Diet quality and blood pressure among pregnant women with overweight or obesity: A secondary analysis of two randomized controlled trials

Abstract

Introduction

Maternal obesity is a significant risk factor for hypertensive disorders of pregnancy. High diet quality may protect against this, yet data regarding the relationship between diet quality and blood pressure among pregnant women with raised body mass index (BMI) is limited.

Material and methods

This is a secondary analysis (n = 543) of women with BMI ≥25 kg/m² from two randomized controlled trials; PEARS (Pregnancy Exercise and nutrition Research Study with smartphone application support) and ROLO (Randomized cOntrol trial of LOw glycemic index diet to prevent macrosomia in euglycemic women). Blood pressure was measured at 10–18 weeks and 28 weeks of pregnancy. Mean arterial pressure was calculated as (diastolic blood pressure $+\frac{1}{3}\times$ [systolic blood pressure $-$ diastolic blood pressure]). Diet quality was assessed using 3‐day food diaries, and Alternative Healthy Eating Index for Pregnancy (AHEI‐P) scores were generated, quantifying alignment of food intakes with dietary guidelines in first and early third trimesters. The cohort was divided based on AHEI‐P tertiles to explore differences at an alpha significance value of <0.05.

Results

The mean age of the group was 32.21 ± 4.39 years with a median body mass index (BMI) of 28.13 (IQR 3.47) kg/m². Mean arterial pressures in the first and third trimesters were 81.07 ± 9.00 mmHg and 82.33 ± 7.53 mmHg, respectively. Rates of elevated blood pressure (≥120/80 mmHg) were 22.33% in trimester 1 and 24.48% in early trimester 3. Mean AHEI‐P scores in trimester 1 and early trimester 3 were 53.90 ± 10.43 and 54.05 ± 10.76, respectively. There was no correlation between AHEI‐P score and blood pressure and no differences in blood pressure between AHEI‐P tertiles at either timepoint (all P‐values <0.05). A higher proportion of those with elevated early third trimester blood pressure had a BMI of ≥30 kg/m² compared with those with normal blood pressure (40.31% vs 28.64%, P = 0.016).

Conclusions

While diet remains an important factor in maternal health and wellbeing, we did not find a relationship between diet quality as measured by AHEI‐P and blood pressure among pregnant women with BMI ≥25 kg/m². High BMI remains a risk factor for hypertensive disorders of pregnancy.

Among women with overweight and obesity, diet quality as quantified by AHEI‐P score had no association with blood pressure in either the first or early third trimesters.

Abbreviations

AHEI‐P

Alternative Healthy Eating Index for Pregnancy

BMI

body mass index

BP

blood pressure

DASH

Dietary Approaches to Stop Hypertension diet

DBP

diastolic blood pressure

HDP

hypertensive disorders of pregnancy

MAP

mean arterial pressure

PEARS

Pregnancy Exercise and nutrition Research Study

ROLO

Randomized cOntrol trial of LOw glycemic index diet

SBP

systolic blood pressure

Key message.

Among pregnant women with overweight or obesity, we did not find a relationship between Alternative Healthy Eating Index for Pregnancy and blood pressure in the first or early third trimesters. High body mass index remains a risk factor for elevated blood pressure in pregnancy.

1. INTRODUCTION

Hypertension is one of the leading causes of maternal and fetal morbidity and mortality worldwide. ¹ Hypertensive disorders of pregnancy (HDP) increase the risk of serious maternal complications such as stroke, intracerebral hemorrhage and kidney disease, while also contributing to an increased risk of intrauterine growth restriction and stillbirth. ² , ³

Pregnancy may be considered as a “stress test” for later life, as hypertension during pregnancy confers higher risk of cardiometabolic disease. ⁴ , ⁵ Dunietz et al. found that even those with moderately elevated blood pressure (BP) during pregnancy (two readings of either systolic blood pressure (SBP) ≥120 mmHg or diastolic BP ≥80 mmHg) had an increased risk of developing hypertension after 7 to 10 years, and longitudinal analysis of the Randomized Control trial of Low Glycemic Index Diet to Prevent Macrosomia in Euglycemic Women trial (ROLO trial) demonstrated that those with moderately elevated BP at 28 and 34 weeks’ gestation were more likely to be diagnosed with Stage 1 or 2 hypertension at a 5‐year follow‐up. ⁶ , ⁷ , ⁸ It is therefore of great clinical relevance to evaluate lifestyle factors which may prevent elevation of BP during pregnancy.

High body mass index (BMI) is one of the most significant risk factors for the development of hypertension. ⁹ With a rising prevalence of overweight and obesity in the pregnant population (usually defined by BMI 25–30 kg/m² and ≥30 kg/m², respectively), obesity in pregnancy has become a critical focus of current research. ¹⁰ A systematic review of international observational studies found that the majority of pregnant women do not adhere to the dietary guidelines for pregnancy. ¹¹ Pregnancy is a unique time in which women are in regular contact with healthcare professionals, with increased motivation to make lifestyle changes. ¹² , ¹³ This highlights the great potential to support women to make positive dietary changes that may reduce risks associated with obesity in pregnancy, including hypertension.

Intake of specific dietary components including sodium, polyunsaturated fatty acids, fruits, vegetables, sugar and calcium have been found to impact maternal BP when assessed in isolation. ¹⁴ , ¹⁵ Current research, however, suggests there may be synergistic effects of dietary components when consumed together. ¹⁶ Overall diet quality can be assessed by quantifying adherence to healthy dietary patterns such as the Mediterranean Diet or the Dietary Approaches to Stop Hypertension (DASH) diet in non‐pregnant populations. ¹⁷ Adherence to the DASH diet is associated with lower maternal BP in pregnancy among healthy women without HDP. ¹⁸ The Alternative Healthy Eating Index for Pregnancy (AHEI‐P) is a tool which has been adapted for pregnancy from the Alternate Healthy Eating Index as a metric quantifying adherence to a diet with proven health benefits across nine categories; vegetables, fruit, fiber, trans fats, white meat to red meat ratio, polyunsaturated to saturated fat ratio, calcium, folate and iron. ¹⁹ Although the AHEI‐P has been included in research regarding diet quality and conception, maternal glucose levels and behavior and cognition of offspring, it has yet to be used in research related to BP during pregnancy. ²⁰ , ²¹ Compared with DASH score, AHEI‐P scoring is better directed towards current health recommendations for pregnancy rather than the specific effect of foods on BP.

The primary aim of this study was to investigate the relationship between BP and overall diet quality as measured by the AHEI‐P at 10–18 and 28 weeks’ gestation in women with BMI > 25 kg/m². A secondary aim was to evaluate and compare the dietary composition, energy intake and BMI of the subgroup of women who had elevated BP during pregnancy with those with normal BP.

2. MATERIAL AND METHODS

We conducted a secondary analysis of data from two randomized controlled trials (RCTs): the Pregnancy Exercise and Nutrition Research Study with smartphone application support (PEARS) and ROLO, conducted. Both trials were conducted in the National Maternity Hospital, Dublin, Ireland, a tertiary level maternity hospital. The results of PEARS and ROLO RCTs were published in 2018 and 2012, respectively. ²² , ²³

In PEARS, pregnant women between the ages of 18 and 45 years with BMI ≥25 kg/m² at their first antenatal visit were invited to participate. Exclusion criteria included lack of a smartphone, inability to give informed consent, multiple pregnancy, medical disorders requiring treatment, previous poor obstetric outcome or gestational diabetes mellitus. Those in the intervention arm received a single education session at the randomization visit involving encouragement to choose foods with low glycemic index and information about healthy carbohydrate portions, delivered by a research dietitian or nutritionist and education regarding exercise aligning with guidelines from the American College of Obstetricians and Gynecologists, delivered by an obstetrician. This information was reinforced via a mobile phone application, emails and follow‐up visits at 28 and 34 weeks’ gestation. The primary outcome was decreased incidence of gestational diabetes mellitus in the intervention group. Whereas no differences were found in the primary outcome, the intervention group had both reduced dietary glycemic index and gestational weight gain.

In ROLO, participants aged over 18 in their second pregnancy with previous delivery of an infant with macrosomia were included. Exclusion criteria included a diagnosis of diabetes mellitus or gestational diabetes mellitus, regular medication, inability to give informed consent or multiple pregnancy. BMI was not part of the inclusion/exclusion criteria. The intervention arm consisted of face‐to‐face education session at a mean gestational age of 15.7 (SD 3.0) weeks with a registered dietitian advising on healthy eating guidelines for pregnancy and exchange of high for low glycemic index carbohydrates. The primary outcome was a difference in birthweight between control and intervention groups. Results showed that a low glycemic index diet had no significant effect on birthweight. Women in the intervention group had significantly lower energy intake and mean glycemic load with higher fiber intake.

For this analysis, participants in both studies with a BMI of ≥25 kg/m² at the first antenatal visit were selected. A subsection of the ROLO cohort (n = 431) and all PEARS participants were initially eligible for inclusion (n = 493) based on BMI. Women were excluded from analysis if they had not completed a 3‐day food diary in either the first or third trimester (n = 218) or if they had no first or third trimester BP measurements recorded (n = 381). The result was 543 women with a BMI of ≥25 kg/m² with at least one food diary and one BP measurement available for analysis. Of these, 368 (67.77%) were participants in the PEARS study and 175 (32.22%) in ROLO.

All participants attended the National Maternity Hospital, Dublin, Ireland, from 2007 to 2016. Age, BMI, education level and ethnicity were recorded at the first antenatal visit. In both ROLO and PEARS, educational attainment was self‐reported as either complete third level or less than complete third level. Participant addresses obtained from medical files were used to determine the Pobal Haase and Pratschke deprivation index (HP index) score, a marker of social deprivation based on three dimensions of affluence and disadvantage. ²⁴ The index scores indicate how a specific address compares with other areas at that time point using latest census data.

Physical activity in both studies was assessed with questionnaires previously validated in pregnancy adapted from the SLÁN 2002 survey. ²⁵ For statistical analysis, the average metabolic equivalent of task (MET) minutes per week were calculated for each participant.

Body weight and BP were obtained at the first antenatal visit (10–18 weeks’ gestation) and again at 28 weeks’ gestation. BMI was calculated as kg/m² from height and weight measured using standardized protocols and instruments.

Primary outcome variables for this secondary analysis were SBP, diastolic blood pressure (DBP) and mean arterial pressure (MAP). MAP was calculated as [DBP $+\frac{1}{3}\times$ (SBP − DBP)]. ²⁶ Auscultatory BP measurements (mmHg) were obtained by doctors or midwives, using a mobile Trimline aneroid sphygmomanometer (Hillrom, Chicago, IL, USA). Sitting with their arm at the level of the heart, after resting for 3–5 minutes, the observer measured BP in the right arm using an appropriately sized cuff according to mid‐arm circumference. BP recordings of <120/80 mmHg were regarded as “normal” as per the American College of Cardiology Guidelines, and two readings of either SBP ≥120 mmHg or diastolic BP ≥80 mmHg were considered elevated. ²⁷

Participants completed 3‐day food diaries in trimesters one and three, recording intake as accurately as possible for two weekdays and one weekend day. Food diaries in ROLO were analyzed using NETWISP, version 3.0 (Tinuviel software, Llanfechell, Anglesey, UK), and Nutritics NUTRITION ANALYSIS Software, version 5.53 (Nutritics, Dublin, Ireland) was used in PEARS. Mean daily intake was calculated for sodium, white and red meat, fats, folate, iron, calcium, fruit, vegetables and sugar. Mean energy intakes of <500 or >3500 kcal/day were excluded from the analysis as they were likely a result of misreporting or measurement errors (n = 1).

Diet quality was measured using the AHEI‐P, adapted from the AHEI. ²⁸ The AHEI‐P differs from the AHEI in that alcohol, nuts and soy protein are excluded and folate, calcium and iron intake are included as important dietary nutrients for pregnancy. Scores from 0 to 10 were assigned according to Table 1, with a participant scoring 10 points for that component if they met the recommended daily intake. For every 10% decrease in intake for a component, 1 point is subtracted. ²⁹ Summation of component scores resulted in the overall AHEI‐P score, ranging from 0 (worst) to 90 (best).

TABLE 1.

Criteria for Alternative Healthy Eating Index for Pregnancy scores.

	Minimum score (0)	Maximum score (10)
Vegetables (servings/day) a	0	≥5
Fruit (servings/day) b	0	≥5
Fiber (g/day)	0	≥25
Trans fat (% of energy)	≥4	≤0.5
White:red meat ratio	0	≥4
PUFA:SFA ratio	≤0.1	≥1
Calcium (mg/day)	0	≥1200
Folate (μg/day)	0	≥600
Iron (mg/day)	0	≥27

Abbreviations: PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids.

^a 1 serving = 0.5 cup of vegetables (1 cup = 236.59 g).

^b 1 serving = 1 medium piece of fruit or 0.5 cup equivalent (0.5 cup = 118.30 g).

2.1. Statistical analyses

Data were assessed for normality using the Kolmogorov–Smirnov tests and through visual inspection of histograms.

Participants were stratified based on AHEI‐P score tertiles first at trimester one and then at trimester three to facilitate subgroup analysis. One‐way analysis of variance (ANOVA) was used to assess for differences in demographic characteristics between tertile groups. Mean SBP, DBP and MAP between AHEI‐P tertile groups were compared using ANOVA and independent‐samples t‐tests.

Correlations were used to assess relationships between the continuous variables SBP, DBP and MAP and early and late AHEI‐P scores. Potential confounders determined a priori based on previous literature were included in correlation analysis including ethnicity, smoking status, BMI and activity in METs.

Multiple regression analysis was performed to assess the relationship between diet quality and BP at trimester one and three. Both adjusted and unadjusted models of regression were carried out.

The cohort was then divided into two groups based on elevated (≥120/80 mmHg) or normal BP readings. Independent t‐tests were used to determine differences in AHEI‐P score and dietary intake of sodium; white to red meat ratio; polyunsaturated to saturated fatty acid ratio; fruit; vegetables; sugar; calcium; iron; and folate between those with normal and elevated BP. These dietary components were selected for analysis given associations between these food groups and BP in pregnancy in the current literature. ¹² , ¹³

Bonferroni correction was applied across P‐values to give an adjusted ɑ significance of P = 0.00089. Nominal P‐values were used in reporting of results. A calculation to clarify what effect sizes we are powered to detect was performed, showing that 543 subjects confer 80% power to detect true standardized regression slopes of ± 0.120 at a Type I error probability associated with this test of the null hypothesis that this slope equals zero, is 0.05.

3. RESULTS

A total of 543 women from the ROLO and PEARS studies were included in analysis. Table 2 summarizes the baseline characteristics and demographics of the population. The sample comprised healthy pregnant women with a mean age of 32.2 ± 4.4 years, mostly white (94.5%), with the majority (60.1%) having completed third level education. The median (interquartile range [IQR]) BMI was 28.13 (4.50) kg/m². Mean (standard deviation) BPs in the first trimester were SBP 111.20 ± 10.28 mmHg, DBP 66.77 ± 7.19 mmHg and MAP 81.07 ± 9.00 mmHg. In the early third trimester, mean BPs were SBP 112.92 ± 10.33 mmHg, DBP 67.05 ± 6.89 mmHg and MAP 82.34 ± 7.22 mmHg. Incidence of elevated BP (≥120/80 mmHg) in the first and early third trimesters was 22.3% (n = 113) and 24.5% (n = 129), respectively. There was no significant difference between SBP, DBP or MAP from first to early third trimesters (P = 0.005, 0.480, 0.004). The incidence of elevated BP was not statistically significant from first to early third trimester (24.48% vs 22.33%, P = 0.001).

TABLE 2.

Maternal characteristics of the ROLO and PEARS participants with BMI ≥25 kg/m², blood pressure measurements and returned, completed 3‐day food diaries (n = 543).

Characteristic	Overall
	n	Mean or n	SD, IQR or %
Age (year)a	542	32.21	4.39
Parityb	543
Nulliparous		196	36.1%
Multiparous		347	63.9%
BMIc	543	28.13	3.47
BMI categoryb	543
Overweight (25–29.9 kg/m2)		374	68.9%
Obese Class I (30–34.9 kg/m2)		126	23.2%
Obese Class II (35–39.9 kg/m2)		42	7.7%
Obese Class III (≥40 kg/m2)		1	0.2%
Educational attainmentb	537
Below third level		214	39.9%
Completed third level		323	60.1%
Ethnicity	532
White Irish		430	80.8%
Other white		73	13.7%
African/any other black background		5	0.9%
Chinese		9	1.7%
Any other Asian		8	1.5%
Mixed		6	1.1%
Pobal HP index	543	5.50	11.28
Physical activity
T1 METs	275	632.49	417.81
T3 METs	400	533.61	347.01
Energy intake
T1 intake	516	1860.92	427.53
T3 intake	442	1828.83	444.14
AHEIPc
T1 AHEIP	515	54.15	14.00
T3 AHEIP	440	54.00	14.00

Note: Data are presented as ^amean (SD), ^b n (%), ^cmedian (interquartile range).

Abbreviations: AHEIP, Alternative Healthy Eating Index for Pregnancy; BMI, body mass index classification per World Health Organization cut‐offs; energy intakes, mean kcal intakes as reported in 3‐day food diaries; IQR, interquartile range; MET, metabolic equivalent; n, number of participants; PEARS, Pregnancy Exercise and nutrition Research Study; Pobal HP Index, Pobal Haase‐Pratschke Deprivation Index; ROLO, Randomized cOntrol trial of LOw glycemic index diet; T1, trimester one; T3, trimester three.

The median AHEI‐P score in trimesters 1 and early trimester 3 was 54.15 (14.00) and 54.00 (14.00), respectively. The overall AHEI‐P scores represented as tertiles were in the range: tertile 1 (0–49.00); tertile 2 (49.00 to <58.00); tertile 3 (>58.00), for trimester one, and tertile 1 (0 to <48.66); tertile 2 (48.66–<58.08); tertile 3 (>58.08) for trimester three.

The incidence of pregnancy‐induced hypertension and preeclampsia was 6.1% and 1.4%, respectively. In all, 16.7% of infants were born large‐for‐gestational age and 4.6% small‐for‐gestational age, with an overall mean birthweight of 3777.25 ± 555.42 g. The median gestational age at delivery was 40.43 (1.71) weeks, with an incidence of preterm delivery (<37 weeks’ gestation) of 3.0% (Table 3).

TABLE 3.

Pregnancy outcomes of the ROLO and PEARS participants with body mass index ≥25 kg/m², blood pressure measurements and returned, completed 3‐day food diaries and delivery information recorded (n = 538).

Characteristic	Overall
	n	Mean, median or n	SD, IQR or %
Pregnancy‐induced hypertensionb	510	31	6.1%
Preeclampsiab	497	7	1.4%
Birthweighta	538	3777.25	555.42
Large for gestational ageb	503	84	16.7%
Small for gestational ageb	503	23	4.6%
Gestational age at deliveryc	535	40.43	1.71
Delivery <37 weeksb	535	16	3.0%

Note: Data are presented as ^amean (SD), ^b n (%), ^cmedian (interquartile range).

Abbreviations: IQR, interquartile range; PEARS, Pregnancy Exercise and nutrition Research Study; Pobal HP Index, Pobal Haase‐Pratschke Deprivation Index; ROLO, Randomized cOntrol trial of LOw glycemic index diet.

Table S1 examines population demographics with regard to the trimester 1 AHEI‐P score. Mean age was lower among those in the lowest AHEI‐P tertile and higher among those in the highest tertile (30.98 ± 5.11 vs 32.36 ± 3.81 vs 33.13 ± 3.88, P < 0.001). Those in the lowest AHEI‐P tertile had a lower average daily energy intake than those in tertiles 2 and 3 (1728.57 ± 379.86 kcal vs 1807.66 ± 386.25 kcal vs 2025.14 ± 450.26 kcal, P < 0.001). Daily activity levels as quantified by metabolic equivalent did not differ between tertiles. A higher proportion of those with higher diet quality had completed third level education (71.67% vs 58.49% vs 49.69%, P < 0.001) but there was no significant difference in Pobal HP Deprivation Index between tertiles (P = 0.004).

Table 4 outlines the average SBP, DBP and MAP for each AHEI‐P tertile group at both trimester one and early trimester three timepoints. In trimester one, those in the highest tertile for diet quality trended towards a lower SBP (109.75 ± 9.7 mmHg vs 111.97 ± 10.77 mmHg vs 112.11 ± 10.32 mmHg, P = 0.065). There was no significant difference in average MAP, SBP or DBP when AHEI‐P tertiles were compared at trimester three (P = 0.574, 0.771 and 0.379, respectively).

TABLE 4.

Mean blood pressure (BP) as measured in trimesters one and three by AHEI‐P score tertile as calculated in trimesters one and three, respectively.

BPa	Overall			Tertile 1			Tertile 2			Tertile 3			P‐value
	n	Mean	SD	n	Mean	SD	n	Mean	SD	n	Mean	SD
Trimester 1
T1 systolic BP (mmHg)	481	111.20	10.28	153	112.11	10.32	151	111.97	10.77	177	109.75	9.7	0.065
T1 diastolic BP (mmHg)	481	66.77	7.19	153	67.43	7.14	152	67.00	7.57	176	65.99	6.85	0.126
T1 MAP (mmHg)	484	81.07	9.00	155	81.26	10.54	152	81.75	8.54	177	80.33	7.84	0.078
T3 systolic BP (mmHg)	501	112.92	10.33	160	113.50	9.98	157	113.59	11.18	184	111.83	9.82	0.196
T3 diastolic BP (mmHg)	501	67.05	6.89	160	67.82	6.46	157	67.13	7.13	184	66.31	7.01	0.056
T3 MAP (mmHg)	501	82.34	7.22	160	83.05	6.94	157	82.62	7.71	184	81.48	6.96	0.089
Trimester 3
T3 systolic BP (mmHg)	427	112.96	10.72	128	112.18	10.57	151	113.39	10.95	148	113.21	10.63	0.771
T3 diastolic BP (mmHg)		67.02	7.16		66.25	7.13		67.55	6.42		67.15	7.86	0.379
T3 MAP (mmHg)		82.33	7.53		81.56	7.53		82.83	7.12		82.50	7.89	0.574

Note: Data are presented as mean (SD).

Abbreviations: AHEIP, Alternative Healthy Eating Index for Pregnancy; MAP, mean arterial pressure; T1, trimester one; T3, trimester three.

Table 5 demonstrates the intake of several dietary components during the first and third trimesters across those with normal and elevated (≥120/80 mmHg) BP. There was no significant difference in intake of sodium, red meat, fruit, vegetables, calcium, folate and iron between groups (all P > 0.001). Intake of sugar in the third trimester appeared to be higher among those with elevated BP at a significance level of P < 0.05 but failed to reach significance when adjustments were made for multiple testing (93.33 ± 44.26 vs 84.36 ± 33.19, P = 0.038). Those with elevated BP did not have a higher average daily energy intake in either the first or third trimesters compared with those with normal BP (1828.43 ± 390.70 vs 1866.61 ± 431.86, P = 0.416; 1856.59 ± 508.36 vs 1819.08 ± 425.61, P = 0.452).

TABLE 5.

Average dietary component intake of those with normal blood pressure (BP) compared with those with elevated BP.

	Overall			Normal BP			Elevated BP			P‐value
	n	Mean or median	SD or IQR	n	Mean or median	SD or IQR	n	Mean or median	SD or IQR
Trimester 1
Sodium (mg)a	434	2463.57	788.64	337	2470.19	804.27	97	2440.60	735.25	0.745
White:red meat ratiob	352	1.02	1.42	271	1.98	2.80	81	2.19	3.87	0.591
PUFA:SFA ratiob	434	1.22	0.59	337	1.44	0.96	97	1.40	0.95	0.743
Fruit (servings)a	383	2.17	1.38	301	2.20	1.40	82	2.04	1.29	0.338
Vegetables (servings)a	357	2.11	1.60	278	2.16	1.68	79	1.90	1.27	0.204
Sugar (g)a	434	89.58	33.33	337	89.33	34.05	97	90.44	30.81	0.773
Calcium (mg)a	482	900.98	301.17	378	906.47	305.24	104	881.04	286.45	0.446
Folate (μg)a	482	226.73	73.59	378	225.83	70.49	104	229.99	84.19	0.611
Iron (μg)a	482	11.16	3.61	378	11.22	3.84	104	10.95	2.61	0.496
Energy intake (kcal)a	482	1858.37	423.21	378	1866.61	431.86	104	1828.43	390.70	0.416
Trimester 3
Sodium (mg)a	369	2455.69	782.56	272	2469.30	757.85	97	2417.55	851.03	0.577
White:red meat ratiob	314	1.02	1.31	235	1.90	3.00	79	1.41	1.62	0.163
PUFA:SFA ratiob	369	1.33	0.69	272	1.63	1.23	97	1.45	0.72	0.195
Fruit (servings)a	312	2.04	1.32	230	2.06	1.27	82	1.96	1.44	0.563
Vegetables (servings)a	305	1.98	1.52	221	2.08	1.57	84	1.71	1.35	0.059
Sugar (g)a	369	86.72	36.58	272	84.36	33.19	97	93.33	44.26	0.038
Calcium (mg)a	429	936.26	329.46	321	945.68	330.32	108	908.25	326.80	0.308
Folate (μg)a	429	233.98	85.87	321	233.22	83.66	108	236.24	92.48	0.753
Iron (μg)a	429	11.34	3.43	321	11.35	3.48	108	11.33	3.30	0.977
Energy intake (kcal)a	429	1828.52	447.55	321	1819.08	425.61	108	1856.59	508.36	0.452

Note: Data are presented as ^amean (SD), ^bmedian (interquartile range). Elevated blood pressure refers to measurements ≥120/80 mmHg.

Abbreviations: IQR, interquartile range; kcal, kilocalorie; PUFA:SFA ratio, ratio of polyunsaturated fats to saturated fats; SD, standard deviation; white:red meat ratio, ratio of white to red meat intake.

Multiple regression analysis demonstrated an increase of 0.642 mmHg in early third trimester SBP and of 0.453 mmHg in MAP with every 1‐point increase in BMI >25 kg/m² (P = 0.0002, P = 00017).

4. DISCUSSION

This study is the first to examine the relationship between diet quality using AHEI‐P and BP in a cohort of healthy pregnant women with overweight and obesity. We observed no significant association between diet quality and BP in pregnancy at either the first or early third trimesters. Additionally, analysis of individual dietary components revealed no significant difference across intakes of sodium, sugar, calcium, folate, iron, red meats, polyunsaturated fatty acids, fruit or vegetables when those with normal and elevated BP were compared. Elevation of BMI >25 kg/m² is associated with an elevation of SBP and MAP at 28 weeks’ gestation.

Current data on the relationship between maternal diet quality as assessed by the AHEI‐P and BP during pregnancy is limited. Rifas‐Shiman et al. found that among 1777 healthy pregnant women, higher second trimester AHEI‐P score was associated with a lower risk of developing preeclampsia (odds ratio 0.87, 95% confidence interval [CI] 0.76–1). ²⁸ However, overall trends in mean BPs and BPs below diagnostic level for HDP were not examined. Rifas‐Shiman et al. evaluated dietary quality in a cohort made up primarily of women with a normal BMI and high diet quality (mean [SD] first trimester AHEI‐P score 61.00 ± 10.00). Our cohort, with an elevated BMI, have a different baseline risk profile for development of HDP when compared with this group, precluding direct comparison of findings.

Although the difference in sugar intake between those with elevated and normal BP failed to reach significance at an adjusted significance level of P < 0.00089, existing literature suggests an association between high intake of added sugars and HDP. A Norwegian study involving 32 933 healthy nulliparous women revealed that intake of added sugar was higher in those who developed preeclampsia, and intake of foods high in natural sugars, such as fruit, were associated with a decreased risk of preeclampsia. ³⁰ Sugar is thought to modify endothelial, inflammatory and vascular responses, leading to preeclampsia. This may be via direct effect of hyperglycemia with or without insulin resistance on vascular endothelial cells and by contributing to the development of metabolic syndrome. A working hypothesis for the differential effects of naturally occurring vs added sugars posits that foods with naturally occurring sugars are also higher in dietary fiber, which is associated with reduced risk of preeclampsia. ³¹ , ³²

In our study, self‐reported sodium intake of those with elevated BP did not differ significantly from those with normal BP. This is despite the well established link between excess sodium intake and hypertension in the non‐pregnant population through its effect on the adrenergic and RAAS systems. ³³ Recent studies suggest that an individual's salt sensitivity, rather than quantitative intake, may play a role in the pathogenesis of preeclampsia. ³⁴ As such, adhering to daily sodium recommendations of <2 g/day in pregnancy may be more important in those with a diagnosis of preeclampsia. ³⁵ Further research involving a cohort with higher incidence of HDP may help to clarify the risk of excess sodium intake on development of elevated BP in pregnancy.

Intake of fruits, vegetables and healthy oils as part of the Mediterranean diet has been associated with lower BP and lower incidence of HDP. ³⁶ Among 3187 pregnant women with normal BMI, low adherence to a Mediterranean diet was associated with higher SBP in early pregnancy (118.1 mmHg ± 12.0 vs 116.2 mmHg ± 11.4, P < 0.01). ³⁷ Although there was no difference between those with elevated and those with normal BP as regards fruit and vegetable consumption in our cohort, this may be explained by the fact that neither group met the AHEI‐P cut‐off intake of at least five daily portions of fruit and vegetables. In our cohort overall, fruit consumption was 2.17 ± 1.38 portions in early and 2.19 ± 1.39 portions in late pregnancy, and vegetable intake was 2.11 ± 1.60 and 2.10 ± 1.59 portions, respectively.

The observed relationship between early‐pregnancy BMI and early third trimester BP supports the existing body of evidence documenting the role of BMI in development of HDP. ³⁸ The mechanism by which adiposity may contribute to hypertension in late pregnancy is yet to be fully understood. Current hypotheses focus on the release of proinflammatory cytokines and complement factors from adipose tissue causing systemic inflammation and endothelial dysfunction. Lipid accumulation in the placenta may also affect placental development promoting preeclampsia. ³⁹ Despite the association between pre‐pregnancy BMI and third trimester BP, latest guidelines from international bodies discourage intentional weight loss during pregnancy. ⁴⁰ , ⁴¹ , ⁴² Although associated with decreased rate of cesarean section and large‐for‐gestational age pregnancies, gestational weight loss is associated with small‐for‐gestational age pregnancies, a key predictor of neonatal morbidity and mortality. ⁴³ The International Federation of Gynaecology and Obstetrics offers recommendations in the care of women with overweight and obesity prior, during and post pregnancy, with emphasis on optimization of weight and nutrition in the preconceptual period. ⁴⁴ During pregnancy, gestational weight gain may exert some influence on risk of HDP; however, we did not assess this in our study.

The use of 3‐day food diaries is a well validated method of assessing average dietary intake, allowing better quantification of portion sizes compared with a food frequency questionnaire and eliminating recall bias with real‐time completion. The use of the AHEI‐P score also adds value to this study. It is specifically tailored to capture key areas of maternal nutrition and facilitates statistical analysis with a discrete numeric variable describing diet quality.

The structure of this analysis with a cohort of participants from two separate trials may act as a limitation of this study. However, the demographic homogeneity of the cohort was optimal given that both trials were conducted by the same research group, at the same location, within 10 years. Exclusion and inclusion criteria of the trials differed, as they were directed towards the individual trial outcomes. No participants in our analysis had a history of hypertension or HDP, limiting selection bias.

The absence of data on BP in late third trimester is a limitation of this study. As BP was not a primary outcome of either ROLO or PEARS RCTs, BP measurements beyond 28 weeks regrettably were not available. It is possible that a relationship may exist between overall dietary quality and BP in late pregnancy. A specifically designed prospective cohort study may better capture this information and could be considered in the future.

Sample size limits the ability to evaluate the effect of diet quality on overall incidence of HDP due to the relatively low incidence of these diseases in our cohort. However, this study offers valuable insights into a group of pregnant women often excluded from research due to BMI.

The results of our study are relevant for doctors, dietitians and midwives involved in care for women with overweight and obesity. Although diet quality may not have a defined impact on BP during pregnancy, the health consequences of the rising incidence of overweight and obesity among the pregnant population in Ireland and worldwide cannot be understated. Preconception counseling for those with BMI ≥25 kg/m² should emphasize the effect of obesity on BP along with other maternal and fetal health outcomes and encourage healthy weight loss. During pregnancy, women with BMI ≥25 kg/m² should be supported to follow dietary recommendations with specific focus on intake of added sugars. On a national and global scale, policy makers should prioritize allocation of resources towards nutritionist staffing to support a population with increasing BMI at primary and tertiary levels. Obesity in pregnancy should also remain a focus of clinical research. Building on our results, a study examining dietary quality and potential association between BP in late third trimester and pregnancy outcomes would be important given the known trend in BP and onset of HDP in later third trimester. ¹

5. CONCLUSION

Within a cohort of women with overweight and obesity, our findings suggest that diet quality during pregnancy, as assessed by the AHEI‐P, may not have a significant association with BP in first or early third trimesters. Pre‐pregnancy BMI is associated with incremental elevation of BP during pregnancy. As such, support in preconceptual weight loss at the level of primary care may represent a critical opportunity to reduce the risk of HDP associated with overweight and obesity.

AUTHOR CONTRIBUTIONS

Sorcha Lynch wrote the paper and carried out statistical analysis. Sarah Louise Killeen helped to edit the paper. Eileen O'Brien was involved in study design. Kate Mullane, Emma Hokey and Grace Mealy were involved in data collection. Fionnuala M. McAuliffe designed the study.

FUNDING INFORMATION

The PEARS Study was supported by the University College Dublin, National Maternity Hospital Medical Fund. The ROLO Study was supported by the Health Research Board Ireland, the Health Research Centre for Health and Diet Research, The National Maternity Hospital Medical Fund and the European Union's Seventh Framework Programme (FP7/2007‐2013), project EarlyNutrition under grant agreement no. 289346.

CONFLICT OF INTEREST STATEMENT

The authors have stated explicitly that there are no conflicts of interest in connection with this article.

ETHICS STATEMENT

Both trials had written maternal consent and received institutional ethical approval from the National Maternity Hospital on December 20, 2006 for ROLO ²³ and October 22, 2012 for PEARS. ²²

Supporting information

Table S1.

Lynch S, Killeen SL, O’Brien E, et al. Diet quality and blood pressure among pregnant women with overweight or obesity: A secondary analysis of two randomized controlled trials. Acta Obstet Gynecol Scand. 2024;103:1073‐1082. doi: 10.1111/aogs.14821