Dietary intake and physical activity in severely obese pregnancy in Scotland
View/ Open
Date
30/06/2012Author
Mohd Shukri, Nor Azwani
Shukri, Nor Azwani Mohd
Metadata
Abstract
Maternal obesity is associated with adverse effects for mothers and offspring. The
primary aim of this thesis was to assess food intake and physical activity (PA) using
validated self-administered questionnaires, and whether these were associated with
gestational weight gain (GWG) and birthweight (BWT), in severely obese (body mass index,
BMI≥40kg/m2) compared with lean pregnant women (BMI 20-25kg/m2). The secondary
aims were to validate self-reports against food diary (FD) and accelerometry; to assess the
prevalence of under or over-reporting of energy intake; and to carry out a pilot study to
assess total energy expenditure, as well as self-reporting accuracy, by using doubly-labelled
water (DLW) technique, in subgroups of participants.
Pregnant women were recruited from an ongoing study of severe obesity in pregnancy
at the Royal Infirmary of Edinburgh, UK. Assessments were done in early (12-20 weeks)
and late (28-32 weeks) pregnancy. A subgroup of women also completed questionnaires on
appetite, general nutrition knowledge, and eating behaviours. All results were adjusted for
age, parity, ethnic origin and deprivation category score.
Self-reported total energy intake was not significantly different between obese and
lean during early (median 2,444 vs 2,312 kcal/day) and late (2,173 vs 2,354 kcal/day)
pregnancy. However when validated with FD, the relative validity of the food frequency
questionnaire was lower in obese compared to lean. Under-reporting of total energy intake
was higher in obese compared to lean (49% vs 15%, P<0.01) through comparison of selfreported
energy intake with estimated total energy expenditure, and this was supported by
the DLW pilot study results. The DLW also showed possible over-reporting of PA by the
obese group. Obese women reported significantly lower appetite than lean throughout
pregnancy (P<0.01). They also had lower scores in general nutrition knowledge, but these
were no longer significant after controlling for confounders. Obese women had significantly higher scores of restrained and emotional eating behaviours than, and similar scores of
external eating behaviours to, lean. Appetite, nutrition knowledge, restraint and emotional
eating behaviours scores were not associated with food intake in either obese or lean. On the
other hand, increased intakes of total calories and fats were influenced by increasing score of
external eating behaviour in both groups. Obese women reported doing similar amounts of
total PA but significantly less of vigorous and sports and exercise activities than lean
(P<0.05). Accelerometry showed obese women had lower average activity counts/day,
although they did have significantly greater energy expenditure in light-intensity activity
than lean (P<0.01). Obese women had less GWG than lean (Mean ± SD, 5.3± 52 vs
10.8±3.7kg, P<0.001). Increased GWG was associated with increased self-reported total
energy intake in lean, but this was not seen in obese. GWG was not associated with PA in
either group. BWT was not significantly different between obese and lean (3,547±549g vs
3,567±516g). In lean, increased BWT was associated with increased energy intake and total
PA. BWT in obese was not associated with diet but with increased with PA in early
pregnancy.
In conclusion, self-reported methods were less reliable in assessment of diet and PA in
severely obese compared to lean pregnant women. These exploratory studies found that
obese women did not appear to have the same factors as lean women affecting GWG and
BWT, though this may be complicated by the poor reliability of self-reports. Therefore,
quantitative assessments such as measurement of serum micronutrient levels (to evaluate
nutritional status), and accelerometry (to assess physical activity) may be necessary in this
poorly understood population.