CLINICAL ORTHOPAEDICS AND RELATED RESEARCH
Number 421, pp. 260–267
Magnetic Resonance Imaging in Stress Fractures and
Shin Splints
Yoshimitsu Aoki, MD, PhD*; Kazunori Yasuda, MD, PhD†; Harukazu Tohyama, MD, PhD†;
Hirokazu Ito, MD*; and Akio Minami, MD, PhD*
The purpose of the current study was to determine whether
stress fractures and shin splints could be discriminated with
MRI in the early phase. Twenty-two athletes, who had pain
in the middle or distal part of their leg during or after sports
activity, were evaluated with radiographs and MRI scans.
Stress fractures were diagnosed when consecutive radio-
graphs showed local periosteal reaction or a fracture line,
and shin splints were diagnosed in all the other cases. In all
eight patients with stress fractures, an abnormally wide high
signal in the localized bone marrow was the most detectable
in the coronal fat-suppressed MRI scan. In 11 patients with
shin splints, the coronal fat-suppressed MRI scans showed a
linear abnormally high signal along the medial posterior sur-
face of the tibia, and in seven patients with shin splints, the
MRI scans showed a linear abnormally high signal along the
medial bone marrow. No MRI scans of shin splints showed
an abnormally wide high signal in the bone marrow as ob-
served on MRI scans of stress fractures. This study showed
that fat-suppressed MRI is useful for discrimination between
stress fracture and shin splints before radiographs show a
detectable periosteal reaction in the tibia.
Athletes frequently experience chronic pain induced by
exercise in the medial aspect of the lower leg. Because the
pain disturbs their sports activities, it is necessary for a
patient with chronic leg pain to be diagnosed correctly as
early as possible. Stress fractures and shin splints are well-
known as causes of chronic leg pain. Although shin splints
and tibial stress fracture are involved in overuse injuries,
Received: June 28, 2002
Revised: March 27, 2003, June 12, 2003
Accepted: June 24, 2003
From the *Department of Orthopaedic Surgery, Hokkaido University Gradu-
ate School of Medicine, Sapporo, Japan; and the †Department of Medical
Bioengineering and Sports Medicine, Hokkaido University Graduate School
of Medicine, Sapporo, Japan
Correspondence to: Yoshimitsu Aoki, MD, PhD, Department of Orthopaedic
Surgery, Hokkaido University Graduate School of Medicine, Kita-15 Nishi-
7, Kita-ku, Sapporo, 060-8638, Japan. Phone: 81-11-706-5935; Fax: 81-11-
706-6054; E-mail:yoshaok@med.hokudai.ac.jp.
DOI: 10.1097/01.blo.0000126333.13806.87
260
© 2004 Lippincott Williams & Wilkins
they are different diseases.14,15,19 According to the stan-
dard nomenclature of the American Medical Association
as reported by Slocum, 19 shin splints are defined as pain
and discomfort in the leg from repetitive running on hard
surfaces or excessive use of foot flexors. The diagnosis
should be limited to musculotendinous inflammation ex-
cluding fractures and ischemic disorders. However, stress
fractures are recognized as the continuum of changes from
microfracture to frank fracture under excess stress. However,
patients with stress fractures or shin splints have similar
symptoms and they have similar clinical signs, at least
during the early phase of each disease. Because treatment
of these diseases is different, early discrimination between
shin splints and stress fractures is important for athletes.
Athletes with shin splints do not need a long rest time, but
all athletes with stress fractures of the tibia should cease
sports activity for at least 4 or 6 weeks. Early discrimina-
tion between shin splints and stress fractures can allow
athletes a return to sports activity as early as possible.
The differential diagnosis between these two diseases is
difficult in the early phase as determined from the physical
examination and the radiographs. Radiographs are not use-
ful in the early phase of stress fractures, although they are
useful in the late phase because they show periosteal re-
action, callus formation, or sclerotic fracture line.2 The
time from onset of pain to positive radiographic evidence
of a stress fracture can vary from 2 weeks to 3 months.7,23
Although bone scintigraphy for a tibial stress fracture
commonly shows abnormal uptake at the site of the stress
fracture, shin splints also sometimes have an abnormal
uptake. Therefore, differentiation between shin splints and
stress fractures is not easy. Magnetic resonance imaging
has been used to diagnose many bone diseases. No studies,
however, have been done to discriminate between shin
splints and stress fracture in the early phase with MRI. The
purpose of the current study was to determine whether
MRI can discriminate between shin splints and stress frac-
tures in the early phase before radiographs show periosteal
reaction or callus formation.
Number 421
April 2004 Magnetic Resonance Imaging in Stress Fractures and Shin Splints
MATERIALS AND METHODS
Patients
Between April 1994 and June 1999, 22 athletes who had chronic
pain in the middle or distal portion of the medial aspect of the leg
during or after sports activities were enrolled in this study. Pa-
tients who had histories of acute trauma such as a direct blow to
the tibia and who had a disease predisposing them to a stress
fracture were excluded. They were screened for disease by ex-
amination of the blood. The inclusion criteria involved moderate
or severe tenderness along the medial posterior border of the
tibia and no abnormal findings seen on radiographs taken at the
initial examination. There were 13 males and nine females, rang-
ing in age from 13–33 years (average age, 16.3 years). The patients
included eight runners, seven basketball players, three soccer
players, two volleyball players, and two kendo players. Twenty
patients were involved in competition at the college, high school,
or junior high school level, and two patients were involved in
recreational activities. In the current study, our protocol for hu-
man subjects was approved according to the relevant laws and
regulations of our country.
Study Design
Each patient was followed up at the outpatient clinic at our
institution every 2 weeks for 8 weeks. For each patient, antero-
posterior (AP) and lateral radiographs were taken at the initial
examination, and at 2 weeks, 4 weeks, and 8 weeks after the
initial examination. Magnetic resonance imaging also was done
with 0.5-T systems (Yokogawa Medical Co, Tokyo, Japan) for
each patient at the initial examination. Magnetic resonance im-
aging was done once for 17 patients at the initial examination,
and twice for five patients at the initial examination, and at 6–8
weeks after the initial examination. Fast spin-echo short inver-
sion-recovery (STIR) (TR, 2500–3700; TE, 20–40; inversion
time, 90–100 ms) was obtained in the coronal plane and axial
plane in all patients. Spectral fat saturation was used with fast
spin-echo sequences. T1-weighted spin-echo (SE) imaging (rep-
etition time [TR], 400 ms; echo time [TE], 25 ms) and T2-
weighted SE imaging (TR, 3000; TE, 110) were obtained in the
coronal plane and sagittal plane in 12 of the 22 patients. Fourteen
of 22 athletes had a technetium-99m bone scan.
Radiographs and MRI scans were examined carefully by two
of us (YA, KY). On MRI scans, high signal intensity lesions
were specifically looked for in the tibia. The coronal MRI find-
ings were classified into three images, subperiosteal high signal
(Fig 1A), subperiosteal high signal and linear high signal in
medial bone marrow (Fig 1B), and wide high signal extending
throughout the whole bone marrow (Fig 1C). Axial MRI scans
also showed a subperiosteal high signal, partially high signal, or
wide high signal in the bone marrow.
Using the radiographs taken at 12 weeks, a definite diagnosis
was made for each patient. Tibial stress fractures were diagnosed
when consecutive radiographs showed local periosteal reaction,
callus formation, or a fracture line 4–12 weeks after onset of
symptoms. However, shin splints were diagnosed when consecu-
tive radiographs showed no periosteal reaction or callus forma-
261
tion throughout the 12 weeks. Based on the diagnosis, the 22
patients were divided into two groups, patients with stress frac-
tures and patients with shin splints. Eight patients had stress
fractures and 14 patients had shin splints. Magnetic resonance
imaging findings and duration of clinical symptoms were com-
pared between the groups. Statistical analysis was done with the
Student’s t test, Mann-Whitney’s U test, and the chi square test.
RESULTS
In the eight patients with stress fractures, plain radiographs
showed periosteal reaction, callus formation, or a fracture
line at 4, 8, or 12 weeks after the initial examination (Fig
1D). In the 14 patients with shin splints, no abnormal
radiographic findings were observed at 2, 4, 8, or 12 weeks
after the initial examination. The mean duration of symp-
toms until MRI was done in all patients ranged from 7
days to 6 weeks (average, 19 days). The mean duration of
symptoms until MRI was done was 17.2 days for patients
with stress fractures and 20.7 days for patients with shin
splints. There were no statistically significant differences
in duration of symptoms and activity level between pa-
tients with stress fractures and patients with shin splints.
In all eight patients with stress fractures, coronal T2-
weighted and fat-suppressed MRI scans showed an abnor-
mally wide high signal in the localized bone marrow (Fig
1C). In the fat-suppressed MRI scans in particular, abnor-
mal findings were more detectable than those in the T1-
and T2-weighted MRI scans (Fig 2A-C). The abnormally
high signal area in the STIR fat-suppressed MRI scans was
continuous with the cortex. The plain radiographs also
showed callus formation 8 weeks after the initial exami-
nation (Fig 2D,E). Also, in axial MRI scans, the abnor-
mally high signal extending throughout the entire bone
marrow was detectable in all patients with stress fractures
(Fig 3).
In the 12 patients with shin splints, the coronal fat-
suppressed MRI scan showed a linear abnormally high
signal along the medial posterior surface of the tibia (Fig
1A). In seven patients with shin splints, the coronal fat-
suppressed MRI scan showed a linear abnormally high
signal along the medial aspect of the bone marrow (Fig
1B). Those longitudinal linear high signals were located
just along the medial aspect of the tibial bone marrow and
never extended throughout the entire tibial bone marrow.
Abnormally high signals along the medial posterior sur-
face of the tibia and the medial aspect of the bone marrow
on coronal MRI also were detectable on axial MRI scans
(Fig 4A,B), but a radiograph did not show the periosteal
reaction (Fig 4C). The periosteal high signal and the linear
high signal along the medial bone marrow on the fat-
suppressed MRI scan were seen in five patients with shin
splints. No MRI scans of patients with shin splints showed
 Clinical Orthopaedics
262 Aoki et al and Related Research

Fig 1. (A) A coronal MRI scan of a 14-year-old tennis player, diagnosed with shin splints, shows an abnormally high signal along
the medial border of the tibia. (B) A coronal MRI scan of a 17-year-old female kendo player, diagnosed with shin splints, shows
a linear abnormally high signal intensity along the medial periosteal region and the medial aspect of the bone marrow. Radio-
graphs obtained 4 and 8 weeks later revealed no periosteal reaction. (C) A coronal MRI scan of a 14-year-old male soccer player,
diagnosed with a stress fracture, shows an abnormally wide high signal in the localized bone marrow. The abnormally high signal
area was continuous with the cortex. (D) A radiograph of a 14-year-old male soccer player with a stress fracture, obtained 4 weeks
after the first examination, revealed a periosteal reaction at the medial side of the tibia.
Number 421
April 2004 Magnetic Resonance Imaging in Stress Fractures and Shin Splints 263

Fig 2. Coronal MRI scans of a 17- year–old male runner, diagnosed with a stress fracture, show a strikingly wide low signal
intensity on the (A) T1-weighted scan, and a high signal intensity on the (B) T2-weighted scan, and (C) STIR fat-suppressed scan
in the localized bone marrow. Particularly in the STIR fat-suppressed MRI scan, the abnormal finding was more detectable.
Radiographs obtained 8 weeks after the first examination revealed callus formation at the lateral and posterior side of the tibia in
the (D) AP and (E) lateral views.

an abnormally wide high signal in the bone marrow (Fig

1C) (Table 1).
In the second MRI scans, taken 4 weeks after the first
examination in five patients with shin splints, the abnor-
mally high signal area was reduced in all sequences, com-
pared with that observed on the initial MRI scans. Al-
though those five patients continued sports activity, no
second MRI scans showed the development of abnormal-
ity or extension of the abnormally high signal in the bone
marrow such as observed in earlier MRI scans (Fig 1C). In
all patients with shin splints, T1-weighted and T2-
weighted MRI scans did not show any detectable abnormal
findings.
Bone scintigraphs revealed abnormal local uptake in all
five patients with stress fractures who had a bone scan (Fig
5A). Nine patients with shin splints had a bone scan. Bone
scintigraphs revealed longitudinal linear uptake in four
patients and longitudinally spindle-shaped uptake in one
patient with shin splints (Fig 5B). In the other four patients
with shin splints, no bone scintigraphs showed abnormal
uptake. Magnetic resonance imaging scans of all five pa-
tients with shin splints, which revealed abnormal longitu-
dinal uptake in the bone scan, showed a linear abnormally
high signal along the medial aspect of the bone marrow
(Fig 1B). Three cases of normal bone scintigraphs showed Fig 3. An axial MRI scan of a patient with a stress fracture
a linear abnormally high signal along the medial posterior shows an abnormal high signal extending through the bone
surface of the tibia (Fig 1A). marrow.
 Clinical Orthopaedics
264 Aoki et al and Related Research

Fig 4. (A) An axial MRI scan of a patient with shin splints shows an abnormally high signal along the medial posterior surface of
the tibia and the medial aspect of the bone marrow. (B) A coronal MRI scan also shows a linear abnormally high signal intensity
along the medial periosteal region and the medial aspect of the bone marrow. A radiograph obtained 4 weeks after the first
examination did not show the periosteal reaction in the (C) AP and (D) lateral views.

DISCUSSION Complete rest from the activity may not be necessary in

Chronic leg pain frequently disrupts the athlete’s ability to
participate in sports activities. Accurate diagnosis is es-
sential in the early phase after the onset of pain to apply
specific treatment and to ensure an early return to sports
activity. Commonly, chronic leg pain in athletes is caused
by stress fractures, shin splints (medial tibial stress syn-
drome), compartment syndrome, tendinitis, venous dis-
ease, and arterial occlusion.4 According to Styf, 21 shin
splints are the most common cause of posterior medial leg
pain. Shin splints generally are considered to be an over-
use syndrome involving the fascia of the soleus muscle
and the tibialis posterior muscle. Almost all athletes with
shin splints can be treated conservatively with success.
most patients with shin splints. A gradual increase of ac-
tivity should be emphasized to prevent recurrence. How-
ever, relative rest is an important treatment to prevent
progression of the fracture in athletes with tibial stress
fractures. Occasionally, patients require immobilization or
must be nonweightbearing until acute symptoms are re-
lieved. Therefore, patients with shin splints or tibial stress
fractures require different treatment methods and periods.
Nielsen et al16 think the term stress fracture should be
used only for patients who have an intensive abnormal
finding on bone scintigraphs and callus formation as seen
on radiographs taken at followup. Plain radiographs in
patients with stress fractures show periosteal new bone
formation in the late phase. However, almost all initial

TABLE 1. Magnetic Resonance Imaging Scans and Patients With Stress Fractures or Shin Splints
Diagnosis MRI Scans
A linear abnormally high signal along the medial posterior An abnormally wide high signal in the localized bone
surface of the tibia (Fig 1A) marrow (Fig 1C)
A linear abnormally high signal along the medial aspect of the
bone marrow (Fig 1B)

Stress fracture 0 patients 8 patients

Shin splints 14 patients 0 patients
Number 421
April 2004
Fig 5. Bone scintigraphs revealed abnormal local uptake in the (A) AP and (B) lateral views of the patient with stress fracture, and
longitudinal linear uptake in the (C) AP and (D) lateral views of the patient with shin splints.
radiographs are negative.20 In the current study, therefore,
the diagnosis of stress fracture was defined by positive
findings in consecutive radiographs.
Bone scintigraphy is recommended for the diagnosis of
stress fractures in the early stage.14 The degree of abnor-
mal uptake depends on the rate of bone turnover and local
blood flow. The sensitivity of bone scans is high for stress
fractures, and bone scans of shin splint lesions occasion-
ally show linear longitudinal uptake.17 However, bone
scintigraphic findings of stress reaction are nonspecific.12
Some authors reported that 20–40% of the lesions seen on
scintigraphs were asymptomatic.16,25 It is difficult to dif-
ferentiate shin splints from stress fractures in the early
phase with a bone scan.14 Examination with a bone scan is
more invasive than MRI.
In the current study, STIR or fat-suppressed MRI scans
showed an abnormally wide high signal at a localized area
of bone marrow in all patients with stress fractures. The
Magnetic Resonance Imaging in Stress Fractures and Shin Splints 265
fat-suppressed MRI scans of stress fractures consistently
showed an intramedullary wide high signal, which was
continuous with the cortex. However, no MRI scans of
shin splints showed an abnormally wide high signal at the
bone marrow as in stress fractures. Therefore, the current
study showed that STIR fat-suppressed MRI scans are
useful in early discrimination between stress fractures and
shin splints, because these MRI scans can distinguish the
two diseases before radiographs show a detectable perios-
teal reaction of the tibia.
Stress fractures exist when microfractures or overt frac-
tures appear in the bone marrow and the cortex because of
the inability to adapt to slow rhythmic stress applied in a
subnormal manner.3,20 In the current study, the MRI fea-
tures of stress fractures included an intramedullary area of
high signal intensity that is continuous with the cortex on
the STIR fat-suppressed image. This corresponds to areas
of fracture, hemorrhage, and edema.24 Characteristic find-

266 Aoki et al
ings on fat-suppressed MRI scans of stress fractures such
as an abnormally wide high signal in the localized bone
marrow could be seen in the early phase before radio-
graphs showed a detectable periosteal reaction in the tibia.
Lee and Yao10 reported that high signal intensity within
the marrow on T2-weighted images was most prominent
when MRI was done within 3 weeks from the onset of
symptoms. Although T1-weighted and T2-weighted MRI
scans could show an abnormally wide signal in the bone
marrow in stress fractures, the abnormal finding was more
detectable in STIR fat-suppressed MRI. Tyrrell and Dav-
ies24 also reported that marrow edema was best shown on
the STIR sequence.
The MRI scans in patients with shin splints were clas-
sified into three groups. No relationship was found be-
tween the MRI scans and the duration of symptoms. A
subperiosteal high signal (Fig 1A) representing subperios-
teal edema implies that one cause of shin splints may be
traction periostitis along the insertion of the soleus fascia
and tibialis posterior.5,13,15 Abnormal signal intensity in
the bone marrow (Fig 1B) is considered to be secondary to
edema or hemorrhage related to microdamage and the as-
sociated reparative response. Schweitzer and White18 re-
ported that altered weightbearing could be a cause of in-
creased medullary signal intensity on MRI scans. Johnell
et al9 reported the increased metabolic bone activity at the
medial edge of the tibia in patients with shin splints
through the biopsy specimens. The linear bone marrow
signal change (Fig 1B) may relate to changes in the in-
creased bone metabolism of bone marrow. The patho-
mechanism of shin splints may be varied. Traction peri-
ostitis and increased bone metabolism of cortical bone and
marrow, which are recognized as stress reactions of bone,
may be two important pathomechanisms of shin splints.
Magnetic resonance imaging may help to differentiate be-
tween these pathomechanisms. Consequently, treatment of
patients with shin splints should be decided from the find-
ings of MRI.
The relationship between shin splints and stress frac-
tures is controversial. Some authors think that shin splints
may result in a stress fracture, whereas others contend that
the stress reaction of bone actually is the underlying origin
of the syndrome.6,9,11 In the current study, the MRI scans
showed a linear bone marrow signal change in some
patients with shin splints. Some authors reported that
shin splints were part of the continuum of fatigue damage
to stress fractures.1,8 These authors implied that pa-
tients with shin splints whose MRI scans show linear high
signal intensity in the bone marrow may experience stress
fractures when strenuous activity is not ceased. In the
current study, however, no patients with shin splints
had stress fractures at the late phase despite continuing
Clinical Orthopaedics
and Related Research
vigorous sports activity. Even patients with a linear
high signal in the medial bone marrow in MRI scans
did not have findings of stress fractures on the sec-
ond MRI scans or callus formation on the consecutive
plain radiographs. Therefore, we think that although shin
splints may have some relation to a stress reaction of bone,
they are likely different clinical entities from stress frac-
tures.
There are some limitations in this study. The first limi-
tation is the small number of patients and few patients
followed up with MRI. However, all patients in the current
study had abnormal signs on MRI scans. Furthermore, use
and results of second MRI scans and consecutive radio-
graphs, such as in the current study, have not been re-
ported. Although all patients should be followed up with
MRI, costs of the second MRI examination for the same
disease were expensive. The second limitation is that this
study involved patients participating in various levels of
sports. However, all patients in this study had a history of
vigorous exercise, including running and jumping, before
their symptoms occurred and there was no difference in
activity level between patients with stress fractures and
patients with shin splints. The third limitation is that the
interval between the onset of symptoms and MRI varied.
However, the mean duration of symptoms before MRI was
short in both groups and there was no significant differ-
ence between the patients with stress fractures and patients
with shin splints. This study showed that fat-suppressed
MRI is useful in early discrimination between stress frac-
tures and shin splints before radiographs show a detectable
periosteal reaction of the tibia.
References
1. Anderson MW, Ugalde V, Batt M, et al: Shin splints: MR appear-
ance in a preliminary study. Radiology 204:177–180, 1997.
2. Brukner P: Exercise-related lower leg pain. Bone. Med Sci Sports
Exerc 32(3 Suppl)S15–26, 2000.
3. Carter DR, Caler WE: A cumulative damage model for bone frac-
ture. J Orthop Res 3:84–90, 1985.
4. Clanton TO, Solcher BW: Chronic leg pain in the athlete. Clin
Sports Med 13:743–759, 1994.
5. Detmer DE: Chronic shin splints: Classification and management of
medial tibial stress syndrome. Sports Med 3:436–446, 1986.
6. Devas MB: Stress fractures of the tibia in athletes or “shin soreness.
J Bone Joint Surg 40B:227–239, 1958.
7. Engber W: Stress fractures of the medial tibial plateau. J Bone Joint
Surg 59A:767–769, 1977.
8. Fredericson M, Bergman AG, Hoffman KL, et al: Tibial stress
reaction in runners: Correlation of clinical symptoms and scintig-
raphy with a new magnetic resonance imaging grading system. Am
J Sports Med 23:472–481, 1995.
9. Johnell O, Rausing A, Wendeberg B, et al: Morphologic bone
change in shin splints. Clin Orthop 167:180–184, 1982.
10. Lee JK, Yao L: Stress fractures: MR imaging. Radiology 196:217–
220, 1988.
11. McBride A: Stress fractures in runners. Clin Sports Med 4:737–752,
1985.
Number 421
April 2004 Magnetic Resonance Imaging in Stress Fractures and Shin Splints
12. Matheson GO, Clement DB, McKenzie DC, et al: Scintigraphic
uptake of 99mTc at non-painful sites in athletes with stress fractures:
The concept of bone strain. Sports Med 4:65–75, 1987.
13. Michael RH, Holder LE: The soleus syndrome: A cause of medial
tibial stress (shin splints). Am J Sports Med 13:87–94, 1984.
14. Monteleone GP: Stress fractures in the athlete. Orthop Clin North
Am 26:423–432, 1995.
15. Mubarak SJ, Gould RN, Lee YH, et al: The medial tibial stress
syndrome: A cause of shin splints. Am J Sports Med 10:201–205,
1982.
16. Nielsen MB, Hansen K, Holmes P, et al: Tibial periosteal reactions
in soldiers: A scintigraphic study of 29 cases of lower leg pain. Acta
Orthop Scand 62:531–534, 1991.
17. Rupani HD, Holder LE, Espinola DA, et al: Three-phase radionu-
clide bone imaging in sports medicine. Radiology 156:187–196,
1985.
18. Schweitzer ME, White LM: Does altered biomechanics cause mar-
row edema? Radiology 198:851–853, 1996.
267
19. Slocum DB: The shin splint syndrome: Medical aspects and differ-
ential diagnosis. Am J Surg 114:875–881, 1967.
20. Stering JC, Edelstein DW, Calvo RD, et al: Stress fractures in
the athlete-diagnosis in management. J Sports Med 14:336–346,
1992.
21. Styf J: Diagnosis of exercise-induced pain in the anterior aspect of
the lower leg. Am J Sports Med 16:165–169, 1988.
22. Styf J: Chronic exercise- induced pain in the anterior aspect of
the lower leg: An overview of diagnosis. Sports Med 7:331–339,
1989.
23. Sullivan D, Warren RF, Pavlov H, et al: Stress fractures in 51
runners. Clin Orthop 187:188–192, 1984.
24. Tyrrell PNM, Davies AM: Magnetic resonance imaging appear-
ances of fatigue fractures of the long bones of the lower limb. Br
J Radiol 67:332–338, 1994.
25. Zwas ST, Elkanovitch R, Frank G: Interpretation and classification
of bone scintigraphic findings in stress fractures. J Nucl Med 28:
452–457, 1987.