Long-term neurological consequences of hypertensive disorders of pregnancy: Implications for postpartum monitoring and intervention

Table 1 Final values for categorical outcomes reported in included studies

Outcome category	No. of studies	Typical maternal age at outcome	Typical follow-up duration
Stroke	[23]	45-55 years	15-20 years (range: 1-50)
Structural/MRI	[10]	50-65 years	15-25 years (range: 5-40)
Cognitive	[9]	55-70 years	20-30 years (range: 10-40)
Mechanistic/biomarker	[6]	35-45 years	7-10 years (range: 1-15)

Follow-up durations were averaged across all studies.

Table 2 Summary of representative studies reporting long-term stroke and vascular outcomes in women with a history of hypertensive disorders of pregnancy

Ref.	Design/setting	Exposure/comparison	Sample (exposed/ref)	Follow-up (years)	Stroke outcome(s)	Effect (95%CI)	Notes
Verburgt et al[19], 2025	Case-control, young women with ischemic stroke vs population controls	Any APO; HDP (PE/PIH) vs none	358 stroke/714 ref	Up to prior pregnancies	Ischemic stroke	PE OR: 4.0 (2.4-6.8); HDP OR: 2.0 (1.4-2.7); SGA OR: 2.8 (2.0-3.9); Preterm OR: 2.7 (1.9-4.0)	HDP strongest with large-artery disease; suggests atherosclerotic mechanism in some cryptogenic strokes
Crump et al[20], 2025	Nationwide cohort, Sweden	APOs (PTB, SGA, PE, other HTN, GDM) vs none	35824 stroke/2201393 total	Up to 46	All stroke	PE aHR: 1.36 (1.31-1.41); other HTN 1.82 (1.67-1.98); GDM 1.86 (1.69-2.04); PTB 1.40 (1.36-1.45); SGA 1.26 (1.22-1.29)	Risks persisted 30-46 years; partly explained by shared familial factors
Hung et al[21], 2022	Nationwide cohort, Taiwan	HDP subtypes vs non-HDP	13617 HDP/54468 ref	≤ 17	Any, ischemic, hemorrhagic	Any aHR: 1.71 (1.46-2.00); Isch aHR: 1.60 (1.35-1.89); Hem aHR: 2.98 (2.13-4.18)	Isch peak 1-3 years (aHR: 2.14); Hem peak 10-15 years (aHR: 4.64); highest with CH superimposed PE (aHR: 3.86)
Auger et al[36], 2020	Provincial cohort, Canada	Very/moderate PTB vs term; mediation by maternal vascular disorders (including PE)	1199364	Approximately 10-20	Ischemic stroke hospitalization	Incidence higher with PTB. For 95%CI PE explained 8.3% (very PTB) and 11.0% (moderate PTB) of PTB-stroke association	Specific stroke HRs by PTB reported; PE key mediator of PTB → stroke/CVD
Garovic et al[23], 2020	Historical cohort, Olmsted County (United States)	Prior HDP vs matched referents	571 HDP/1142 ref	Median 36	Any stroke	HR: 2.27 (1.37-3.76)	HDP also increase CAD, CKD, arrhythmia; robust to adjustments
Arnott et al[24], 2020	National registry, Australia	HDP (early-onset < 34 weeks) × smoking vs no HDP	528106 (first births)	10 (risk estimate)	Composite CVD (including ischemic stroke)	EO-HDP, non-smokers HR: 4.90 (3.00-7.80); EO-HDP + smoking HR: 23.5 (13.5-40.5)	Stroke included in composite; strong interaction with smoking
Gastrich et al[25], 2020	Matched cohort, United States	History of PE vs matched controls	6360 PE/325347 ref	Up to several years	Hospitalized stroke	HR: 1.81 (0.75-4.37) (NS)	MI and CV death increase; stroke directionally increase but underpowered
Kuo et al[26], 2018	Nationwide cohort, Taiwan	PE/eclampsia vs age-matched controls	1295 cases/5180 ref	Approximately 10	Cerebrovascular disease (stroke)	Eclampsia HR: 10.71 (3.45-33.24); PE HR: 3.47 (1.46-8.23)	Hemorrhagic stroke: Eclampsia HR: 19.74; PE NS
Lobitz et al[27], 2024	National admin cohort, Australia	Caesarean vs vaginal; adjusted including HDP/DM	14179299	≤ 365 days PP	Stroke readmission	HR: 1.40 (1.26-1.56)	Increase 180-365 days (HR: 1.94)
Lin et al[29], 2016	Nationwide cohort, Taiwan	PIH vs non-PIH	28346 PIH/113384 ref	Approximately 10	Intracranial hemorrhage	aHR: 2.81 (1.58-4.99)	Hemorrhagic risk increasing strongly
Nelander et al[30], 2016	Cohort, Sweden (≥ 65 years)	Any HDP vs none	3232	To late life	Any stroke	HR: 1.36 (borderline)	Attenuated at oldest ages
Akhter et al[35], 2014	Vascular imaging cohort, Norway	Severe PE vs normotensive	42 PE/44 ref	9 years (range 1-13)	Vascular surrogate (carotid IMT/plaques)	Intima 013 ± 0.02 mm vs 0.08 ± 0.01 mm (+63%); I/M ratio 0.27 ± 0.07 vs 0.15 ± 0.03 (+80%); IMT 063 ± 0.12 mm vs 0.61 ± 0.12 mm (+3%); media -7%	Increased carotid intimal thickening and arterial remodeling consistent with early subclinical atherosclerosis after severe PE

Of 23 included studies, 12 with detailed effect estimates, or national cohort data are shown here. CI: Confidence interval; APO: Adverse pregnancy outcome; HDP: Hypertensive disorder of pregnancy; PE: Preeclampsia; PIH: Pregnancy-induced hypertension; Ref: Reference/control group; OR: Odds ratio; SGA: Small for gestational age; HTN: Hypertension; GDM: Gestational diabetes mellitus; PTB: Preterm birth; aHR: Adjusted hazard ratio; CH: Chronic hypertension; HR: Hazard ratio; Isch: Ischemic stroke; Hem: Hemorrhagic stroke; PTB → stroke/CVD: Pathway from preterm birth to stroke or cardiovascular disease; CAD: Coronary artery disease; CKD: Chronic kidney disease; EO-HDP: Early-onset hypertensive disorder of pregnancy; NS: Not significant; MI: Myocardial infarction; CV: Cardiovascular; DM: Diabetes mellitus; I/M: Intima/media ratio; IMT: Intima-media thickness.

Table 3 Structural and neuroimaging findings after hypertensive disorders of pregnancy per PubMed analysis, including white matter hyperintensities, brain atrophy, and vascular changes

Ref.	Country/design	Exposure (HDP subtype)	Sample (exposed/ref)	Time since index pregnancy	Imaging modality/measure	Main finding(s)	Notes/adjustments
Hussainali et al[41], 2025	ORACLE substudy, Netherlands	Gestational HTN or PE	63 GH, 30 PE/445 ref (n = 538 total)	Median 14.6 years	MRI markers of CSVD (WMH volume, lacunes, microbleeds)	WMH volume increase after HDP [β = 0.32 (0.08-0.56)]; driven by GH [β = 0.39 (0.10-0.67)]; no differences in lacunes or microbleeds. Trend increase with later chronic HTN	Adjusted for age, BMI, education, ICV
Alers et al[42], 2023	Netherlands; cohort MRI + cognition	Prior PE	96/96 ref	Approximately 15 years	MRI-gray and white matter volumes, WML load, and cognitive testing	Cortical and subcortical gray and white matter volumes decrease after PE; WML burden increase; slower processing speed and impaired executive function vs controls (all P < 0.05)	Adjusted for age, BMI, and education
Canjels et al[43], 2022	Netherlands; case-control MRI	Prior PE	22/13 ref	6.6 years PE vs 9.0 years ref	Contrast-enhanced MRI: BBB leakage (Ki, vl)	BBB leakage higher in PE across whole cerebrum-global WM Ki increase (P = 0.001) and GM Ki increase (P = 0.02); aOR for high Ki approximately 48 (3.5-651) WM, 3.5 (1.1-30.8) GM	Adjusted for age, HTN at MRI, Fazekas
Akhter et al[34], 2019	Sweden; vascular imaging	Prior PE	23/35 ref	Approximately 7 years	Carotid ultrasound: Intima thickness increase, I/M ratio increase, IMT decrease	At 7 years, intima 012 mm vs 0.09 mm, I/M increase (P < 0.001); IMT paradoxically lower. Correlated with MAP and biomarkers	Adjusted for BMI, BP, MAP
Siepmann et al[44], 2017	Germany; cohort imaging	Prior PE	34/40 ref	5-15 years	3T MRI: WM lesions, microstructure (DTI), cortical volumes	Temporal-lobe WM lesion volume increase after PE (P = 0.04); cortical GM volume decrease and DTI showed lower FA in parietal-occipital regions (P < 0.05)	Associations persisted after adjustment for MAP, HDL, and age
Postma et al[52], 2016	Netherlands; cross-sectional	Prior PE	41 eclampsia, 49 PE/47 ref	Approximately 6 years	MRI: WML presence, infarcts, periventricular/subcortical	WML more frequent in PE (40% vs 21%, P = 0.03). Infarcts only in PE. WML not tied to objective cognition	Adjusted for age and risk factors; subcortical WML P = 0.06; infarcts 4 vs 0
Mielke et al[46], 2016	United States; community cohort	History of HDP	286 HDP/297 ref	Approximately 35 years median follow-up	MRI: Total brain volume, WM lesion volume; cognition	Smaller brain volume in HDP (P = 0.023). WMH volume slightly increase (8.9 mL vs 8.1 mL, NS). Processing speed slower in HDP	Adjusted for demographics and vascular risk
Akhter et al[35], 2014	Sweden; case-control	Prior severe PE	42/44 ref	Approximately 9 years (range 1-13)	Carotid wall imaging (intima, I/M, IMT)	Intima 013 mm vs 0.08 mm (+63%); I/M ratio 0.27 vs 0.15 (+80%); IMT no difference	Structural differences persisted after adjustment for BMI, age, BP
Aukes et al[47], 2012	Netherlands; cross-sectional	Prior PE (early vs late onset)	73/75	Approximately 5 years	MRI: WM lesion volume, periventricular WML, infarcts	WML burden increase after PE (β = 0.77; P = 0.04); early-onset PE highest (P < 0.01); periventricular WML 5 vs 0	Current HTN independently increase WML risk
Aukes et al[48], 2009	Netherlands; case-control	Former eclampsia	39/29	Median 7 years (IQR 3-14)	MRI: WML load (volume and severity)	WML volume greater in eclampsia (0.041 mL vs 0.004 mL, P = 0.016). More WML in women with seizures (P = 0.01)	WML load is correlated with seizure recurrence

HDP: Hypertensive disorder of pregnancy; Ref: Reference/control group; HTN: Hypertension; PE: Preeclampsia; GH: Gestational hypertension; MRI: Magnetic resonance imaging; CSVD: Cerebral small-vessel disease; WMH: White-matter hyperintensity; BMI: Body mass index; ICV: Intracranial volume; WML: White-matter lesion; BBB: Blood-brain barrier; K_i: Leakage rate constant; v_l: Leakage volume fraction; WM: White matter; aOR: Adjusted odds ratio; GM: Gray matter; aOR: Adjusted odds ratio; I/M: Intima/media ratio; IMT: Intima-media thickness; BP: Blood pressure; MAP: Mean arterial pressure; DTI: Diffusion tensor imaging; FA: Fractional anisotropy; HDL: High-density lipoprotein; NS: Not significant; β: Regression coefficient; IQR: Interquartile range.

Table 4 Cognitive outcomes following hypertensive disorders of pregnancy per PubMed analysis, including objective neuropsychological testing and subjective patient-reported measures

Ref.	Country/design	Exposure (HDP subtype)	Sample (exposed/ref)	Time since index pregnancy	Cognitive measure(s)	Main finding(s)	Notes/adjustments
Alers et al[42], 2023	Netherlands; population cohort	Prior PE	1036/527 ref	≤ 19 years	BRIEF-A executive and working memory composite	Executive function and working memory decrease (aRR: 9.20 and 7.94 at one year; sustained significance at up to 19 years postpartum)	Adjusted for age, BMI, and education
Wang et al[49], 2022	United States; Framingham Offspring study (longitudinal)	Prior PE	142/1107 ref (n = 1249)	Median 12 years	Dementia diagnosis (all-cause, Alzheimer’s, vascular)	PE → increase risk of all-cause dementia (HR: 1.56) and Alzheimer’s dementia (HR: 1.65)	Longitudinal adjudicated outcomes; adjusted for cardiovascular risk
Garovic et al[23], 2020	United States; retrospective cohort (Olmsted County)	GH, PE, superimposed HDP	571 HDP/1142 ref (mothers total number = 7544)	Median approximately 36 years	Dementia diagnosis	Dementia signal present but ns after multiple testing	Adjusted for education, smoking, obesity; CVD endpoints significant
Dayan et al[50], 2018	United States; retrospective cohort (CARDIA)	Prior HDP	193/375 ref (n = 568 total)	25 years at testing	Neuropsychological testing (digit symbol substitution test, Stroop Test Trial 3, Rey Auditory Verbal Learning Test)	Processing speed and executive function decrease (unadjusted P = 0.01-0.05; differences attenuated and ns after adjustment)	Adjusted for age, BMI, hypertension, education, and depression at approximately 18 years postpartum
Fields et al[51], 2017	United States; case-control (Mayo Clinic)	Prior eclampsia	40/40 ref	Approximately 35 years	Neuropsych battery (executive function, memory, attention; trail making B, logical memory I/II, verbal learning, auditory attention)	Executive and memory decrease (d = 0.5-2.0); cognitive impairment 20% vs 8%; CAC increase in impaired hPE group (P = 0.043)	Adjusted for age, education, hypertension, BMI, and medication use; ApoE-4 and mood symptoms were not explanatory
Mielke et al[46], 2016	United States; community cohort + MRI	History of HDP	208/1071 ref	Mid-late life	Processing speed (digit symbol, TMT-A, Stroop) decrease in HDP (all P ≤ 0.035); memory, language, executive ns	Processing speed decrease on digit symbol (P = 0.005), TMT-A (P = 0.035), and Stroop (P = 0.002); memory/Language/executive ns. Processing speed correlated with lower brain injury volume (b = -0.36, P = 0.004)	Adjusted for hypertension duration, nulliparity, and eGFR; results robust in sensitivity models
Postma et al[52], 2016	Netherlands; case-control MRI + neuropsych battery	Prior PE	46 eclamptic + 51 PE/47 ref	Approximately 6 years	Motor speed (TMT-5, P < 0.01); other domains ns	CFQ increase (44 ± 16 vs 36 ± 11, P < 0.001) and HADS increase (11 ± 6 vs 8 ± 5, P < 0.001) in HDP vs controls; no objective cognitive impairment on neuropsych testing	Subjective dysfunction persisted despite normal objective scores; WML not predictive of cognition
Nelander et al[30], 2016	Sweden; population cohort (≥ 65 years)	HDP ± proteinuria	419 HDP/2646 ref	Decades later	Registry diagnosis of dementia	Dementia: 7.6% vs 7.4% (HR: 1.19, 95%CI: 0.79-1.73); CVD: 22.9% vs 19.0% (HR: 1.29, 95%CI: 1.02-1.61); stroke: 13.4% vs 10.7% (HR: 1.36, 95%CI: 1.00-1.81)	Subgroup analysis by proteinuria; registry follow-up
Postma et al[45], 2014	Netherlands; cross-sectional questionnaires ± MRI subset	Eclampsia and PE	46 eclamptic + 51 PE/48 ref	Approximately 6-7 years	CFQ, HADS, neuropsych battery	CFQ and HADS increase in PE/eclampsia vs controls (P < 0.01); visuomotor speed slower (Trail Making B P < 0.001); other domains ns	Persistent subjective cognitive and mood symptoms post-HDP; not time-dependent

HDP: Hypertensive disorders of pregnancy; PE: Preeclampsia; Ref: Reference/control group; BRIEF-A: Behavior Rating Inventory of Executive Function - Adult Version; aRR: Adjusted risk ratio; BMI: Body mass index; HR: Hazard ratio; GH: Gestational hypertension; CVD: Cardiovascular disease; CAC: Coronary artery calcification; hPE: Hypertensive preeclampsia; TMT: Trail Making Test; eGFR: Estimated glomerular filtration rate; CFQ: Cognitive Failures Questionnaire; HADS: Hospital Anxiety and Depression Scale; ApoE-4: Apolipoprotein E ε4 allele; WML: White matter lesion; CI: Confidence interval.

Table 5 Biomarker and mechanistic studies in women with a history of hypertensive disorders of pregnancy, including vascular imaging, molecular markers, and cardiovascular risk profiles

Ref.	Design and setting	HDP subtype	Sample (exposed/ref)	Time since index pregnancy	Biomarker/mechanistic focus main findings	Notes/adjustments
Venkatesh et al[32], 2025	Secondary analysis, HAPO follow-up study	Prior HDP	476 HDP/3250 ref	10-14 years	ASCVD risk increase in HDP (median 7.0% vs 5.8%, P < 0.001) vs normotensive/no GDM; risk additive when both present. HDL decrease, BMI increase, BP increase, HbA1c increase in exposure groups	Adjusted for BMI, BP, lipid profile
Hromadnikova et al[37], 2019	Case-control, prospective follow-up	Prior PE	101 PE/89 ref	3-11 years	Circulating microRNAs (e.g., miR-17-5p, miR-20b-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, etc.) linked to CVD risk	Standardized serum profiling; corrected for age and BMI
Akhter et al[34], 2019	Case-control, vascular imaging + serum assays	Prior PE	23 PE/35 ref	Approximately 7 years	Intima increase (0.12 mm vs 0.09 mm, P < 0.0001); I/M ratio increase (P < 0.001); Endostatin increase and Apo B increase correlated with vascular remodeling (rs approximately 0.35-0.38, P < 0.01)	Independent of carotid IMT confounders; adjusted for BMI, MAP, and time since follow-up
Orabona et al[53], 2019	Echo/cardiac imaging study	EO-PE, LO-PE	30 EO-PE, 30 LO-PE/30	Approximately 2 years (6 months-4 years)	LV mass increase, concentric remodeling (60% of EO-PE; 53% of LO-PE), diastolic dysfunction in EO-PE > LO-PE	Adjusted for age and hemodynamic parameters
Siepmann et al[44], 2017	MRI + risk factor analysis	Prior PE	34 PE/40 ref	5-15 years	WM lesion volume increase (23.2 ± 24.9 μL vs 10.9 ± 15.0 μL, P < 0.05); lower FA on DTI; correlated with BP and HDL levels	Structural-functional link between microstructure and vascular status
Fields et al[51], 2017	Cognitive cohort with imaging + CVD risk factors	Prior PE	40 PE/39 ref	Approximately 35 years	Coronary artery calcification (endothelial remodeling) increase (67.5 AU vs 0.0 AU, P = 0.043) linked to cognitive impairment; ApoE ε4 ns	Adjusted for HTN, BMI, medications

HDP: Hypertensive disorders of pregnancy; Ref: Reference/control group; HAPO: Hyperglycemia and Adverse Pregnancy Outcome; ASCVD: Atherosclerotic cardiovascular disease; GDM: Gestational diabetes mellitus; HDL: High-density lipoprotein; BMI: Body mass index; BP: Blood pressure; HbA1c: Hemoglobin A1c; PE: Preeclampsia; CVD: Cardiovascular disease; I/M: Intima/media ratio; IMT: Intima-media thickness; Apo B: Apolipoprotein B; MAP: Mean arterial pressure; EO-PE: Early-onset preeclampsia; LO-PE: Late-onset preeclampsia; LV: Left ventricular; WM: White matter; DTI: Diffusion tensor imaging; FA: Fractional anisotropy; HTN: Hypertension; ApoE-4: Apolipoprotein E ε4 allele; ns: Not significant.