractures at
various sites and illnesses reported to be associated with fat
embolism syndrome were determined. The NHDS consists of data
obtained annually from approximately 270,000 sampled inpatient records from about 500 non-Federal short-stay hospitals
(average length of stay 30 days) in 50 states and the District of
Columbia.17
Identification of Fat Embolism Syndrome
The International Classification of Diseases, Ninth Revision,
Clinical Modification (ICD-9-CM) code used for fat embolism was
958.1.
Identification of Fractures
The ICD-9-CM codes used for fractures of the pelvis were 808,
humerus 812, radius or ulna 813, femoral neck 820, femur other
and unspecified sites 821, and tibia and fibula 823.
Identification of Procedures
Procedures that we identified with ICD-9-CM codes were bone
graft 78.0, limb shortening 78.2, limb lengthening 78.3, internal
fixation 78.5, joint replacement of lower extremity 81.5, arthroplasty and repair of shoulder and elbow 81.8.
Identification of Illnesses
Illnesses identified by ICD-9-CM codes were decompression
sickness 993.3, burns, third degree 941.3, bone marrow transDecember 2008 Volume 336 Number 6

Stein et al

Table 1. Incidence of Fat Embolism Syndrome According to

Year
Years
(5-year Interval)

Incidence
(Fat Embolism Syndrome/5 yrs)

19811985
19861990
19911995
19962000
20002005

8000
10,000
7000
5000
9000

plant 996.85, V42.4, V42.8, bone marrow transplant donor V59.2,

V59.3, crush 929.9, pancreatitis 577.0, diabetes mellitus 250.0,
panniculitis 729.30, fatty liver 571.0, 571.8, and sickle cell anemia 282.6.
Statistical Analysis and Methodological Considerations
Relative risk and 95% confidence intervals (CI) were calculated
using calculator for confidence intervals of relative risk (www.
sign.ac.uk/methodology/risk.xls).

Results
From 1979 through 2005 among 928,324,000 patients discharged from short-stay hospitals, 41,000
(0.004%) had fat embolism syndrome. The incidence
of fat embolism remained relatively unchanged over
the interval of study (Table 1). Among 23,829,000
patients with isolated or multiple fractures involving the femur (any site), tibia, fibula, pelvis, ribs,
humerus, radius, or ulna, 41,000 (0.17%) developed
fat embolism syndrome. The fracture site most frequently responsible for fat embolism was the femur
(Figure 1). The incidence of fat embolism was higher
in patients with multiple fractures of these sites
[16,000 of 2,291,000 (0.70%)] than in patients with
isolated fractures of these sites [25,000 of 21,538,000
(0.12%) (P 0.0001)].
Among 388,000 patients with multiple fractures
that included the femur (excluding neck), 1.29% had
fat embolism syndrome (Table 2). Among 1,643,000

patients with isolated fractures of the femur (excluding neck), fewer (0.54%) developed fat embolism
syndrome (P 0.0001) (Table 2). The incidence of
fat embolism syndrome was lower in patients with
isolated fractures of the tibia or fibula (0.30%) and
even lower in patients with isolated fractures of the
neck of the femur (0.06%) (Table 2). The incidence of
fat embolism was too low to calculate with isolated
fractures of the pelvis, ribs, humerus, radius, or
ulna. Even taken together, patients with any of
these isolated fractures showed an incidence of fat
embolism too low to calculate accurately (Table 2).
The incidence of fat embolism syndrome in patients who had internal fixation, with or without
open fracture reduction was 0.15% of 11,862,000
(Table 2). The incidence of fat embolism syndrome in
patients with joint replacement of the lower extremity, arthroplasty and repair of shoulder or elbow,
bone graft, limb shortening, limb lengthening, osteotomy, and spinal fusion were too low to calculate
accurately.
Nonorthopedic conditions including decompression sickness, third degree burns, bone marrow
transplantation, crushing injury (excluding fracture), pancreatitis, diabetes mellitus, panniculitis,
fatty liver, and sickle cell anemia were either not
accompanied by fat embolism syndrome or only
rarely, and the incidences were too low to calculate
accurately.
The fat embolism syndrome was more frequent in
men (relative risk 5.71) (Table 3). A higher proportion of men had fractures of the femur (excluding
neck), tibia or fibula than women (25% versus 14%),
and fewer men had isolated fractures of the neck of
the femur (24%) than women (45%). Fractures of the
pelvis, ribs, humerus radius, or ulna occurred in
51% of men and 41% of women.
Among 1,178,000 patients aged 0 to 9 years who
had isolated fractures of the femur (any site), tibia,

Figure 1. Percentage of hospitalized patients with

the fat embolism syndrome (syn) who had isolated
or multiple fractures of the femur (other than
neck), tibia or fibula, neck of the femur, or pelvis,
humerus, ribs, or ulna.

Fat Embolism Syn (%)

40

30

34
25

20

24
17

10

0
Femur Other than
Neck

Tibia or Fibula

Femur Neck

Pelvis/Humerus/
Ribs/Radius or
Ulna

Isolated or Multiple Fractures

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

473

Fat Embolism Syndrome

Table 2. Incidence of Fat Embolism Syndrome According to Location of Fracture or Procedure

Multiple or Isolated Fractures
Fat Emb/No. Fx (%)

Multiple Fractures
Fat Emb/No. Fx (%)

Isolated Fractures
Fat Emb/No. FX (%)

Relative Risk
(95% CI)

14,000/2,031,000 (0.68)
7,000/8,387,000 (0.09)
10,000/2,747,000 (0.40)
10,000/10,664,000 (0.09)

5,000/388,000 (1.29)
b
/7,551,000
b
/411,000
b
/942,000

9,000/1,643,000 (0.54)
5,000/7,836,000 (0.06)
7,000/2,336,000 (0.30)
b
/9,722,000

2.35 (2.272.43)a

Femur (excluding neck)

Femur, neck
Tibia or fibula
Pelvis, ribs, humerus,
radius, ulna
Internal fixation (with or
without open fracture reduction)

18,000/11,862,000 (0.15)

Multiple fractures involving the femur (excluding neck), compared with isolated fractures of the femur (excluding neck).
Too few fat emboli to accurately calculate incidence.
Emb indicates embolism; Fx, fracture.
b

fibula, pelvis, ribs, humerus, radius, or ulna, the

incidence of fat embolism syndrome was too low to
calculate accurately (Table 3). The fat embolism
syndrome in patients aged 10 to 19 years and 20 to
39 years was more frequent (0.37% in both age
groups) than in patients aged 40 years and older
(incidence 0.05%). Comparing patients aged 20 to 39
years with patients 40 years the relative risk was
7.34 (Table 3).
Discussion
The incidence of fat embolism syndrome that we
observed among 3,979,000 patients with isolated
fractures of the femur (excluding neck), tibia or
fibula, 0.40%, was in the range observed by others in
case series of at least 3000 patients, 0.3% to
1.2%4,5,12,15,16 (Table 4). The reported incidences of
fat embolism syndrome among patients with single
or multiple fractures or various injuries ranges from
0 to 17%.3 6,10,12,14 16,18 22 (Table 4). The relative
risk of fat embolism syndrome that we observed in
patients with multiple fractures involving the femur
(excluding neck), compared with isolated fractures
of the femur (excluding neck), was 2.35. Similarly,
Table 3. Incidence of Fat Embolism Syndrome According to
Gender and Age Among Patients with Isolated Fractures of
Femur (any site), Tibia, Fibula, Pelvis, Ribs, Humerus, Radius,
or Ulna

Men
Women
Age (years)
09
1019
2039
40
a

Fat Emb/No. Fx (%)

Relative Risk (95% CI)

20,000/8,872,000 (0.23)
5,000/12,665,000 (0.04)

5.71 (5.545.89)a

b
/1,178,000
6,000/1,621,000 (0.37)
11,000/2,947,000 (0.37)
8,000/15,729,000 (0.05)

7.34 (7.137.55)c

Comparing men with women.

Too few fat emboli to accurately calculate incidence.
c
Comparing patients aged 20 to 39 years with patients 40 years.
Emb indicates embolism; Fx, fracture.
b

474

published data show a relative risk of fat embolism syndrome in patients with multiple fractures
compared with isolated fractures of 2.29 (Table
4).2,6,10,18,2325 The higher risk of patients with multiple fractures has been described.5,7,8,26,27
The incidence of fat embolism syndrome that we
observed was much higher in patients with long
bone fractures of the lower extremity than that of
the arm, as has been observed previously.4 The incidence of fat embolism syndrome that we observed
was also 7.6 times higher in patients with isolated
fractures of the femur (excluding neck) than in patients with isolated fractures of the neck of the
femur. The neck of the femur is characterized by red
bone marrow in adults, rather than yellow bone
marrow.28
We observed a higher incidence of fat embolism
syndrome in men than in women (relative risk of
5.7). Others, among 92 patients with fractures, also
showed a higher relative risk for fat embolism syndrome in men (relative risk 3.6).3 In our patients, a
higher proportion of men than women had fractures
of the femur (excluding neck), tibia, or fibula. Fewer
men had fractures of the neck of the femur. The
higher incidence in men has been suggested to relate
to men being more subjected to trauma.29 The degree of trauma and extent of associated soft-tissue
damage is likely to be less in women with fractures
than in men because fracture of the neck of femur
may occur with minimal trauma because of osteoporosis.30 There is a rapid and diffuse bone loss in
perimenopausal women31 in addition to the universal age-related bone loss that occurs in both men and
women.32
Children develop fat embolism syndrome almost
100 times less frequently than adults with comparable injuries.26 Among 1,178,000 children aged 0 to
9 years with isolated fractures of the femur (any
site), tibia, fibula, pelvis, ribs, humerus, radius, or
ulna, we found virtually no cases of fat embolism
syndrome. It has been suggested that the low incidence of fat embolism in children is due to the fact
that up to 14 years, the fat content of the bone
December 2008 Volume 336 Number 6

Stein et al

Table 4. Incidence of Fat Embolism

Fat Embolism Syndrome, n/N (%)

Incidence of fat embolism in patients
with isolated fractures
1/70 (1.4)
11/274 (4.0)
6/175 (3.5)
7/64 (10.9)
0/60 (0.0)
Incidence of fat embolism in patients
with multiple fractures or injuries
6/25 (24.0)
7/20 (35.0)
10/36 (27.8)
3/211 (1.4)
Incidence of fat embolism with
multiple or single fractures
38/4,197 (0.9)
47/3,650 (1.3)
208/16,706 (1.2)
7/95 (7.4)
5/43 (11.6)
0/50 (0.0)
7/80 (8.8)
41/812 (5.0)
27/3,026 (0.9)
17/6,564 (0.3)
17/100 (17.0)
10/92 (10.9)
11/82 (13.4)
9/62 (14.5)

Characteristics of Population at Risk

First Author, Year

(Reference)

Isolated femur (excluding neck)

Isolated femur (excluding neck)
Isolated femur (excluding neck)
Isolated fracture of femur (excluding neck), tibia,
or fibula
Isolated fracture of femur (excluding neck), or tibia

Chow, 198018
Pinney, 199823
Ten Duis, 198824
Myers, 197710

Multiple fractures of femur and other sites

Multiples fractures involving femur or tibia and
pelvis or other sites
Multiple fractures of femur (excluding neck), tibia,
or fibula and other sites
Long bone fractures and multiple injuries

Chow, 198018
Chan, 19846

Fractures
Fractures
Hospitalized with injuries
Fractures of femur (excluding neck)
Fractures of femur (excluding neck)
Fractures of extremities
Fractures of femur (excluding neck), or tibia, and
pelvis or multiple other sites
Lower extremity and pelvic fractures
Fracture of femur, tibia, fibula, or humerus
Fracture of femur (excluding neck), tibial shaft, or
pelvis
Fracture of femur (excluding neck), tibial, or fibula
Fracture of femur, tibia or pelvis
Fracture of long bones
Fracture of femur, tibia, or fibula

Magerl, 196615
Pelzl, 196916
Kroupa, 198812
Chow, 198018
Allardyce, 197419
McCarthy, 197314
Chan, 19846

Chan, 19846

Myers, 197710
Riska, 198225

Boricov, 200620
Bulger, 19974
Robert, 19935
Myers, 197710
Fabian, 19903
Kallenbach, 198721
Schoenfeld, 198322

n indicates number with fat embolism syndrome; N, number at risk.

marrow is low, with a greater proportion of higher

melting fats, palmitin and stearin, and less of the
fluid fat, olein.29 Palmitin and stearin (in the bone
marrow of children) are not as likely to produce
emboli as olein, found in the marrow of adults.2 An
orderly progression of red to yellow marrow conversion in the femur was appreciated first in the diaphysis (ages 110 years) and then in the distal metaphysis (ages 10 20 years), with an adult pattern seen
by age 24 years.33
Among patients aged 10 to 19 and aged 20 to 39
years, the relative risk compared with patients aged
40 years or older was 7.34. Others, among 321 patients with long bone injury, found a higher relative
risk for fat embolism syndrome among patients aged
11 to 40 years than age 41 years and older (relative
risk 13.15).34 In another investigation, no fat embolism syndrome was observed among 105 patients
aged 35 years or older, but 6.5% developed fat embolism syndrome among 169 patients aged 13 to 34
years.23 They speculated that patients under age 35
years (who are prodominantly male) in general sustain injuries as a consequence of high-energy trauTHE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

ma.23 Bone mineral content and bone mineral density decrease with age beginning at around age 40
years35 or perhaps 30 to 35 years.36 Minimal trauma
fractures may result in patients with decreased bone
density, which particularly occurs in women after
menopause.32,37 Because fractures of the neck of the
femur often are minimal trauma fractures and are
associated with a low risk of fat embolism syndrome,
it may be that minimal trauma fractures which
occur predominantly in sites rich in cancellous bone
(ie, richer in red marrow), result in a low risk of fat
embolism syndrome.
Several clinical settings for fat embolism syndrome have been named according to the broad
categories of trauma, surgery, and nontrauma.38
The medical conditions and nonorthopedic procedures that we evaluated were only rarely accompanied by fat embolism syndrome, or in some instances, or not at all. Occasional cases of fat
embolism syndrome occur, however, in many clinical
settings38 including diabetes mellitus,39 pancreatitis,40 burns,2 bone marrow transplant,41 sickle cell
disease,42 osteomyelitis,43 fatty liver,44 and soft tis475

Fat Embolism Syndrome

sue injury.45 Regarding orthopedic procedures, we

found an incidence of fat embolism syndrome in
patients who had internal fixation, with or without
open fracture reduction of 0.15%. The incidence of
fat embolism syndrome in patients that we identified with other orthopedic procedures was too low to
calculate accurately. Delayed stabilization is associated with a higher incidence of fat embolism syndrome, compared with prompt stabilization.5,23
Closed fractures are associated with a higher incidence of fat embolism than open fractures.4
The incidence of fat embolism syndrome is probably underestimated.18 It is often not possible to
make the diagnosis of fat embolism syndrome.4 Major clinical features are (1) respiratory symptoms
plus bilateral signs on chest examination with positive radiographic changes, (2) petechial rash, and
(3) cerebral signs unrelated to head injury or other
condition.1 Respiratory findings were later described as respiratory insufficiency46 or hypoxia
(PaO2 60 mm Hg).4 Minor features are fever,
tachycardia, retinal changes (fat or petechiae), renal
changes (anuria, oliguria, fat globules), sudden drop
in hemoglobin, sudden thrombocytopenia, high
erythrocyte sedimentation rate, and fat globules in
the sputum.1 Jaundice was subsequently added to
the minor features, and circulating fat globules were
identified as a laboratory feature.46 Clinical diagnosis is the key, because laboratory and roentgenographic findings are nonspecific.38
Strengths of this investigation are the huge patient sample, the diversity of the population in terms
of age, race, gender, and geographic region (all 50
states and the district of Columbia), the extensive
duration of observation (27 years), and the meticulous and statistically robust methods of sampling.17
Limitations include an inability to determine the
basis of the diagnosis of fat embolism syndrome and
whether some patients were diagnosed with fat embolism but did not have the findings of the clinical
syndrome. Also we do not know the severity of the
fractures, only the location of isolated fractures.

9.
10.
11.
12.
13.
14.
15.
16.
17.

18.
19.

20.
21.
22.

23.
24.
25.

References
1. Gurd AR. Fat embolism: an aid to diagnosis. J Bone Joint
Surg Br 1970;52:732737.
2. Levy D. The fat embolism syndrome: a review. Clin Orthop
Relat Res 1990:281286.
3. Fabian TC, Hoots AV, Stanford DS, et al. Fat embolism
syndrome: prospective evaluation in 92 fracture patients. Crit
Care Med 1990;18:42 46.
4. Bulger EM, Smith DG, Maier RV, et al. Fat embolism
syndrome. A 10-year review. Arch Surg 1997;132:435 439.
5. Robert JH, Hoffmeyer P, Broquet PE, et al. Fat embolism syndrome. Orthop Rev 1993;22:567571.
6. Chan KM, Tham KT, Chiu HS, et al. Post-traumatic fat
embolismits clinical and subclinical presentations. J Trauma
1984;24:45 49.
7. Peltier LF, Collins JA, Evarts CM, et al. A panel by
correspondence. Fat embolism. Arch Surg 1974;109:1216.
8. Rokkanen P, Lahdensuu M, Kataja J, et al. The syn-

476

26.
27.
28.
29.
30.
31.

32.

drome of fat embolism: analysis of thirty consecutive cases

compared to trauma patients with similar injuries. J Trauma
1970;10:299 306.
Gergen M. Clinical aspects and results of treatment of
post-traumatic fat embolism. Monatsschr Unfallheilkd Versicherungsmed 1961;64:365383.
Myers R, Taljaard JJ. Blood alcohol and fat embolism
syndrome. J Bone Joint Surg Am 1977;59:878 880.
Moreau JP. Fat embolism: a review and report of 100 cases.
Can J Surg 1974;17:196 199.
Kroupa J, Unger K. Early complications in patients with
multiple injuries and polytraumatism with special regard to
traumatic fat embolism. Czech Med 1988;11:217240.
Talbot M, Schemitsch EH. Fat embolism syndrome: history, definition, epidemiology. Injury 2006;37S:S3S7.
McCarthy B, Mammen E, Leblanc LP, et al. Subclinical
fat embolism: a prospective study of 50 patients with extremity fractures. J Trauma 1973;13:9 16.
Magerl F, Tscherne H. On diagnosis, therapy and prophylaxis of fat embolism. Langenbecks Arch Chir 1966;314:292
306.
Pelzl H. Clinical aspects and prevention of fat embolism.
Med Klin 1969;64:11071110.
US Department of Health and Human Services, Public
Health Service, National Center for Health Statistics National Hospital Discharge Survey 1979 1999 Multi-year
Public-Use Data File Documentation. Available at: http://www.
cdc.gov/nchs/about/major/hdasd/nhdsdes.htm. Last accessed
February 28, 2008.
Chow SP, Hoaglund FT, Ma A, et al. Fat embolism in
Hong Kong Chinese. J Bone Joint Surg Am 1980;62:1138
1143.
Allardyce DB, Meek RN, Woodruff B, et al. Increasing
our knowledge of the pathogenesis of fat embolism: a prospective study of 43 patients with fractured femoral shafts.
J Trauma 1974;14:955962.
Borisov MB, Gavrilin SV. The fat embolism syndrome in
severe combined traumas. Vestn Khir Im I I Grek 2006;165:
68 71.
Kallenbach J, Lewis M, Zaltzman M, et al. Low-dose
corticosteroid prophylaxis against fat embolism. J Trauma
1987;27:11731176.
Schonfeld SA, Ploysongsang Y, DiLisio R, et al. Fat
embolism prophylaxis with corticosteroids. A prospective
study in high-risk patients. Ann Intern Med 1983;99:438
443.
Pinney SJ, Keating JF, Meek RN. Fat embolism syndrome
in isolated femoral fractures: does timing of nailing influence
incidence? Injury 1998;29:131133.
Ten Duis HJ, Nijsten MW, Klasen HJ, et al. Fat embolism
in patients with an isolated fracture of the femoral shaft.
J Trauma 1988;28:383390.
Riska EB, Myllynen P. Fat embolism in patients with
multiple injuries. J Trauma 1982;22:891 894.
Gossling HR, Pellegrini VD Jr. Fat embolism syndrome: a
review of the pathophysiology and physiological basis of
treatment. Clin Orthop Relat Res 1982:68 82.
Peltier LF. Fat embolism. A current concept. Clin Orthop
Relat Res 1969;66:241253.
Feller JF. MRI of bone marrow: advanced MRI 2002: from
head to toe. Available at: http://mri.cpson.com/pdf/MRI_
of_the_Bone_Marrow.pdf. Last accessed November 1, 2007.
James ES. Fat embolism. Can Med Assoc J 1950;62:548
550.
Riggs BL, Baylink DJ, Kleerekoper M, et al. Incidence of
hip fractures in osteoporotic women treated with sodium
fluoride. J Bone Miner Res 1987;2:123126.
Chapurlat RD, Gamero P, Sornay-Rendu E, et al. Longitudinal study of bone loss in pre- and perimenopausal
women: evidence for bone loss in perimenopausal women.
Osteoporos Int 2000;11:493 498.
Jones G, Nguyen T, Sambrook P, et al. Progressive loss of

December 2008 Volume 336 Number 6

Stein et al

33.

34.

35.

36.

37.

bone in the femoral neck in elderly people: longitudinal findings from the Dubbo osteoporosis epidemiology study. BMJ
1994;309:691 695.
Moore SG, Dawson KL. Red and yellow marrow in the
femur: age-related changes in appearance at MR imaging.
Radiology 1990;175:219 223.
Moore P, James O, Saltos N. Fat embolism syndrome:
incidence, significance and early features. Aust N Z J Surg
1981;51:546 551.
Qin MW, Yu W, Xu L, et al. Bone mineral and body composition analysis of whole body in 292 normal subjects assessed
by dual X-ray absorptiometry. Zhongguo Yi Xue Ke Xue Yuan
Xue Bao 2003;25:66 69.
Harris ST, Jaffe RB, Shoback D. Menopause and Bone
Loss. J Clin Endocrinol Metab 2006;91. Available at: www.
jcem.endojournals.org/cgi/eletter-submit/91/3/0. Accessed November 3, 2008.
Kroger H, Huopio J, Honkanen R, et al. Prediction of
fracture risk using axial bone mineral density in a perimenopausal population: a prospective study. J Bone Miner Res
1995;10:302306.

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

38. Johnson MJ, Lucas GL. Fat embolism syndrome. Orthopedics 1996;19:41 48.
39. Purdiaev IUS, Asmolova ND. Fat embolism in diabetes
mellitus. Sud Med Ekspert 1983;26:49 50.
40. Guardia SN, Bilbao JM, Murray D, et al. Fat embolism
in acute pancreatitis. Arch Pathol Lab Med 1989;113:503
506.
41. Lipton JH, Russell JA, Burgess KR, et al. Fat embolization and pulmonary infiltrates after bone marrow transplantation. Med Pediatr Oncol 1987;15:24 27.
42. Shapiro MP, Hayes JA. Fat embolism in sickle cell disease.
Report of a case with brief review of the literature. Arch
Intern Med 1984;144:181182.
43. Broder G, Ruzumna L. Systemic fat embolism following
acute primary osteomyelitis. JAMA 1967;199:150 152.
44. Schulz F, Puschel K. An unusual form of a pulmonary fat
embolism in fulminant viral hepatitis. Pathologe 1996;17:
154 156.
45. Hiss J, Kahana T, Kugel C. Beaten to death: why do they
die? J Trauma 1996;40:2730.
46. Gurd AR, Wilson RI. The fat embolism syndrome. J Bone
Joint Surg Br 1974;56B:408 416.

477