Effect of Prenatal Docosahexaenoic Acid Supplementation on Blood Pressure in Children With Overweight Condition or Obesity

Key Points Question Is maternal intake of prenatal docosahexaenoic acid associated with reductions in childhood blood pressure? Findings In this prespecified secondary analysis of a randomized clinical trial and follow-up of 171 mothers and their offspring, the children of women randomized to receive prenatal docosahexaenoic acid did not experience an increase in systolic blood pressure and diastolic blood pressure if they became overweight or obese compared with children of women randomized to placebo. A statistically significant interaction was found between prenatal treatment and child weight status. Meaning Docosahexaenoic acid intake during pregnancy may protect children from the blood pressure–elevating consequence of childhood overweight condition or obesity.


Introduction
Consumption of docosahexaenoic acid (DHA) and eicosapentaenoic acid from fish oil is well known to reduce blood pressure (BP) in both adults and children. 1,2However, there has been recent interest in the potential programming association of DHA in utero and in early infancy with long-term physiological functions, including BP.Consistent with this possibility, a large population-based prospective cohort study from the Netherlands found an association between higher in utero DHA exposure and lower systolic BP (SBP) at age 6 years. 3,4Another 4-country European study found lower diastolic BP (DBP) at age 6 years in children randomized to an infant formula containing DHA for only the first 4 months of infancy. 5ly randomized clinical trials of DHA supplementation during pregnancy can determine if improving DHA nutrition during fetal life improves childhood BP.The Kansas University DHA Outcome Study (KUDOS) trial randomized pregnant women to either a DHA supplement of 600 mg per day or a placebo and followed up the growth and development of their offspring to 6 years of age.The follow-up included semiannual BP measurements from 4 to 6 years of age among 179 children (92 DHA and 87 placebo recipients) as an indicator of child health.Based on the findings of Forsyth et al, 5 we hypothesized that maternal DHA supplementation would be associated with reductions in BP in childhood.Results from this randomly exposed cohort agree with those of the Forsyth et al 5 report and 2 recent reports, 3,4 suggesting that prenatal DHA exposure is associated with reductions in childhood BP; however, this report suggests that children who become overweight or obese are the ones protected from higher BP by maternal DHA supplementation.

Methods
The University of Kansas Institutional Review Board approved this prespecified secondary analysis and both the parent randomized clinical trials, KUDOS (NCT00266825) 6 and its follow-up (NCT02487771).The trial protocol is available as Supplement 1.Both the research protocol and written informed consent adhered to the Declaration of Helsinki. 7This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

Participants
From January 10, 2006, through November 17, 2009, 350 pregnant women were enrolled in the KUDOS trial, a phase 3, double-blind, randomized, placebo-controlled clinical trial of DHA supplementation delivered at several local hospitals in the Kansas City, Kansas, metropolitan area.
Women were randomized to either placebo (3 capsules per day containing an equal mixture of soybean and corn oil) or DHA (3 capsules per day of algal oil that provided a total of 600 mg of DHA).alcohol or other drug dependency, and BMI higher than 40 at the time of enrollment into the trial.
Other details of study design and demographic characteristics were included in a report of pregnancy outcome. 6Information on childhood diet 8 and 5-year body composition of the offspring 9 has also been reported.

Anthropometric Assessment
Height was measured without shoes to the nearest 0.1 cm using a wall-mounted stadiometer (Health O Meter; PORTROD).Body weight was obtained to the nearest 0.01 kg using the system integrated electronic scale.Anthropometric data were converted to BMI percentiles with Epi Info, version 3.5.4 (Centers for Disease Control and Prevention).

Dietary Intake
A registered dietitian (E.H.K., J.M.T., or J.M.H.) trained in the multipass 24-hour dietary recall method and neutral probing questions interviewed parents about their child's food intake at each visit throughout the KUDOS and follow-up trials.Multiple dietary recalls have been shown to represent usual dietary intakes. 10

Blood Pressure Measurement
We measured childhood BP using an automated system (CARESCAPE V100 Monitor; GE), with the cuff size recommended for the child's arm circumference.Blood pressure was measured in triplicate: the mean of the last 2 measures for SBP and DBP was taken unless the coefficient of variance was greater than 0.095.If the coefficient of variance was greater than 0.095 (9.2% of SBP and 19.4% of DBP measurements), we used the mean of the 2 closest measurements for each indicator.We excluded from analysis the results of assessments with only 1 reliable data point (n = 23).Irritated or restless children or mechanical failure were the main reasons for exclusion.Members of the staff who measured BP and made decisions related to data quality and calculations were blinded to the mothers' study randomization.[470] mg/d) compared with the DHA group that was obese or overweight (Table 1).However, it was apparent from examining the large variability in both human milk feeding and sodium intake that these differences were not statistically significant.In addition, our preliminary analysis eliminated these factors as important to childhood BP.

Secondary Analyses
We tested a model for SBP and DBP that included maternal DHA randomization, child age and sex, and child BMI (Յ85th or >85th percentile) at 5 years of age as well as important covariates (socioeconomic status, maternal BMI, and additional supplemental DHA taken during pregnancy).
The aim of this analysis was to determine whether these additional factors mediated BP outcomes.
Analyses of SBP yielded a marginal effect for DHA group and a substantial main effect for BMI (Յ85th or >85th percentile).These associations were quantified by a statistically significant P for interaction = .04(Table 2).Children who were obese or overweight (with a 5-year BMI >85th percentile) whose mothers were randomized to the placebo group had a mean (SE) SBP of 104.28 (1.37) mm Hg compared with obese or overweight children in the DHA group (100.34 [1.02]), a difference of 3.94 mm Hg.Maternal prepregnancy BMI was a statistically significant covariate with higher BMI associated with higher childhood BP (r = 0.284, P = .001).The analysis also yielded a statistically significant main effect for child age as SBP was higher at 5.5 years of age than any other age (F = 7.385; P = .001).For example, BP at 5.5 years in overweight condition and obese children in

Discussion
The prevalence of high BP during childhood is on the rise, partly owing to the high rates of obesity during childhood.A study of more than 20 000 US children aged 2 to 17 years found 36% had high BP (>90th percentile for weight, age, and height) at least once in a year. 11A meta-analysis found that BP in childhood tracks to BP in adulthood 12 and that higher BP in childhood is associated with risk for hypertension in adulthood. 13Some of us began measuring childhood BP in 2010 as part of monitoring the overall health of children born to women in the KUDOS trial and because a report of early postnatal DHA supplementation 5 linked lower BP at age 6 years to higher DHA exposure.
Subsequent reports from the Netherlands 3,4 associating prenatal DHA exposure with lower a These values represent the interaction between randomization group and weight status.The P for interactions = .04for systolic blood pressure and P for interactions = .01for diastolic blood pressure.
b Individual value for SBP and DBP of the placebo group with BMI greater than the 85th percentile differs from the 3 other groups (P < .05).We found a statistically significant difference in DBP, which was lower in the DHA group compared with the placebo group; however, childhood weight status emerged as a major factor in both SBP and DBP.In the mixed model analysis that included influential covariates, it became clear that the advantages from maternal DHA supplementation accrued only to children at risk for elevated BP owing to obesity or overweight condition.Both SBP and DBP showed a substantial interaction between maternal DHA randomization and child weight status, suggesting that maternal DHA supplementation exerted a protection for offspring who became overweight or obese in childhood.The increase in BP among children who were obese or overweight in the placebo group compared with the DHA group was large (3.94 mm Hg for SBP vs 4.97 mm Hg for DBP) and statistically significant.Although we knew that obesity or overweight condition is associated with higher BP in adults and children, we did not hypothesize a priori that children who were overweight or obese as a group would gain an advantage from intrauterine exposure to DHA.
We chose weight status at 5 years of age for the mixed model analysis as the midpoint in the ages of assessment.We realized that the weight status of individual children may have changed between 4 and 6 years of age, but the BMI of subgroups varied little among ages 4, 5, and 6 years as shown in Table 1.In fact, the groups who were obese or overweight at 5 years of age already had a higher BMI by 2 years of age than did their leaner counterparts from the same prenatal randomization.We doubt we would have observed an association between DHA supplementation and childhood BP had the incidence of overweight condition and obesity not been so high (29%) in this cohort of children.Nevertheless, the incidence is not markedly discrepant from the base rate in children in general in the United States, in which 1 in 5 school-aged children is obese. 14rsyth et al 5 did not report BMI or state that they adjusted for child BMI in their analysis.They percentile at all ages.The authors of the Generation R study 3 adjusted for BMI in linear regression models and still observed a substantial association between maternal ω-3 status and BP; however, the difference in BP was quite small.
Although smoking was variable within the subgroups, the mean number of smoking days during pregnancy appeared higher in mothers whose children were obese or overweight in both maternal randomizations.[17][18] It is important to consider that the protection against BP in children who are obese or overweight during the preschool-and early school-aged periods is the result of a relatively brief intervention that occurred as much as 6 years earlier.A persistent physiological implication after an intrauterine intervention supports early developmental programming.Hypotheses derived from the Developmental Origins of Health and Disease model posit that less-than-optimal conditions or events experienced in utero are linked to adult disorders, including hypertension.
and reduced cortisol response to a stressor in infancy. 23If these effects were prolonged after birth, either one could affect childhood BP.
Our results suggest an advantage of increased intrauterine DHA exposure to programming of BP response to excess body weight.The children studied in these trials were all fed human milk (a source of DHA) or infant formula with DHA (DHA was added to US formulas after 2002) exclusively or in combination.Consequently, we interpret these findings as an advantage of intrauterine DHA exposure for childhood BP.In the only other randomized clinical study that measured childhood BP after exposure to DHA early in development, Forsyth et al 5 found that infants who were fed a formula with DHA for the first 4 months of life had lower SBP and DBP at age 6 years.Their results suggest that DHA exposure can program lower childhood BP for at least some period after birth.
The absolute SBP and DBP for the 4 randomization-weight status groups are shown in Figure 2.
In searching the literature, we were unable to find chronological reports of BP in children at these ages.We believe these results may be of interest to those who will study BP in young children in the future.

Limitations
A limitation of the study is that mothers in the placebo group with children who were overweight or obese provided a fewer number of days of human milk, and their children consumed more sodium than did children who were overweight or obese in the DHA group.Although these subgroup differences were clearly not statistically significant because of large SDs, they do suggest that subgroups of children may not be entirely comparable.Consequently, they limit the conclusion that DHA alone mitigated the association between overweight or obese weight status and childhood BP.
The generalizability of this secondary analysis may be limited to children of mothers expected to have a healthy pregnancy and not be morbidly obese.

Conclusions
Children of women randomized to DHA supplementation compared with those randomized to receive placebo during pregnancy appeared to be protected against the increase in SBP and DBP associated with childhood overweight condition and obesity.

Figure 2 . 4
Figure 2. Systolic and Diastolic Blood Pressure at Each Age by Maternal Docosahexaenoic Acid (DHA) Randomization

reported 2 . 3 mm
Hg lower SBP and 3.6 mm Hg lower DBP at 6 years of age in children randomized to a formula containing DHA for the first several months of life.If our findings are correct that the advantage of DHA supplementation for BP occurs only in children who are obese or overweight, we speculate that the Forsyth et al 5 cohort may also have had a high proportion of children with overweight condition or obesity.Children in the Generation R study 3 (n = 4455) were leaner than the children in our cohort: their mean BMI was just above the 50th percentile (actual BMI, 15.8; range, 13.6-21.2) at 6 years of age, whereas the mean BMI in this secondary analysis exceeded the 60th

JAMA Network Open | Nutrition, Obesity, and Exercise Effect
Randomization schedules generated by the biostatistics department were for 2 maternal age groups of Docosahexaenoic Acid on Blood Pressure in Children With Overweight Condition or Obesity 25.99 years and 26-35.99years).Each 8-random number sequence included 4 assignments per group stratified by treatment.The Investigational Pharmacy at the University of Kansas Medical Center assigned women to placebo or DHA on the basis of age shared by study personnel.Both the placebo and DHA capsules were orange flavored and provided by DSM Nutritional Products.Women were enrolled at a mean (SD) of 14.5 (3.7) weeks' gestation but had to be enrolled before 20 weeks' gestation.The primary aims of the KUDOS trial were to evaluate the effect of prenatal DHA JAMA Network Open.2019;2(2):e190088.doi:10.1001/jamanetworkopen.2019.0088(Reprinted) February 22, 2019 2/10 Downloaded From: https://jamanetwork.com/ on 09/17/2023 (16- During infancy, the child's dietary recalls were collected at 6 weeks and 4, 6, 9, and 12 months as well as from ages 1 to 6 years at 6-month intervals.Recalls were entered into the Nutrition Data System for Research, version 2008-2014 (University of Minnesota) and evaluated for accuracy by a second registered dietitian (one of E.H.K, J.M.T., or J.M.H.).If the parent or child was unable to recall 1 or more meals in a 24-hour period, the recall was excluded from analysis.The mean results of all 24-hour recalls obtained between 1 and 5 years of age were calculated to estimate the dietary sodium intake (grams per 1000 kcal).

Table 1 .
6n total, 171 children (88 [51.5%] female) were included in this analysis.Of these children, 89 (52.0%) were randomized to the DHA group and 82 (47.9%) to the placebo group.The characteristics of children in the maternal randomization groups by weight status (BMI Յ85th or >85th percentile) are included in Similar to those in the KUDOS trial cohort,6children in the DHA group had a higher mean (SD) Data analysis was performed (J.W. and J.C.) fromMay 23, 2017, to July 10, 2018.Children's SBP and DBP measured at 4, 4.5, 5, 5.5, and 6 years of age were the dependent variables for this study.We conducted a type 3 test of fixed effects to test for the association of maternal randomization (placebo or DHA) and other independent variables with BP.Birth weight, mean sodium intake per 1000 kcal, days fed human milk, maternal gestational weight gain, and race and days smoked during pregnancy were eliminated as influential covariates.The secondary mixed model analysis included maternal DHA randomization, child weight status (BMI Յ85th or >85th percentile) at 5 years of age, 6 years of age.Two-sided P Յ .05 for a 2-tailed analysis was considered statistically significant.Results

Table 1 .
Child Characteristics by Maternal Docosahexaenoic Acid (DHA) Randomization and 5-Year Weight Status aAnalysis of DBP yielded statistically significant main effects for the DHA group and BMI.These associations were quantified by a statistically significant interaction (Table2), such that children who were obese or overweight had higher DBP only if their mothers did not receive the DHA supplement Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); DHA, docosahexaenoic acid.a No statistical comparisons were made.JAMA Network Open | Nutrition, Obesity, and Exercise Effect of Docosahexaenoic Acid on Blood Pressure in Children With Overweight Condition or Obesity JAMA Network Open.2019;2(2):e190088.doi:10.1001/jamanetworkopen.2019.0088(Reprinted) February 22, 2019 5/10 Downloaded From: https://jamanetwork.com/ on 09/17/2023

Table 2 .
Systolic and Diastolic Blood Pressure as a Function of Maternal DHA Randomization and 5-Year Weight Status 21rdling 19first proposed a theory for cancer that Knudson20termed a 2-hit or multiple-hit hypothesis.The hypothesis, then adopted by Li et al21in Effect of Docosahexaenoic Acid on Blood Pressure in Children With Overweight Condition or Obesity higher BP after the second hit of childhood overweight condition or obesity.We cannot know the biological mechanism, but evidence from 2 trials of DHA supplementation during pregnancy suggests 2 plausible mechanisms: intrauterine DHA resulted in more mature fetal autonomic nervous system relation to the Developmental Origins of Health and Disease model, is that in utero insults may only manifest as disease after a physiological second hit.Intrauterine malnutrition is a developmental insult.In the case of our findings, we suggest that less-than-optimal development of some organ system or systems associated with poor maternal DHA status is the first hit, which manifests as JAMA Network Open | Nutrition, Obesity, and Exercise JAMA Network Open.2019;2(2):e190088.doi:10.1001/jamanetworkopen.2019.0088(Reprinted) February 22, 2019 7/10 Downloaded From: https://jamanetwork.com/ on 09/17/2023