April 28, 2008
Can Prenatal Care Prevent Childhood Obesity?
by Rebecca L. Utz, PhD, Assistant Professor of Sociology, University of Utah
Risk factors for obesity include, but are not
limited to, low socioeconomic status, minority race/ethnicity status, increased
caloric consumption, and sedentary behavior (Robinson,
1999; Speiser, 2005; Zhang & Wang, 2004). Recent research adopting the “fetal origins hypothesis” also
suggests that the development of later-life health outcomes such as obesity may
be programmed prior to birth (Salsberry & Reagan, 2004). To understand the multidimensional nature of
obesity risk among adolescents (age 15-19), we developed an epidemiological
study to investigate the unique role that prenatal and maternal characteristics
may play in increasing or decreasing a child’s risk for becoming obese.
The Research Questions
- Do the resources of the mother at the time
of the birth (such as her age, marital status, or education level)
influence one’s obesity risk?
- Do features associated with the birth
itself (such as whether it was a vaginal or cesarean delivery, whether a
child was the mother’s first-born, the child’s APGAR score 1-minute at
birth, the child’s birth weight and gestation) increase one’s obesity
risk?
- Do prenatal behaviors adopted by the mother
(such as initiation of prenatal health care, smoking during pregnancy,
pregnancy related weight gain, and pre-pregnancy body weight) influence a
child’s obesity risk?
The motivation of this project comes from a
desire to understand which of these characteristics are most associated with
the development of obesity, and which are potentially modifiable and thus able
to become a specific policy recommendation to help reduce the spread of obesity
from generation to generation.
Data & Methods
Our data consists of linked birth certificates and
driver license records. The birth
records provide information about the birth mother, the prenatal care she
received, and the characteristics of the child at the time of delivery. The driver’s license provides a measure of the
child’s height and weight at ages 15 to 19.
This unique data source yielded a sample of nearly 200,000 Utahans who
were born between 1983 and 1990 and reached their late teens in the late 1990s
and 2000s.
From the driver’s license records, we calculated
a measure of body mass index (BMI) for each child. BMI is calculated as body weight in kilograms
divided by height in meters squared. Adolescents
who have a BMI-for-Age that is at or above the 95th percentile,
according to the CDC height-weight growth charts, are considered
overweight.
We then estimated a series of
statistical models that predicted which features of the prenatal environment
were associated with increased or decreased odds of being overweight during
adolescence. We also adjusted our
statistical models based on whether the adolescent was male or female, whether
s/he was white, black, Hispanic, or other race, and what year s/he was born
since the prevalence of obesity varies quite dramatically by these demographic characteristics.
Findings
First, some have contested the
validity of measuring BMI from driver’s license records (i.e., people misreport
their height and especially weight when getting a license), but the data
presented in Figure 1 suggest that the self-reported data from driver’s license
records is comparable to other established data sources measuring BMI among
adolescents in Utah. It also shows that
Utah adolescents are less likely to be overweight than their national
counterparts: Utah ranks as the 8th
most lean among adults and the leanest among children (Levi, Segal, Gadola,
2007).
Figure Notes:
A.
Overweight is defined as having a BMI > 95th percentile.
B. The Youth Risk
Behavior Surveillance Survey (YRBSS) is a telephone survey of students in
grades 9-12 which is conducted by each state.
The National YRBSS estimate is a weighted average of all state surveys.
C. Drivers license
data (DLD) provide a measure of self-report height and weight for Utah
adolescents age 15-19 (M=16.5)
D.
All data were collected in 2005.
As for which characteristics
predict a child’s risk of being overweight, our analyses offer the following
findings:
- Demographic Differences: Female adolescents are less likely to
be overweight than males (or perhaps more likely to underreport
height/weight on driver license than men). Older adolescents are more likely to be overweight than the
younger adolescents. American
Indians, Blacks, and Pacific Islanders all have over two times greater the
odds of being overweight than whites.
Hispanics also have significantly higher odds of being overweight
than non-Hispanic adolescents. Finally,
we find increased odds of being overweight by birth year: persons born in 1989-1990 have 34%
greater odds of being overweight than the persons born in the early
1980s.
- Maternal Pre-Birth Resources: A child is less likely to be overweight
had the mother been married at the time of the birth. A child born to an older mother is more
likely to be overweight during adolescence. As well, children of mothers receiving less than high school
had significantly higher odds of being overweight, while children of
mothers receiving a college degree or more had significantly lower odds.
- Birth Experiences: Birth weight is associated with obesity
risk: babies who were greater than
3500 grams (approx 7.7 pounds) were 50% more likely to be overweight than
the babies born in the 2500 to 3500 gram range (5.5-7.5 pound). Children born through a cesarean
delivery were 1.4 times more likely to be overweight than children born
through a vaginal delivery. The
most hearty or robust children at birth (i.e., those with the highest
APGAR scores at birth) were less likely to be overweight.
- Prenatal Behaviors: Those children whose mothers initiated
prenatal health care in the first trimester were significantly less likely
to be overweight than those whose mothers began prenatal appointments
after their 13th week or never received care. If a mother smoked during pregnancy,
the child is almost twice as likely to be overweight. Furthermore, if the mother gained too
much weight (as defined by the physician recommended weight categories),
the child has 65% higher odds of being overweight during adolescence,
compared to a child whose mother gained the recommended weight range or
less.
- Maternal Pre-Pregnancy Weight Status:
The most striking finding of our
analysis is the overwhelming effect that mother’s pre-pregnancy weight
status has on an adolescent’s risk of being overweight. The child of a mother who was
overweight pre-pregnancy is approximately twice as likely to be overweight
by the time they reach adolescence, while the child of a mother who was
obese pre-pregnancy was five times more likely to be overweight during
adolescence, compared to the child of a mother who had a normal pre-pregnancy
weight.
Policy Implications
We found that prenatal health care is indeed important for
the prevention of obesity among children.
As such, pregnant women ought to be encouraged to initiate prenatal
health care early in the pregnancy, to not gain too much weight during
pregnancy, and to not smoke during pregnancy.
Our analyses also revealed that the mother’s pre-pregnancy weight status
is a strong predictor of a child’s overweight risk. Although it is unclear from this analysis whether this effect
indicates a genetic/heritable trait that increases a child’s risk for being
overweight or whether it indicates a lifestyle that she exposed the child to
throughout his/her childhood, the public health message should focus on how
mothers may inadvertently be passing obesity risk from generation to
generation, setting the child on a trajectory of poor health prior to his or
her conception. Thus, couples who are
contemplating a family should be encouraged to lose weight prior to conception. Healthy weight and/or healthy lifestyles (rather
than unhealthy weight and lifestyles) need to be passed from generation to
generation to ensure that we control the spread of the obesity epidemic in
America.
References
Levi, J. Segal, L. M., & Gadola, E. (2007).
F as in fat: How obesity policies are failing in America. Report from
Trust for America’s Health.
Robinson, T. N.
(1999). Reducing children's television viewing to prevent obesity: a randomized
controlled trial. Jama, 282(16),
1561-1567.
Salsberry, P.
J. & Reagan, P. B. (2004).
Dynamics of early childhood overweight.
Peds, 116(6), 1329-1338.
Speiser, P. W.,
Rudolf, M. C., Anhalt, H., Camacho-Hubner, C., Chiarelli, F., Eliakim, A., et
al. (2005). Childhood obesity. J Clin
Endocrinol Metab, 90(3), 1871-1887.
Zhang, Q., &
Wang, Y. (2004). Trends in the association between obesity and socioeconomic
status in U.S. adults: 1971 to 2000. Obes
Res, 12(10), 1622-1632.
|