Ultrasonographic measurement of cheek-to-cheek

diameter in fetal growth disturbances

Jacques S. Abramowicz, MD, David M. Sherer, MD, and James R. Woods, Jr., MD
Rochester, New York

OBJECTIVE: Our purpose was to assess the value of the cheek-to-cheek diameter and the
cheek-to-cheek diameter/biparietal diameter ratio in the ultrasonographic detection of abnormal fetal
growth.
STUDY DESIGN: The cheek-to-cheek diameter and cheek-to-cheek diameter/biparietal diameter ratio
were examined in 21 small-for-gestational-age « 1Oth percentile for estimated weight) and 87
large-for-gestational-age (> 90th percentile) fetuses. Statistical analysis consisted of Student's t test
comparison between means and analysis of covariance for comparison of regression lines slope
between these 108 fetuses and previously published nomograms of appropriate-for-gestational-age
fetuses.
RESULTS: The mean cheek-to-cheek diameters in small-for-gestational-age fetuses were significantly
smaller than in appropriate-for-gestational-age fetuses (p < 0.0001). In large-for-gestational-age fetuses
the mean cheek-to-cheek diameters were significantly larger (p < 0.005). Although
large-for-gestational-age fetuses of diabetic mothers exhibited higher cheek-to-cheek diameter/biparietal
diameter ratios than did appropriate-for-gestational-age fetuses (p < 0.0001), in fetuses of nondiabetic
mothers this ratio was only minimally larger (p < 0.05).
CONCLUSION: The cheek-to-cheek diameter and cheek-to-cheek diameter/biparietal diameter ratio are
innovative ultrasonographic parameters for detecting abnormal fetal growth. Futhermore, the
cheek-to-cheek diameter/biparietal diameter ratio permits inSight into the possible underlying
pathophysiologic mechanisms of fetal macrosomia. (AM J OBSTET GVNECOL 1993; 169:405-8.)

Key words: Abnormal fetal growth, cheek-to-cheek diameter, intrauterine growth retardation,
macrosomia, ultrasonography

Fetal cheek-to-cheek diameter has recently been de-
scribed in the ultrasonographic assessment of fetal
growth. I In fetuses with normal growth this measure-
ment has been shown to linearly correlate with gesta-
tional age. In the same study the cheek-to-cheek diam-
eterlbiparietal diameter (BPD) ratio between 20 and 40
weeks' gestation was demonstrated to be almost inde-
pendent of gestational age. The present study was
designed to assess the value of measurements of cheek-
to-cheek diameter and cheek-to-cheek diameter/BPD
ratio in fetuses with abnormal growth.
Material and methods
One hundred eight fetuses were entered into the
study conducted at Strong Memorial Hospital, a re-
gional perinatal referral center. Commercially available
real-time equipment was used (AI 5200, Acoustic Imag-
From the Division of Maternal-Fetal Medicine, Department of Ob-
stetrics and Gynecology, Strong Memorial Hospital, The University of
Rochester School of Medicine and Dentistry.
Received for publication September 30, 1992; revised February 17,
1993; accepted February 28, 1993.
Reprint requests: Jacques S. Abramowicz, MD, Department of Obstet-
rics and Gynecology, Division of Maternal and Fetal Medicine,
Strong Memorial Hospital, Box 668, University of Rochester School
of Medicine and Dentistry, 601 Elmwood Ave., Rochester, NY 14642.
Copyright © 1993 by Mosby-Year Book, Inc.
0002-9378/93 $1.00 + .20 6/1/46909
ing, Phoenix; ATL Ultramark 9, ATL, Bothell, Wash.;
Aloka 650, Corometrics, Wallington, Conn.). Gesta-
tional age was obtained in all cases from a well-defined
last menstrual period with confirmation by ultrasonog-
raphy before 20 weeks. Fetal measurements included
cheek-to-cheek diameter, BPD, head circumference, ab-
dominal circumference, and femur length obtained by
standard methods. The method for obtaining the cheek-
to-cheek diameter has been described in detail. I A
coronal view of the fetal face at the level of the lips and
nostrils provides the measurement plane (Fig. 1). In the
current study only fetuses at gestational age between 25
and 42 weeks and with estimated fetal weights < 10th or
> 90th percentile were considered. None of the fetuses
from the previous study were included. Estimated fetal
weight was obtained from BPD and abdominal circum-
ference measurements." Data were analyzed for com-
parison between means and by analysis of covariance
for regression lines of slopes with Student's t test; a value
of p < 0.05 was considered statistically significant.
Results
Twenty-one fetuses were small for gestational age
(SGA) and 87 were large for gestational age (LGA).
These fetuses had ultrasonographic estimation and
postnatal confirmation of deviant growth. Table I rep-
resents distribution of these fetuses by gestational age.

405
406 Abramowicz, Sherer, and Woods August 1993
Am J Obstet Gynecol

Fig. 1. Cheek-to-cheek diameter. Large arrows (e), Each cheek; smalln arrow (n), nostril.

Table I. Distribution of SeA and LeA fetuses LeA fetuses demonstrated significantly larger mean
by gestational age cheek-to-cheek diameters when compared with control
fetuses with normal growth (p < 0.005). These LeA
Gestational age (wk) SGA (No.) LGA (No.)
fetuses could be separated into two groups by their
25 5 respective cheek-to-cheek diameter/B PD ratios: fetuses
26 2
27 3
of insulin-dependent diabetic mo thers had mean ratios
28 6 well above controls (p < 0.000 1), whereas LeA fetuses
29 2 of nondiabetic mothers demonstrated higher mean ra-
30 6
31 2 2
tios than control but smaller ratios than fetuses of
32 I 6 diabetic mothers (p < 0.05). Comparison of slopes by
33 5 analysis of covariance demonstrated parallelism be-
34 2 4
35 3 8
tween the slopes (p = 0.47) and intercepts (p = 0.17),
36 6 6 but a tendency for the LeA fetuses of nondiabetic
37 2 9 mothers to have larger cheek-to-cheek diameters than
38 4 7
39 3
controls was demonstrated (one-sided t-test, p =
40 5 0.085). For LGA fetuses of diabetic mothers, the same
41 6 results were demonstrated with parallelism of the two
42 I 2
TOTAL 21 87 regression lines (p for the slopes and the intercepts 0.45
and 0.15, respectively). However, LGA fetuses demon-
strated a trend toward larger cheek-to-cheek diameters
than the control (one-sided test, p = 0.075). Mean
When the regression equation for the nomogram cheek-to-cheek diameters were roughly 0.2 cm larger
previously published was used, cheek-to-cheek diame- than control over the range of gestational age in LGA
ter = - 0.908 + 0.195 gestational age (for gestational fetuses of nondiabetic mothers and 0.87 cm larger in
age ~ 20 weeks), mean cheek-to-cheek diameters of LeA fetuses of diabetic mothers. In Fig. 3 regression
SeA fetuses were found to be significantly smaller lines from LGA fetuses of diabetic and nondiabetic
(P < 0.0001) than for fetuses with appropriate growth mothers and regression line for controls are displayed.
of the same gestational age. When we used the pub- During this study three patients with poor prenatal
lished nomogram fOl" the cheek-Io-cheek diameter/BPD care had their first prenatal visit at 34, 36, and 39
ratio, SeA fetuses showed smaller ratios (p < 0.000 I). weeks, respectively. At that time they were examined by
Slope comparison demonstrated lhat the fetuses with ultrasonography. All three fetu ses had estimated fe-
intrauterine growth retardation had different (smaller) tal weights > 90th percentile, cheek-to-cheek diameter,
cheek-to-cheek diameters than ap propriately grown fe- and cheek-to-cheek diameter/BPD ratios well above the
tuses (p = 0.0068). Fig. 2 displays the regression line of 90th percentile. Diabetes mellitus was suspected on the
cheek-to-cheek diameter versus gestational age in SeA basis of these findings and confirmed in each case by
fetuses and the regression line for predicted means. chemical testing. Postnatal exam ination of these infants
Volume 169, Number 2, Part I Abramowicz, Sherer, and Woods 407
Am J Obstet Gynecol

eeD (em)
8'-~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Control
-£--- SGA

2~--------~--------~----------L---------~--------~

25 30 35 40
Gestational Age (weeks)
Fig. 2. Regression line of cheek-to-cheek diameter (CCD) versus gestational age in SGA fetuses
compared with regression line of predicted means (control).

eeD (em)
8,-~~--------------~----------------------~---------,

4
Control
-+- LGA Diabetic Mother
-A- LGA Non Diab. Mother
2L----------L----------L----------L~~------~--~----~

25 30 35 40
Gestational Age (weeks)
Fig. 3. Regression line of cheek-to-cheek diameter (CCD) versus gestational age in LGA fetuses of
nondiabetic mothers (triangles) and LGA fetuses of diabetic mothers (stars) compared with regression
line of predicted means (control).

demonstrated the often referred to "chubby cheeks"
(Fig. 4).
Comment
Results of this study extend our previous experience
with this new ultrasonographic parameter for assessing
fetal subcutaneous fat tissue in association with growth
disturbances. In our first study a small sample of fetuses
with abnormal growth was considered: 10 were SGA
and 39 macrosomic (lOin diabetic mothers and 29 in
nondiabetic). Nine appropriate-for-gestational-age fe-
tuses of diabetic mothers were also included. A trend
408 Abramowicz. Sherer. and Woods
Fig. 4. Neonatal photograph of infant of a diabetic mother.
Note classic "chubby cheeks." This neonate's cheek-to-cheek
diameter/BPD ratio at 41 weeks was 0.98 . Birth weight was
4735 gm.
for smaller cheek-to-cheek diameters was demonstrated
in the SeA group. In LeA fetuses cheek-to-cheek di-
ameters were larger than normal controls. In the cur-
rent study we present our observations from a substan-
tially larger number of patients.
The subcutaneous. nonmuscular component of the
fetal cheeks consists of white fat, which is sensitive to
changes in nutritional status." 4 In SeA fetuses subcu-
taneous fat is depleted. and, as expected. the cheek-to-
cheek diameter is markedly reduced. The bony frame is
much less affected"' 6; this is demonstrated by a
smaller-than-normal cheek-to-cheeklBPD ratio. The ra-
tio would have been maintained within normal range
had both subcutaneous tissues and biparietal diameter
diminished to a similar degree.
The situation is more complex in LeA fetuses. De-
pending on the cause, either subcutaneous tissues or
bony frame or both can be affected. 7 In "metabolic
macrosomia," such as occurs in fetuses of diabetic
mothers. the main overgrowth is in subcutaneous tis-
sues."' 9 Cephalic growth is not affected. 9 . 10 Therefore
the cheek-to-cheek diameter and the cheek-to-cheek
diameter/BPD are expected to be larger than normal.
In "constitutional macrosomia" growth is accelerated.
albeit in a more balanced manner. and therefore the
cheek-to-cheek diameter is expected to be above nor-
mal, but the ratio of cheek-to-cheek diameter to BPD
will be more in the normal range because both these
parameters are proportionally larger.
The importance of diagnosing fetal macrosomia and
differentiating between the two distinctly different types
has repeatedly been stressed. II Perinatal morbidity and
mortality and birth complications are much higher in
August 1993
Am J Obstet Gynecol
fetuses of diabetic mothers. 12. 13 Results of the current
study demonstrate that the difference between the two
types of macrosomia can be detected ultrasonographi-
cally.
Often ultrasonographically derived fetal weight esti-
mation is inaccurate because actual fetal body compo-
sition is not taken into account with standard biometric
methods. '4 If parameters to measure nonbony struc-
tures were available, this could enhance the precision of
fetal weight estimation. In a recent review of macroso-
mia the author refers to "some yet unspecified indicator
of soft tissue mass," which may need to be defined.
measured. and incorporated in fetal weight estima-
tion. '5 We believe that the cheek-to-cheek diameter
represents this elusive ultrasonographic indicator and
that, in conjunction with other established parameters.
it should be used to improve ultrasonographic estima-
tion of fetal weight.
REFERENCES
1. AbramowiczJS. Sherer OM. Bar-Tov E. WoodsJRJr. The
cheek-to-cheek diameter in the ultrasonographic assess-
ment of fetal growth. AM J OBSTET GYNECOL 1991; 165:846-
52.
2. Warsof SL. Gohari p. Berkowitz RC. Hobbins JC. The
estimation of fetal weight by computer-assisted analysis.
AMJ OBSTETGYNECOL 1977;128:881-92.
3. Vorherr A. Factors influencing fetal growth . AM J OBSTET
GYNf.COL 1982;142:577-88.
4. Roord JJ. Ramaekers LHT. Quantification of intrauterine
malnutrition. Bioi Neonate 1978;33:273-7.
5. Abramowicz JS. Jaffe R. Warsof SL. Ultrasonographic
measurement of fetal femur length in growth disturbances.
AM J OBSTET GYNECOL 1989; 161 : 1137 -9.
6. Villar J. Belizan JM . The timing factor in the pathophysi-
ology of the intrauterine growth retardation syndrome.
Obstet Gynecol Surv 1982;37:499-506.
7. Hadlock FP. Deter RL. Harrist RB. Sonographic detection
of abnormal fetal growth patterns. Clin Obstet Gynecol
1984;27:342-51.
8. Naeye RL. Infants of diabetic mothers: a quantitative.
morphologic study. Pediatrics 1955;35:980-8.
9. Landon MB. Mintz MC. Gabbe SG. Sonographic evalu-
ation of fetal abdominal growth: predictor of the large-
for-gestational-age infant in pregnancies complicated by
diabetes mellitus. AM J OBSTET GYNECOL 1989; 160: 115-
21.
10. Ogata ES. Sabbagh a RE. Metzger BE. Phelps RL. Depp R.
Freinkel N . Serial ultrasonography to assess evolving fetal
macrosomia: studies in 23 pregnant diabetic women.
JAMA 1980;243:2405.
11. Deter RL. Hadlock FP. Use of ultrasound in the detection
of macrosomia: a review. JCU 1985;13:519-24.
12. Kitzmiller J . Cloherty J. Younger MD. et al. Diabetic
pregnancy and perinatal morbidity. AM J OBSTET GYNECOL
1978; 131 :560-80.
13. Langer O. Berkus MD. Huff RW. Samuel off A. Shoulder
dystocia: should the fetus weighing 2: 4000 grams be
delivered by cesarean section? AM J OBSTET GYNECOL 1991;
165:831-7.
14. Bernstein 1M. Catalano PM. Influence of fetal fat on the
ultrasound estimation of fetal weight in diabetic mothers.
Obstet Gynecol 1992;79:561-3.
15. Deter RL. Definition. epidemiology and classification of
macrosomia. In: Divon MY. ed. Abnormal fetal growth.
New York: Elsevier. 1991:75-82.