While PE has long been considered a transient and reversible condition that typically resolves within a few months postpartum, increasing evidence suggests a potential long-term impact on maternal health. The relationship between PE and cardiovascular or renal disease is complex and likely bidirectional as these conditions share a common underlying mechanism: endothelial cell damage, which is widely accepted as a key factor in the pathogenesis of PE. Consequently, pre-existing disorders characterized by endothelial dysfunction—such as hypertension, CKD, diabetes, and connective tissue or vascular diseases—may predispose women to PE [5]. On the other hand, PE, along with gestational hypertension and gestational diabetes, may itself promote vascular endothelial dysfunction, arterial stiffness, and arteriosclerosis [22], thereby contributing to long-term renal damage and increasing the risk of CKD [23, 24]. Given this complex interplay, whether women who experience HDP have an increased long-term risk of developing CKD and ESKD remains unclear.

To address this issue, we analyzed the published and updated literature on the association between PE and the future risk of maternal CKD/ESKD later in life, focusing only on women who were healthy prior to pregnancy. Since PE can be associated with reversible acute kidney injury [25, 26], we included only studies with medium- to long-term follow-up. Our analyses revealed a statistically significant increase in the risk of CKD and ESKD later in life following PE, with the risk of developing ESKD being unexpectedly higher than that of developing CKD. However, high statistical heterogeneity was observed, and the only prospective study did not find a significant association between PE and CKD [9].

One important finding of our review is the discrepancy between the prospective and retrospective studies. This inconsistency is likely caused by several methodological differences. Retrospective studies are more prone to misclassification and selection bias, while prospective studies often apply more rigorous exclusion criteria and control for confounding factors. Furthermore, in our specific case, the prospective study by Behboudi-Gandevani et al. had a shorter follow-up period, which may have led to an underestimation of the association with CKD.

The considerable statistical heterogeneity we observed is not unexpected and mirrors findings from previous meta-analyses in this field [1, 7, 8]. It likely reflects methodological (e.g., different exposure and outcome definitions) and clinical (e.g., different populations) variability across studies investigating long-term renal outcomes after preeclampsia. Subgroup or sensitivity analyses—such as excluding studies with broader CKD definitions or shorter follow-up—might have clarified some sources of heterogeneity. However, these would have required excluding the only available prospective study, [9], reducing methodological diversity. Moreover, given the limited number of ESKD studies, similar analyses could not be conducted without compromising statistical robustness.

The higher relative risk of ESKD compared to CKD may be partly explained by the studies’ limitations. While ESKD is uniformly defined, the definition of CKD was not consistent across the analyzed studies and varied over time. In addition, unexplored risk factors for CKD/ESKD [27,28,29], such as cardiovascular disease, hypertension, and diabetes, developed during follow-up may have confounded the observed association. Where performed, statistical analyses adjusting for these factors have shown a significant reduction in the strength of the association between PE and the future risk of either CKD or ESKD [15, 17, 18]. Additionally, other independent risk factors, such as obesity and smoking, as well as obstetric complications should be considered [14, 17, 19, 30, 31]. Furthermore, few studies took the etiology of renal disease into account. Women with undiagnosed congenital or hereditary kidney diseases (e.g., ADPKD) prior to pregnancy were possibly included in the analyses, leading to an overestimation of the observed association, as previously shown by Khashan et al. [19]. By adjusting and/or excluding more confounders, the strength of the observed association between PE and CKD/ESKD could be further reduced.

Preeclampsia is increasingly regarded as a stress test which may unmask an underlying maternal pathology, such as kidney disease. Proteinuria documented before pregnancy or < 20 weeks’ gestation, persistent proteinuria and/or eGFR impairment after 6 to 8 weeks postpartum all suggest pre-existing undetected renal disease [32, 33], leading to a biased attribution of causality between PE and CKD/ESKD later in life. In fact, Kristensen et al. [19] reported a stronger association between PE and future risk of CKD within five years of the latest pregnancy, while the strength of the association decreased with time (aHR 6.11 (95% CI 3.84–9.72) vs aHR 2.06 (95% CI 1.65–2.50), p < 0.001). Based on these findings and on the literature reviewed, we support the hypothesis that PE often reveals a pre-existing subclinical maternal condition, rather than representing a de novo disease triggered by pregnancy. Nonetheless, the respective contributions of pregnancy-induced endothelial injury and pre-existing dysfunction remain difficult to disentangle and require prospective investigation.

Taken together, these findings raise the question of whether the observed association justifies long-term clinical monitoring. As shown by Ayansina et al. [16], the absolute risk of CKD in women with PE was low (1.3% over 2.5 million person-years of follow-up), resulting in a hypothetical number needed to treat (NNT) of approximately 150 individuals to prevent/detect a single individual with CKD. Similar findings were reported by Covella et al. [1] (NNT 310 for ESKD and 157 for CKD). Given these high NNTs, long-term follow-up for women who have experienced PE to detect conditions with such a low incidence may not be clinically justified.

Endothelial damage is a common pathophysiological feature of various diseases, but it remains unclear whether such damage in PE contributes directly to the long-term risk of kidney disease. Women who have experienced PE may have undiagnosed pre-existing conditions that could potentially accelerate or exacerbate the development of CKD and ESKD. Although the data collected were carefully selected from the database, we cannot completely exclude the presence of an underlying maternal disease before pregnancy. Therefore, the question of causality remains unresolved. Further meta-analyses are unlikely to provide new insights: only well-designed prospective studies with pre-pregnancy baseline assessments–and ideally a nationwide registry–can definitively clarify this relationship.

In conclusion, we suggest that women who suffered from PE should be monitored regularly for one year postpartum. However, in the absence of any signs of renal involvement or hypertension, they should be considered clinically healthy and not at elevated long-term risk for CKD and/or ESKD. Importantly, they should not face unnecessary long-term surveillance or discrimination in insurance or healthcare access.

Most of the studies included in our meta-analysis had large sample size, were longitudinal and had low risk of bias. However, an underlying bias due to the mostly retrospective nature of the studies must be considered, and the possibility of bias due to misclassification of exposure, outcome, or covariate (as well as missing data in the registries used) exists in all studies selected.

Another limiting factor of our research is the possible inclusion of some studies involving populations with other conditions that may lead to long-term kidney damage, such as diabetes and/or autoimmune diseases [29, 34]. It has also been suggested that the severity of pre-pregnancy renal function impairment, rather than the etiology of the renal disease itself, is a risk factor for PE and for accelerated progression of CKD [35].

These limitations show, as reported before, the need for prospective cohort studies to finally assess the question of the causality of the relationship between PE and kidney diseases.