Long-term cardiometabolic health in people born after assisted reproductive technology: a multi-cohort analysis

Abstract

Aims: To examine associations of assisted reproductive technology (ART) conception (vs. natural conception: NC) with offspring cardiometabolic health outcomes and whether these differ with age.

Methods and results: Differences in systolic (SBP) and diastolic blood pressure (DBP), heart rate (HR), lipids, and hyperglycaemic/insulin resistance markers were examined using multiple linear regression models in 14 population-based birth cohorts in Europe, Australia, and Singapore, and results were combined using meta-analysis. Change in cardiometabolic outcomes from 2 to 26 years was examined using trajectory modelling of four cohorts with repeated measures. 35 938 (654 ART) offspring were included in the meta-analysis. Mean age ranged from 13 months to 27.4 years but was <10 years in 11/14 cohorts. Meta-analysis found no statistical difference (ART minus NC) in SBP (-0.53 mmHg; 95% CI:-1.59 to 0.53), DBP (-0.24 mmHg; -0.83 to 0.35), or HR (0.02 beat/min; -0.91 to 0.94). Total cholesterol (2.59%; 0.10-5.07), HDL cholesterol (4.16%; 2.52-5.81), LDL cholesterol (4.95%; 0.47-9.43) were statistically significantly higher in ART-conceived vs. NC offspring. No statistical difference was seen for triglycerides (TG), glucose, insulin, and glycated haemoglobin. Long-term follow-up of 17 244 (244 ART) births identified statistically significant associations between ART and lower predicted SBP/DBP in childhood, and subtle trajectories to higher SBP and TG in young adulthood; however, most differences were not statistically significant.

Conclusion: These findings of small and statistically non-significant differences in offspring cardiometabolic outcomes should reassure people receiving ART. Longer-term follow-up is warranted to investigate changes over adulthood in the risks of hypertension, dyslipidaemia, and preclinical and clinical cardiovascular disease.

Keywords: In vitro fertilization; Blood pressure; Glucose; Lipids; Meta-analysis; Pooled longitudinal trajectory analysis.

Structured Graphical Abstract

Meta-analysis of results in 35000 offspring with cardiometabolic outcomes measured at various ages found no robust differences in blood pressure, heart rate, triglycerides, or hyperglycaemic/insulin resistance traits, and higher cholesterol in ART-conceived than NC offspring. Analysis of cardiometabolic trajectories up to age 26 years in 17,000 offspring identified subtle increases to nominally higher blood pressure and triglycerides in young adults who were conceived by ART. ART, assisted reproductive technology; NC, natural conception; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; TC, total cholesterol; HDLc, high-density lipoprotein cholesterol; LDLc, low-density lipoprotein cholesterol; TG, triglycerides; HbA1c, glycated haemoglobin.

Figure 1

Overview of the included cohorts. The figure shows the birth country, birth years, sample size, and type and age of cardiometabolic outcome assessments in offspring from each included cohort study. The sample sizes represent the maximum number of offspring included in any meta-analysis. ABCD, Amsterdam Born Children and their Development Study; ALSPAC, Avon Longitudinal Study of Parents and Children; BASELINE, Babies After SCOPE: Evaluating the Longitudinal Impact on Neurological and Nutritional Endpoints; BIS, Barwon Infant Study; CHART: Clinical review of the Health of 22–33 years old conceived with and without ART; EDEN, Etude de cohorte généraliste, menée en France sur les Déterminants pré et post natals précoces du développement psychomoteur et de la santé de l’Enfant; GASPII, Gene and Environment: Prospective Study on Infancy in Italy; Gen R, Generation R Study; G21, Generation XXI Study; GUSTO, Growing up in Singapore Towards healthy Outcomes; HGS, Healthy Growth Study; SWS, Southampton Women’s Survey; HUNT, The Trøndelag Health Study. Further details on the included cohorts and measurements can be found in Supplementary material online, Text S1 and Table S1.

Figure 2

Pooled mean differences in cardiometabolic health outcomes between assisted reproductive technology-conceived and natural conception offspring from up to 14 birth cohort studies. The figure shows the pooled confounder-adjusted mean differences (and 95% confidence intervals) in cardiometabolic outcomes between assisted reproductive technology-conceived and natural conception offspring from up to 14 cohort studies. Estimates represent standardized mean differences in (A) systolic blood pressure, diastolic blood pressure, and heart rate, (B) mean % difference in lipids, (C) glucose, and insulin, and mean difference in % glycosylation for HbA1c. Cohort-specific models were adjusted (as fully as possible) for maternal age, parity, education, smoking, body mass index, and ethnicity, plus offspring sex and age at outcome assessment. Results from each cohort are presented in Supplementary material online, Figure S2. ART, assisted reproductive technology; NC, natural conception; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; TC, total cholesterol; HDLc, high-density lipoprotein cholesterol; LDLc, low-density lipoprotein cholesterol; TG, triglycerides; HbA1c, glycated haemoglobin. The pooled mean differences in blood pressure and heart rate in original units were −0.53 mmHg (−1.59 to 0.53) for systolic blood pressure, −0.24 mmHg (−0.83 to 0.35) for diastolic blood pressure, and 0.02 b.p.m. (−0.91 to 0.94) for heart rate.

Figure 3

Pooled mean differences in cardiometabolic health outcomes between assisted reproductive technology-conceived and natural conception offspring, stratified by sex, parental subfertility, fresh embryo transfer/frozen embryo transfer, and in vitro fertilization/intracytoplasmic sperm injection. The figure shows pooled confounder-adjusted mean differences in cardio-metabolic outcomes between pre-specified assisted reproductive technology-conceived and natural conception offspring subgroups. Estimates represent standardized mean differences in (A) systolic blood pressure (SBP), (B) diastolic blood pressure (DBP), and (C) heart rate (HR), and mean % difference in (D) total cholesterol (TC), (E) high-density lipoprotein cholesterol (HDLc), (F) low-density lipoprotein cholesterol (LDLc), (G) triglycerides (TG), (H) glucose, and (I) insulin. The horizontal bars represent 95% confidence intervals. Cohort-specific models were adjusted (as fully as possible) for maternal age, parity, education, smoking, body mass index, and ethnicity, plus offspring sex and age at outcome assessment. ART, assisted reproductive technology; NC, natural conception; IVF, in vitro fertilization; ICSI, intracytoplasmic sperm injection; ET, embryo transfer; FET, frozen embryo transfer. The corresponding numerical results, and results from tests of subgroup differences, are given in Supplementary material online, Table S4.

Figure 4

Predicted mean differences in cardio-metabolic trajectories from childhood to adulthood between assisted reproductive technology-conceived and natural conception offspring. The figure shows the predicted mean differences in cardio-metabolic outcomes from childhood to adulthood between the assisted reproductive technology-conceived and natural conception offspring. The horizontal bars represent 95% confidence intervals. Predicted means were obtained from multicohort (ABCD, ALSPAC, G21, and GUSTO) natural cubic spline mixed-effects models that were adjusted for offspring sex, maternal age, parity, body mass index, smoking, education, ethnicity, and cohort. All models included an interaction between assisted reproductive technology and age. SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; TC, total cholesterol; HDLc, high-density lipoprotein cholesterol; LDLc, low-density lipoprotein cholesterol; TG, triglycerides. The corresponding numerical results are given in Supplementary material online, Table S5. The predicted mean trajectories are presented in Supplementary material online, Figure S6.