Content uploaded by Wei Bao
Author content
All content in this area was uploaded by Wei Bao on Oct 14, 2014
Content may be subject to copyright.
ARTICLE
Pre-pregnancy fried food consumption and the risk of gestational
diabetes mellitus: a prospective cohort study
Wei Bao &Deirdre K. Tobias &Sjurdur F. Olsen &
Cuilin Zhang
Received: 6 June 2014 /Accepted: 29 August 2014
#Springer-Verlag (outside the USA) 2014
Abstract
Aims/hypothesis Fried foods are frequently consumed in
Western countries. However, the health effects of frequent
fried food consumption in humans are not well understood.
We aimed to prospectively examine the association between
pre-pregnancy fried food consumption and risk of incident
gestational diabetes mellitus (GDM).
Methods We included 21,079 singleton pregnancies from
15,027 women in the Nurses' Health Study II cohort. Since
1991 and every 4 years thereafter, we collected diet informa-
tion, including consumption of fried foods at home and away
from home, using a validated food frequency questionnaire.
We used generalised estimating equations with log-binomial
models to estimate the RRs and 95% CIs.
Results We documented 847 incident GDM pregnancies dur-
ing 10 years of follow-up. After adjustment for age, parity,
dietary and non-dietary factors, the RRs (95% CIs) of GDM
among women who consumed total fried foods 1–3, 4–6and
≥7 times/week, compared with those who consumed it less
than once/week, were 1.13 (0.97, 1.32), 1.31 (1.08, 1.59) and
2.18 (1.53, 3.09), respectively (pfor trend <0.001). The asso-
ciation persisted after further adjustment for BMI ( pfor
trend= 0.01). When analysed separately, we found a signifi-
cant association of GDM with fried food consumption away
from home, but not with fried food consumption at home.
Conclusions/interpretation Frequent fried food consumption,
particularly away from home, was significantly associated
with a greater risk of incident GDM. Our study indicates
potential benefits of limiting fried food consumption in the
prevention of GDM in women of reproductive age.
Keywords Diet .Fried food .Gestational diabetes
mellitus
Abbreviations
AHEI-2010 Alternate Healthy Eating Index 2010
FFQ Food frequency questionnaire
GDM Gestational diabetes mellitus
GEE Generalised estimating equations
MET Metabolic equivalent
NHS Nurses’Health Study
Introduction
Fried foods are frequently consumed in Western countries [1,
2]. With globalisation, the number of food industries and
outlets that produce and sell fried food, particularly fast food,
is increasing rapidly around the world [2–4]. Frying is a
complex cooking process that modifies both the foods and
the frying medium. It reduces water content, increases energy
density, changes fatty acid composition and deterioratesfrying
Electronic supplementary material The online version of this article
(doi:10.1007/s00125-014-3382-x) contains peer-reviewed but unedited
supplementary material, which is available to authorised users.
W. Bao :C. Zhang (*)
Epidemiology Branch, Division of Intramural Population Health
Research, Eunice Kennedy Shriver National Institute of Child Health
and Human Development, National Institutes of Health, 6100
Executive Blvd, Rockville, MD 20852, USA
e-mail: zhangcu@mail.nih.gov
D. K. Tobias
Department of Nutrition, Harvard School of Public Health,
Boston, MA, USA
D. K. Tobias
Division of Preventive Medicine, Department of Medicine,
Brigham and Women’s Hospital and Harvard Medical School,
Boston, MA, USA
S. F. Olsen
Centre for Fetal Programming, Department of Epidemiology
Research, Statens Serum Institut, Copenhagen, Denmark
Diabetologia
DOI 10.1007/s00125-014-3382-x
oils, especially when reused, through the processes of oxida-
tion and hydrogenation [5]. Frying also makes food crunchy
and aromatic and improves food palatability, which may in
turn lead to excess intake [6]. Recently, frequent consumption
of fried foods has been linked to a higher risk of being
overweight and obesity in two Mediterranean cohorts [6,7].
However, there are few prospective epidemiological studies
examining the association of fried food consumption with
other health outcomes.
Gestational diabetes mellitus (GDM) is a common preg-
nancy complication characterised by glucose intolerance with
onset or first recognition during pregnancy [8]. GDM is not
only associated with short-term perinatal outcomes [9], but is
also related to long-term metabolic risk in both mothers and
their offspring [8,10,11]. Thus, it is crucial to identify
modifiable risk factors that may contribute to the prevention
of GDM. Previous studies have identified a number of dietary
and other lifestyle factors related to GDM risk [12]. Fried
foods contain some deleterious substances, e.g., oil degrada-
tion products, which may have adverse effects on glucose
homeostasis. The association between fried food consumption
and GDM risk, however, remains unknown. In this prospec-
tive cohort study, we examined the association between pre-
pregnancy fried food consumption, both at home and away
from home, and the risk of subsequent GDM.
Methods
Study population The Nurses’Health Study II (NHS II) is an
ongoing prospective cohort study of 116,671 female nurses
aged 25–44 years at study inception in 1989 [13]. The partic-
ipants received a biennial questionnaire regarding disease
outcomes and lifestyle behaviours, such as smoking status
and medication use. Follow-up for each questionnaire cycle
was greater than 90% [14]. This study has been approved by
the Partners Human Research Committee (Boston, MA,
USA), with the consent of participants implied by the return
of the questionnaires.
We included NHS II participants in this analysis if they
reported at least one singleton pregnancy lasting longer than
6 months (electronic supplementary material [ESM] Fig. 1).
The 1991 questionnaire was the first time dietary information
was collected. Thus, we set this year as the baseline for this
analysis and we only included pregnancies after the return of
the 1991 questionnaire. The 2001 questionnaire was the last
time GDM was ascertained, since the majority of NHS II
participants had passed reproductive age by then; therefore,
follow-up was through the return of the 2001 questionnaire.
Individual pregnancies were eligible if there was no GDM
reported in a previous pregnancy, or a prior diagnosis of type 2
diabetes mellitus, cardiovascular disease or cancer. We ex-
cluded pregnancies if the participant did not return at least one
pre-pregnancy food frequency questionnaire (FFQ), left more
than 70 FFQ items blank or reported unrealistic total energy
intake (<2,510 or >14,644 kJ/day, which is equivalent to <600
or >3,500 kcal/day). Women with GDM in a previous preg-
nancy were excluded because they may change their diet and
lifestyle during the next pregnancy to prevent recurrent GDM.
Exposure assessment Beginning in 1991 and every 4 years
thereafter, we asked the participants to report their usual food
intake over the previous year using a previously validated
FFQ [15–17]. For fried food consumption, we asked the
participants ‘How often do you eat fried food away from home
(e.g. French fries, fried chicken, fried fish)?’and ‘How often
do you eat food that is fried at home (exclude “Pam”-type
spray)?’Both questions had four possible frequency re-
sponses: <1/week, 1–3/week, 4–6/week or daily. We analysed
fried food consumption at home and away from home sepa-
rately, as well as their sum (total fried food consumption). In
addition, we asked the participants what kind of frying fat/oils
they usually used at home, with the possible responses as
follows: real butter, margarine, vegetable oil, vegetable short-
ening or lard.
Outcome ascertainment The NHS II participants reported
incident GDM on each biennial questionnaire through to
2001. In the case of more than one pregnancy lasting longer
than 6 months reported within a 2-year questionnaire period,
GDM status was attributed to the first pregnancy. In a prior
validation study among a subgroup of the NHS II cohort, 94%
of GDM self-reports were confirmed by medical records [13].
In a random sample of parous women without GDM, 83%
reported a glucose screening test during pregnancy and 100%
reported frequent prenatal urine screenings, suggesting a high
level of GDM surveillance in this cohort [13].
Covariates assessment Participants reported their height and
weight in 1989 and updated their weight on each biennial
questionnaire. Self-reported weight was highly correlated with
measured weight (r=0.97) in a previous validation study [18].
BMI was computed as weight in kilograms divided by height
in metres squared. Total physical activity was ascertained by
frequency of engaging in common recreational activities, from
which metabolic equivalent (MET)-hours per week were de-
rived. The questionnaire-based estimates correlated well with
detailed activity diaries in a prior validation study (r=0.56)
[19].
To assess the overall diet quality of the participants, we
derived a diet score, the Alternate Healthy Eating Index 2010
(AHEI-2010), for each participant, as described previously
[20]. Briefly, we allotted 0 to 10 points for intake of each of
the following components, with adherence to dietary recom-
mendations receiving 10 points: vegetables (servings/day),
fruits (servings/day), whole grains (g/day), sugar-sweetened
Diabetologia
beverages and fruit juice (inversely scored, servings/day), nuts
and legumes (servings/day), red/processed meat (inversely
scored, servings/day), trans-fat (inversely scored, % of ener-
gy), long-chain n-3 fats (mg/d), polyunsaturated fat (% of
energy), sodium (inversely scored, mg/day) and alcohol
(drinks/day; we assigned the highest score to moderate [0.5–
1.5 drinks/day] and the worst score to heavy [≥2.5 drinks/day]
alcohol consumers). The overall AHEI-2010 ranged from 0 to
110 points, with a higher score indicating a better diet quality
and associated with a lower risk of incident chronic diseases,
including cardiovascular disease, diabetes and cancer [20].
Statistical analysis Descriptive statistics at baseline were pre-
sented as means (SD) for continuous variables and percent-
ages for categorical variables. We used generalised estimating
equations (GEE) with log-binomials models [21]toestimate
RRs and 95% CIs. The GEE method allows us to account for
correlations among repeated observations (pregnancies) con-
tributed by a single participant [22]. In a few instances, the
log-binomials models did not converge, and log-Poisson
models [23], which provide consistent but not fully efficient
risk estimates, were used.
Pre-pregnancy covariates in the multivariable models in-
cluded age (continuous), parity (0, 1, 2 and ≥3 pregnancies
lasting ≥6 months), race/ethnicity (Caucasian, African-
American, Hispanic, Asian, other and missing/not reported),
family history of diabetes (yes, no), cigarette smoking (never,
past, current and missing/not reported), physical activity
(quartiles), total energy intake (quartiles) and diet quality as
represented by the AHEI-2010 dietary pattern score (quar-
tiles). All these variables, except race/ethnicity and family
history of diabetes that were reported in 1989 only, were
updated during the follow-up (ESM Fig. 2). In addition, we
further adjusted for biennially updated measures of BMI (<21,
21–22.9, 23.0–24.9, 25.0–26.9, 27.0–28.9, 29.0–30.9, 31.0–
32.9, 33.0–34.9, ≥35.0 kg/m
2
and missing/not reported). BMI
was modelled separately because it may also be an interme-
diate on the pathway between fried food consumption and
GDM risk. We derived covariates from the questionnaire
preceding each pregnancy. Categorical covariates included
an indicator variable for missing data, if necessary. Less than
3% of covariate data was missing for a given variable. Tests
for linear trends were conducted across categories of fried
food consumption by assigning the median intake for each
category and fitting this continuous variable in the models.
To evaluate effect modification, we performed stratified
analyses according to age (<35 vs ≥35), parity (nulliparous
vs parous), family history of diabetes (yes vs no), pre-
pregnancy physical activity (< median vs ≥median), pre-
pregnancy overall diet quality (AHEI-2010 dietary pattern
score <median vs ≥median) and pre-pregnancy overweight
status (BMI <25 vs ≥25 kg/m
2
). We tested interactions be-
tween these factors and fried food consumption by adding
multiplicative interaction terms of the binary effect modifiers
and continuous linear trend variables in the multivariable
models. For fried food consumption at home, we also per-
formed a stratified analysis by types of frying oils.
All statistical analyses were performed with SAS software
(version 9.2; SAS Institute, Cary, North Carolina, USA). A
pvalue <0.05 was considered statistically significant.
Results
During the 10 years of follow-up, we documented 847 inci-
dent GDM pregnancies among 21,079 eligible singleton preg-
nancies from 15,027 women. At baseline in 1991, women
with higher consumption of fried foods were younger, less
likely to be nulliparous and white, and more likely to be
current smokers (Table 1). They had a lower diet quality as
represented by the AHEI-2010 dietary pattern score, less
physical activity and higher BMI.
We observed an elevated risk of GDM in association with
higher total fried food consumption (Table 2). After adjustment
for age, parity, race/ethnicity, family history of diabetes, ciga-
rette smoking, physical activity, total energy intake and diet
quality, the RR (95% CI) of GDM among women who con-
sumed total fried foods ≥7 times/week was 2.18 (1.53, 3.09)
compared with those who consumed it less than once/week (p
for trend <0.001). After additional adjustment for BMI, the
association was moderately attenuated but remained signifi-
cant, with a corresponding RR (95% CI) of 1.88 (1.34, 2.64)
(pfor trend= 0.01). Further adjustment for French fried potatoes
consumption or trans-fat intake did not substantially alter the
results. In addition, the association was not significantly mod-
ified by other risk factors of GDM, such as age, parity, family
history of diabetes, physical activity, overall diet quality or
overweight status (all pfor interaction >0.20).
When analysed separately, we found a significant associa-
tion of GDM with fried food consumption away from home,
but not with fried food consumption at home (Table 3). The
multivariable-adjusted RR (95% CI) of GDM comparing con-
sumption of ≥4 times/week with less than once/week of fried
food consumption away from home was 1.63 (1.15, 2.33) ( p
for trend <0.001). The association remained significant after
additional adjustment for BMI, with a corresponding RR (95%
CI)of1.46(1.03,2.07)(pfor trend = 0.03). Although women
who consumed fried food consumption at home ≥4 times/week
seemed to have a 30% (adjusted RR 1.30, 95% CI 0.81, 2.10)
higher risk of GDM compared with women who consumed less
than once/week, the association was not statistically significant.
Additional adjustment for types of frying oils at home did not
substantially alter the results. We also conducted a sensitivity
analysis by excluding women who were currently pregnant at
the time of FFQ and found similar results.
Diabetologia
Tabl e 1 Age-standardised characteristics of the study population (n= 15,027) in 1991, according to frequency of total fried food consumption
Characteristic Frequency of total fried food consumption pvalue for trend
<1/week 1–3/week 4–6/week ≥7/week
No. of participants 7,742 4,986 2,027 272 –
Age, years 32.3 (3.3) 31.8 (3.2) 31.4 (3.2) 31.1 (3.2) <0.001
White,% 94939187<0.001
Family history of diabetes, % 11 11 11 16 0.11
Nulliparous, % 38 35 33 28 <0.001
Current smoker, % 8 9 11 12 <0.001
Alcohol, g/day 3.1 (5.1) 3.0 (5.2) 2.9 (5.6) 2.4 (4.5) 0.03
BMI, kg/m
2
23.0 (3.9) 23.7 (4.5) 24.3 (5.0) 24.7 (5.7) <0.001
Physical activity, MET-h/week 26.4 (31.7) 20.8 (25.5) 18.3 (24.4) 16.1 (22.8) <0.001
Total energy intake, kJ/day 7,293 (2,213) 7,828 (2,272) 8,473 (2,385) 9,217 (2,590) <0.001
Carbohydrate, %E 51.9 (7.3) 49.8 (6.8) 48.4 (7.0) 46.3 (7.7) <0.001
Total protein, %E 19.7 (3.4) 18.8 (3.1) 18.4 (3.0) 18.2 (3.3) <0.001
Total fat, %E 29.2 (5.2) 32.1 (4.9) 33.9 (5.1) 36.3 (5.6) <0.001
Saturated fat, %E 10.6 (2.3) 11.6 (2.2) 12.2 (2.3) 13.0 (2.5) <0.001
Monounsaturated fat, %E 10.8 (2.2) 12.2 (2.1) 13.1 (2.2) 14.4 (2.4) <0.001
Polyunsaturated fat, %E 5.3 (1.3) 5.6 (1.3) 5.7 (1.3) 5.8 (1.2) <0.001
trans-Fat, %E 1.3 (0.4) 1.8 (0.5) 2.0 (0.6) 2.5 (0.8) <0.001
Cholesterol, mg/day
a
233 (66) 239 (62) 248 (62) 253 (67) <0.001
Glycaemic index
a
53.5 (3.3) 54.4 (3.1) 54.8 (3.0) 55.4 (2.9) <0.001
Glycaemic load
a
126 (21) 122 (20) 119 (20) 114 (22) <0.001
Tot al f ib re , g/ da y
a
19.2 (5.8) 17.2 (4.8) 16.2 (4.4) 14.9 (3.9) <0.001
Magnesium, mg/day
a
336 (74) 307 (68) 288 (64) 264 (58) <0.001
Haem iron, mg/day
a
1.0 (0.4) 1.1 (0.4) 1.2 (0.4) 1.3 (0.5) <0.001
Potassium, mg/day
a
3,013 (504) 2,780 (468) 2,679 (459) 2,558 (461) <0.001
Calcium, mg/day
a
1,139 (434) 1,034 (406) 938 (373) 822 (352) <0.001
Vitamin C, mg/day
a
273 (297) 230 (250) 201 (225) 188 (259) <0.001
Vitamin E, mg/day
a
21.6 (46.9) 19.7 (42.3) 18.3 (39.9) 17.2 (37.2) 0.04
Red meat, servings/day 0.6 (0.4) 0.8 (0.5) 1.1 (0.6) 1.5 (0.9) <0.001
Poultry, servings/day 0.5 (0.3) 0.5 (0.3) 0.4 (0.3) 0.4 (0.3) 0.01
Fish, servings/day 0.2 (0.2) 0.2 (0.2) 0.2 (0.2) 0.2 (0.3) 0.002
Eggs, servings/day 0.2 (0.2) 0.2 (0.2) 0.2 (0.2) 0.2 (0.2) <0.001
Low-fat dairy, servings/day 1.7 (1.3) 1.5 (1.2) 1.3 (1.2) 1.1 (1.2) <0.001
High-fat dairy, servings/day 0.9 (0.9) 1.0 (0.9) 1.1 (1.0) 1.1 (1.0) <0.001
Nuts, servings/day 0.3 (0.4) 0.3 (0.3) 0.3 (0.3) 0.2 (0.2) 0.04
Legumes, servings/day 0.4 (0.3) 0.3 (0.3) 0.4 (0.3) 0.4 (0.3) 0.09
Vegetables, servings/day 3.4 (2.1) 3.0 (1.8) 3.0 (1.8) 2.8 (1.6) <0.001
Fruits, servings/day 1.4 (1.0) 1.1 (0.9) 1.0 (0.8) 0.8 (0.7) <0.001
Whole grains, servings/day 1.6 (1.2) 1.4 (1.1) 1.2 (1.0) 1.1 (1.0) <0.001
SSBs, servings/day 0.4 (0.7) 0.6 (0.9) 0.8 (1.0) 0.9 (1.1) <0.001
AHEI-2010 score 51.7 (10.4) 45.8 (10.1) 42.3 (9.6) 38.2 (9.2) <0.001
Data are means (SD), unless otherwise specified and are standardised to the age distribution of the study population
Using the conversion factor 1 kcal=4.184 kJ, total calorie intakes across categories of total fried food consumption were 1,743 (529), 1,871 (543), 2,025
(570), and 2,203 (619) kcal/day
SSB, sugar-sweetened beverage; %E, % of total energy
a
Indicates values are energy-adjusted
Diabetologia
In our cohort, women with higher consumption of fried
food consumed more red meat and sugar-sweetened bever-
ages. Although these dietary factors were components of the
AHEI-2010 dietary pattern score, we conducted a sensitivity
analysis further adjusting for red meat and sugar-sweetened
beverages, which slightly attenuated the associations. The
multivariable-adjusted RRs (95% CIs) of GDM among wom-
en who consumed total fried foods 1–3, 4–6and≥7times/
week, compared with those who consumed less than
once/week, were 1.04 (0.89, 1.21), 1.09 (0.90, 1.32) and
1.78 (1.27, 2.51), respectively ( pfor trend= 0.04).
Discussion
In this large prospective cohort study, we found a significant
and positive association between pre-pregnancy fried food
consumption and the risk of GDM. The association was
Tabl e 2 Risk of GDM in association with frequency of pre-pregnancy total fried food consumption
Frequency of total fried food consumption pvalue for trend
<1/week 1–3/week 4–6/week ≥7/week
GDM/pregnancies 355/10,518 304/7,181 156/3,032 32/348
Model 1 1.00 (reference) 1.28 (1.10, 1.49) 1.60 (1.33, 1.92) 3.07 (2.19, 4.30) <0.001
Model 2 1.00 (reference) 1.23 (1.06, 1.43) 1.49 (1.24, 1.79) 2.55 (1.81, 3.59) <0.001
Model 3 1.00 (reference) 1.13 (0.97, 1.32) 1.31 (1.08, 1.59) 2.18 (1.53, 3.09) <0.001
Model 4 1.00 (reference) 1.06 (0.91, 1.23) 1.14 (0.94, 1.38) 1.88 (1.34, 2.64) 0.01
Data are shown as RRs (95% CI)
Model 1: age (months) and parity (0, 1, 2 and ≥3 pregnancies lasting ≥6 months)
Model 2: Model 1 + race/ethnicity (Caucasian, African-American, Hispanic, Asian, others and missing/not reported), family history of diabetes (yes, no),
cigarette smoking (never, past, current and missing/not reported) and physical activity (quartiles)
Model 3: Model 2 + total energy intake (quartiles) and diet quality (i.e. AHEI-2010 score, quartiles)
Model 4: Model 3 + BMI (<21, 21–22.9, 23.0–24.9, 25.0–26.9, 27.0–28.9, 29.0–30.9, 31.0–32.9, 33.0–34.9 and ≥35.0 kg/m
2
)
Tabl e 3 Risk of GDM in association with frequency of pre-pregnancy fried food consumption at home and away from home
Frequency of fried food consumption pvalue for trend
<1/week 1–3/week ≥4/week
At home
GDM/pregnancies 554/15,042 276/5,773 17/264
Model 1 1.00 (reference) 1.19 (1.03, 1.38) 1.55 (0.98, 2.46) 0.004
Model 2 1.00 (reference) 1.18 (1.02, 1.36) 1.37 (0.86, 2.20) 0.01
Model 3 1.00 (reference) 1.14 (0.98, 1.31) 1.30 (0.81, 2.10) 0.054
Model 4 1.00 (reference) 1.07 (0.92, 1.23) 1.37 (0.88, 2.13) 0.18
Away f rom h om e
GDM/pregnancies 475/13,493 341/7,170 31/416
Model 1 1.00 (reference) 1.40 (1.22, 1.61) 2.15 (1.52, 3.05) <0.001
Model 2 1.00 (reference) 1.32 (1.15, 1.52) 1.87 (1.32, 2.66) <0.001
Model 3 1.00 (reference) 1.21 (1.05, 1.40) 1.63 (1.15, 2.33) <0.001
Model 4 1.00 (reference) 1.11 (0.96, 1.28) 1.46 (1.03, 2.07) 0.03
Data are shown as RRs (95% CIs)
Model 1: age (months) and parity (0, 1, 2 and ≥3 pregnancies lasting ≥6 months)
Model 2: Model 1 + race/ethnicity (Caucasian, African-American, Hispanic, Asian, others and missing/not reported), family history of diabetes (yes, no),
cigarette smoking (never, past, current and missing/not reported) and physical activity (quartiles)
Model 3: Model 2 + total energy intake (quartiles) and diet quality (i.e. AHEI-2010 score, quartiles)
Model 4: Model 3 + BMI (<21, 21–22.9, 23.0–24.9, 25.0–26.9, 27.0–28.9, 29.0–30.9, 31.0–32.9, 33.0–34.9 and ≥35.0 kg/m
2
)
Diabetologia
attenuated, but remained significant, after additional adjust-
ment for dietary and non-dietary factors, including BMI. In
addition, the association was not significantly modified by
age, parity, family history of diabetes, physical activity, over-
all diet quality or overweight status. Fried food consumed
away from home may have a stronger association with
GDM risk than fried food consumed at home.
To our knowledge, the association between fried food con-
sumption and risk of GDM has not been previously examined.
There is evidence, although still limited, indicating that frequent
fried food consumption is associated with increased risk of
obesity and type 2 diabetes among non-pregnant individuals.
During a median of 6.1 years of follow-up in the SUN project (a
Mediterranean cohort), participants who consumed fried foods
more than four times/week had a 37% higher risk of developing
overweight/obesity in comparison with those who consumed
fried foods less than twice/week [7]. Similarly, fried food
consumption was also positively associated with general and
central obesity in a cross-sectional analysis in another
Mediterranean cohort [6]. In a prospective cohort study of
African-American women, frequent fried chicken consumption
was associated with risk of incident type 2 diabetes [24].
Recently, an analysis of two large prospective cohorts among
US men and women also reported that frequent fried food
consumption was significantly associated with an increased
risk of incident type 2 diabetes [25].
Frequent fried food consumption could be an indicator of
unhealthy dietary habits. In our study population, we observed
that women who consumed fried foods more frequently had
greater total energy intake and poorer diet quality, which is
consistent with findings from a previous study [26]. However,
the association between fried food consumption and GDM risk
persisted after adjustment for these dietary variables. Moreover,
additional adjustment for specific food groups (i.e. red meat,
processed meat) did not materially alter the observed association.
The potential detrimental effects of fried food consumption
on GDM risk may result from the modification of foods and
frying medium, and generation of harmful by-products during
the frying process. Frying deteriorates oils through the pro-
cesses of oxidation and hydrogenation [5], leading to an
increase in the absorption of oil degradation products by the
foods being fried, and also a loss of unsaturated fatty acids
such as linoleic and linolenic acids and an increase in the
corresponding trans-fatty acids such as trans-linoleic acids
and trans-linolenic acids [27]. Several epidemiological studies
and experimental studies have linked higher intake of trans-
fatty acids to reduced insulin sensitivity and increased risk of
type 2 diabetes [28]. There was suggestive evidence showing
that higher intake of trans-fat may be associated with greater
risk of incident GDM, although the association was no longer
significant after additional adjustment for other types of die-
tary fats [29]. In the present study, additional adjustment of
trans-fat intake did not appreciably alter the association
between fried food consumption and GDM, indicating that
the association cannot be fully explained by trans-fat. Frying
also results in significantly higher levels of dietary AGEs, the
derivatives of glucose–protein or glucose–lipid interactions
[30]. Recently, AGEs have been implicated in insulin resis-
tance, pancreatic beta cell damage and diabetes, partly be-
cause they promote oxidative stress and inflammation [31,
32]. Moreover, intervention studies with a diet low in AGEs
have shown significantly improved insulin sensitivity [33],
reduced oxidant stress and alleviated inflammation [34]. The
above-mentioned deterioration of oils during frying is more
profound when the oils are reused [35], a practice more
common away from home than at home. This may partly
explain why we observed a stronger association of GDM risk
with fried foods consumed away from home than fried foods
consumed at home.
Our study has several strengths, including the prospective
design that establishes the temporal direction of the associa-
tions, the large sample size and the repeated comprehensive
assessment of many lifestyle characteristics that were collect-
ed prospectively with a long duration of follow-up. Therefore,
we were able to assess the confounding effects and potential
modification of the association for fried food consumption by
other dietary components or lifestyle factors. We acknowledge
that there are several limitations. First, we did not have infor-
mation about the specific types (e.g. beef, chicken, fish or
vegetables, etc.), serving size or frying methods (e.g. deep or
pan frying; fresh or reused oil; duration and temperature; types
of oil [i.e. olive, corn, vegetable, etc.] used away from home,
etc.) of fried foods consumed by our participants. Second, our
study population consisted mostly of white American women.
Thus, the ability to directly generalise the observed associa-
tions may be limited to similar populations. However, the
relative homogeneity of the study population reduces potential
confounding owing to unmeasured socioeconomic variability.
In conclusion, we observed that frequent fried food con-
sumption was significantly and positively associated with the
risk of incident GDM in a prospective cohort study. Our study
indicates potential benefits of limiting fried food consumption
in the prevention of GDM in women of reproductive age.
Further studies are warranted to confirm our findings and to
elucidate the underlying mechanisms.
Funding This study was supported by the Intramural Research
Program of the Eunice Kennedy Shriver National Institute of Child
Health and Human Development, National Institutes of Health (contract
No. HHSN275201000020C). The Nurses’Health Study II was funded by
research grants DK58845, CA50385, P30 DK46200 and UM1
CA176726 from the National Institutes of Health. DKT was supported
by a mentored fellowship from the American Diabetes Association (No.
7-12-MN-34).
Duality of interest All authorsdeclare that there is no duality of interest
associated with this manuscript.
Diabetologia
Contribution statement WB contributed to the design and analysis of
the study and wrote the manuscript. DKT interpreted the data, conducted
technique review and reviewed and edited the manuscript. SFO
interpreted the data and reviewed and edited the manuscript. CZ contrib-
uted to the design and analysis of the study, interpreted the data and
reviewed and edited the manuscript. All authors approved the final
version. WB and CZ are the guarantors of this work and, as such, had
full access to all the data in the study and take responsibility for the
integrity of the data and the accuracy of the data analysis.
References
1. Iqbal R, AnandS, OunpuuS et al (2008) Dietary patterns and the risk
of acute myocardial infarction in 52 countries: results of the
INTERHEART study. Circulation 118:1929–1937
2. Hu FB (2008) Globalization of food patterns and cardiovascular
disease risk. Circulation 118:1913–1914
3. Popkin BM (2006) Global nutrition dynamics: the world is shifting
rapidly toward a diet linked with noncommunicable diseases.
Am J Clin Nutr 84:289–298
4. Pan A, Malik VS, Hu FB (2012) Exporting diabetes mellitus to Asia:
the impact of western-style fast food. Circulation 126:163–165
5. Guallar-Castillon P, Rodriguez-Artalejo F, Lopez-Garcia E et al
(2012) Consumption of fried foods and risk of coronary heart disease:
Spanish cohort of the European Prospective Investigation into Cancer
and Nutrition study. BMJ 344:e363
6. Guallar-Castillon P, Rodriguez-Artalejo F, Fornes NS et al (2007)
Intake of fried foods is associated with obesity in the cohort of
Spanish adults from the European Prospective Investigation into
Cancer and Nutrition. Am J Clin Nutr 86:198–205
7. Sayon-Orea C, Bes-Rastrollo M, Basterra-Gortari FJ et al (2013)
Consumption of fried foods and weight gain in a Mediterranean
cohort: the SUN project. Nutr Metab Cardiovasc Dis 23:144–150
8. American Diabetes Association (2004) Gestational diabetes mellitus.
Diabetes Care 27(Suppl 1):S88–S90
9. Metzger BE, Lowe LP, Dyer AR et al (2008) Hyperglycemia and
adverse pregnancy outcomes. N Engl J Med 358:1991–2002
10. Reece EA, Leguizamon G, Wiznitzer A (2009) Gestational diabetes:
the need for a common ground. Lancet 373:1789–1797
11. Bellamy L, Casas JP, Hingorani AD, Williams D (2009) Type 2
diabetes mellitus after gestational diabetes: a systematic review and
meta-analysis. Lancet 373:1773–1779
12. Zhang C, Ning Y (2011) Effect of dietary and lifestyle factors on the
risk of gestational diabetes: review of epidemiologic evidence.
Am J Clin Nutr 94:1975S–1979S
13. Solomon CG, Willett WC, Carey VJ et al (1997) A prospective study
of pregravid determinants of gestational diabetes mellitus. JAMA
278:1078–1083
14. Colditz GA, Hankinson SE (2005) The Nurses' Health Study: life-
style and health among women. Nat Rev Cancer 5:388–396
15. Willett WC, Sampson L, Stampfer MJ et al (1985) Reproducibility
and validity of a semiquantitative food frequency questionnaire.
Am J Epidemiol 122:51–65
16. Willett WC, Sampson L, Browne ML et al (1988) The use of a self-
administered questionnaire to assess diet four years in the past.
Am J Epidemiol 127:188–199
17. Salvini S, Hunter DJ, Sampson L et al (1989) Food-based validation
of a dietary questionnaire: the effects of week-to-week variation in
food consumption. Int J Epidemiol 18:858–867
18. Rimm EB, Stampfer MJ, Colditz GA, Chute CG, Litin LB, Willett
WC (1990) Validity of self-reported waist and hip circumferences in
men and women. Epidemiology 1:466–473
19. Wolf AM, Hunter DJ, Colditz GA et al (1994) Reproducibility
and validity of a self-administered physical activity questionnaire.
Int J Epidemiol 23:991–999
20. Chiuve SE, Fung TT, Rimm EB et al (2012) Alternative dietary
indices both strongly predict risk of chronic disease. J Nutr 142:
1009–1018
21. Wacholder S (1986) Binomial regression in GLIM: estimating risk
ratios and risk differences. Am J Epidemiol 123:174–184
22. Hanley JA, Negassa A, Edwardes MD, Forrester JE (2003) Statistical
analysis of correlated data using generalized estimatingequations: an
orientation. Am J Epidemiol 157:364–375
23. Zou G (2004) A modified poisson regression approach to prospective
studies with binary data. Am J Epidemiol 159:702–706
24. Krishnan S, Coogan PF, Boggs DA, Rosenberg L, Palmer JR (2010)
Consumption of restaurant foods and incidence of type 2 diabetes in
African American women. Am J Clin Nutr 91:465–471
25. Cahill LE, Pan A, Chiuve SE et al (2014) Fried-food consumption
and risk of type 2 diabetes and coronary artery disease: a prospective
study in 2 cohorts of US women and men. Am J Clin Nutr 100:667–
675
26. Taveras EM, Berkey CS, Rifas-Shiman SL et al (2005) Association
of consumption of fried food away from home with body mass index
and diet quality in older children and adolescents. Pediatrics 116:
e518–e524
27. Li A, Ha Y, Wang F, Li W, Li Q (2012) Determination of thermally
induced trans-fatty acids in soybean oil by attenuated total reflectance
Fourier transform infrared spectroscopy and gas chromatography
analysis. J Agric Food Chem 60:10709–10713
28. Mozaffarian D, Aro A, Willett WC (2009) Health effects of trans-
fatty acids: experimental and observational evidence. Eur J Clin Nutr
63(Suppl 2):S5–S21
29. Bowers K, Tobias DK, Yeung E, Hu FB, Zhang C (2012) A prospec-
tive study of prepregnancy dietary fat intake and risk of gestational
diabetes. Am J Clin Nutr 95:446–453
30. Goldberg T, Cai W, Peppa M et al (2004) Advanced glycoxidation
end products in commonly consumed foods. J Am Diet Assoc 104:
1287–1291
31. Cai W, Ramdas M, Zhu L, Chen X, Striker GE, Vlassara H (2012)
Oral advanced glycation endproducts (AGEs) promote insulin resis-
tance and diabetes by depleting the antioxidant defenses AGE
receptor-1 and sirtuin 1. Proc Natl Acad Sci U S A 109:15888–15893
32. Coughlan MT, Yap FY, Tong DC et al (2011) Advanced glycation
end products are direct modulators of beta-cell function. Diabetes 60:
2523–2532
33. Mark AB, Poulsen MW, Andersen S et al (2014) Consumption of a
diet low in advanced glycation endproducts for 4 weeks improves
insulin sensitivity in overweight women. Diabetes Care 37:88–95
34. Vlassara H, Cai W, Goodman S et al (2009) Protection against
loss of innate defenses in adulthood by low advanced glycation
end products (AGE) intake: role of the antiinflammatory AGE
receptor-1. J Clin Endocrinol Metab 94:4483–4491
35. Lee WT, Dawson LE (1973) Chicken lipid changes during cooking in
fresh and reused cooking oil. J Food Sci 38:1232–1237
Diabetologia
