In this retrospective cohort study, all women who underwent LKD between 1980 and 2017 at the age of 45 years or younger in the University Medical Center Groningen (UMCG) or the Erasmus Medical Center Rotterdam (Erasmus MC) were eligible for inclusion. The study was approved by the medical ethical committees of the UMCG (METc 2014/077) and the Erasmus MC (MEC-2013-585).

LKD-specifics and obstetric characteristics of pregnancies pre-LKD and post-LKD were recorded. Check-ups of eGFR, proteinuria, blood pressure, weight and medication-use are (bi-) annually measured as part of the standard care after LKD in the Netherlands. eGFR was calculated using the CKD-EPI formula [12]. Due to the study design, data on pregnancy outcomes were self-reported by questionnaires sent via post or email. Participants who did not respond were contacted twice by telephone to conduct the questionnaire by a direct interview after obtaining direct informed consent from the subject. When this was not possible, the questionnaire was sent again by post or email. If a donor reported gestational hypertension, preeclampsia and/or Intra Uterine Fetal Demise (IUFD) during post-LKD pregnancy in the questionnaire, their medical record was obtained after written informed consent.

Gestational hypertension was defined as ≥140/90 mmHg measured twice or increase in diastolic blood pressure (DBP) >15 mmHg/systolic blood pressure (SBP) >30 mmHg after 20 weeks of pregnancy, without signs of preeclampsia, according to the guideline of the National Institute for Health and Care Excellence (NICE) [13]. Preeclampsia was defined as it was diagnosed by the attending physician according to the guideline in use, defining preeclampsia as the presence of pregnancy induced hypertension at >20 weeks of gestation and proteinuria [14]. Hypertensive disorders of pregnancy (HDP) were defined as a combined endpoint of either gestational hypertension and/or preeclampsia and/or Hemolysis Elevated Liver enzymes Low Platelets syndrome (HELLP). Miscarriage was defined as the spontaneous loss of a pregnancy before 20 weeks and abortion as the deliberate termination of pregnancy before 20 weeks. IUFD was defined as stillbirth after 20 weeks of pregnancy.

During follow-up, hypertension was defined as a SBP ≥140 mmHg, a DBP ≥90 mmHg and/or the use of antihypertensive medication. Proteinuria was defined as urine protein/creatinine ratio >15 g/mol, urine albumin/creatinine ratio >3.5 mg/mmol or when no ratio could be calculated, proteinuria >0.15 g/L. When none of these quantitative urine measurements were available, a positive urine dipstick was defined as proteinuria.

Statistical analysis was performed using SPSS software version 25 and GraphPad Prism version 9.3.1. Continuous variables were reported as mean with standard deviation (SD) in case of a normal distribution and as median with interquartile range (IQR) in case of skewed distribution.

We compared post-LKD eGFR of women with post-LKD pregnancies (study group) to women with only pre-LKD pregnancies or nulliparous (control group). 221 Women were included with 2149 eGFR measurements. Five women were excluded for this analysis because no eGFR levels were available. Multilevel analysis was adjusted for age at LKD, eGFR before LKD, years after LKD and maximum follow-up time. Mean adjusted eGFR in the study group was not significantly different compared to the control group: 71 mL/min/1.73 m² (SEM 1.32, 95% CI: 68.51–73.70) versus 73 mL/min/1.73 m² (SEM 0.57, 95% CI 71.88–74.11, p = 0.23). As shown in Figure 2, eGFR increased in both groups the first 4 years after LKD. Adjusted eGFR slope per year in the study group was −0.21 mL/min/1.73 m² per year (SEM 0.09, 95% CI: −0.38 to −0.04) versus −0.15 mL/min/1.73 m² per year (SEM 0.05, 95% CI −0.25 to −0.05) in the control-group (p = 0.06). When comparing women with and without HDP pre-LKD, no difference was observed in adjusted mean eGFR after LKD (73 mL/min/1.73 m² (SEM 2.39, 95% CI 68.10–77.46) versus 71 mL/min/1.73 m² (SEM 0.70, 95% CI 69.76–72.51, p = 0.51)). Furthermore, there was no difference in eGFR slope in women with and without HDP pre-LKD (−0.29 mL/min/1.73 m² vs. −0.19 mL/1.73 m² (SEM 0.12, 95 CI: −0.33–0.13, p = 0.39)).

Furthermore, a sub-analysis of eGFR slope was performed solely in women with post-LKD pregnancies. Before pregnancy, 108 eGFR measurements were collected in 31 women, which results in a median of 4 measurements per women (IQR 3). After the first pregnancy, 275 eGFR measurements were collected in 37 women: 214 after a first post-LKD pregnancy, 49 after a second post-LKD pregnancy and 12 after a third post-LKD pregnancy. In total, a median of 7 measurements per woman (IQR 10) were included. eGFR analysis was adjusted for age at delivery, years after LKD and eGFR before LKD. The course of adjusted mean eGFR before and after first post-LKD delivery is illustrated in Figure 3. Adjusted mean eGFR before pregnancy was 78 mL/min/1.73 m² (SEM 0.97, 95% CI 76.19–79.99) and after pregnancy 80 mL/min/1.73 m² (SEM 0.60, 95% CI 78.50–80.84) (p = 0.13). eGFR slope per year was decreasing with −0.19 mL/min/1.73 m² per year (SEM 0.42, 95% CI −1.01–0.62) before pregnancy, and after pregnancy with −0.23 mL/min/1.73 m² per year (SEM 0.10, 95% CI −0.42; −0.43 (p < 0.001). Protein-creatinine ratio was not higher after pregnancy than before pregnancy (14 g/mol versus 15 g/mol, p = 0.83). Risk for a protein-creatinine ratio >15 g/mol was not significantly different after pregnancy compared to before pregnancy (OR 0.83, 95% CI 0.44–1.54, p = 0.55). MAP after pregnancy was not significantly different than before pregnancy (93 mmHg versus 89 mmHg, p = 0.09). Longer time between LKD and first delivery was associated with better eGFR (B 0.96, 95% CI 0.66–1.26, p < 0.001).

Long-term outcomes after LKD are demonstrated in Table 2. During our median follow-up time of 12 years after pregnancy, 14% of the women with only pre-LKD pregnancies used anti-hypertensive drugs versus 9% of the women with only post -KD pregnancies. Seven women were affected by a cardiovascular event (CVE) during follow-up. Six of these women had only pre-LKD pregnancies of whom one had a pre-LKD pregnancy with preeclampsia. She had a myocardial infarction 19 years after LKD. The seventh woman was nulliparous. Only two women developed diabetes mellitus type 2, one with only pre-LKD pregnancies and one was nulliparous. Due to the low incidence of these endpoints, we did not perform further statistical analysis.

Pregnancy outcomes of pre-LKD pregnancies and post-LKD pregnancies are illustrated in Table 3. Miscarriages occurred in n = 23/84 (27%) in post-LKD pregnancies versus vs. 56/413 (14%) in pre-LKD pregnancies. Median year of miscarriage was 1996 (IQR 19) for pre-LKD miscarriages and 2012 (IQR 8) for post-LKD miscarriages. Univariable analysis demonstrated a higher risk for miscarriage in post-LKD pregnancies (OR 2.142, 95% CI 1.12–4.100, p = 0.021). Multivariable analysis adjusted for age and multiple pregnancies per women, showed no significant higher risk of miscarriages in post-LKD pregnancies versus pre-LKD pregnancies (OR 1.19, 95% CI 0.04–40.46, p = 0.92).

Further analysis was performed only in women with pregnancies >20 weeks (n = 186 women, n = 402 pregnancies). Two women had an IUFD: one before LKD (at 33 weeks of pregnancy, unknown cause) and one after LKD (at 26 weeks of pregnancy, probably caused by a placenta infarction). According to the self-reported questionnaires 30/413 (7%) of the pregnancies pre-LKD were complicated by HDP and post-LKD 18/86 (21%) of the pregnancies were complicated by HDP. After studying their medical files, this alleged HDP could be confirmed in 15 pregnancies post-LKD: 7 with preeclampsia and 8 with gestational hypertension. Four pregnancies with preeclampsia had preterm birth and low birthweight. An overview of these women are shown in Supplementary Appendix SA.

In Table 4 we show the results of univariable and multivariable analyses on the risk of the composite outcome HDP, adjusted for multiple pregnancies in one woman. Post-LKD pregnancies had a significant higher risk of HDP, as well as higher MAP before LKD. In univariable and multivariable analysis, the risk of gestational hypertension was not higher in post-LKD pregnancies compared to pre-LKD pregnancies (OR 1.69, 95% CI 0.56–5.08, p = 0.35). However, post-LKD pregnancies did have a significantly higher risk of preeclampsia (OR 14.77, 95% CI 3.07–70.99, p = 0.001). Multivariable analysis identified that women with higher BMI at LKD (OR 1.26, 95% CI 1.08–1.46, p = 0.003) and lower parity (OR 0.38, 95% CI 0.18–0.76, p = 0.007) had a significantly higher risk of preeclampsia. All univariable analysis are presented in Supplementary Appendix SB.

Multilevel univariable analysis on adverse pregnancy outcomes such as birthweight <2,500 g and preterm delivery were not associated with post-LKD pregnancies (p = 0.58 and p = 0.43) (Supplementary Appendix SC). In multilevel multivariable analysis preterm birth was associated with preeclampsia (OR 5.24, 95% CI 1.55–17.70, p = 0.008). Birthweight <2,500 g was associated with HDP during pregnancy (OR 4.875, 95% CI 1.607–14.790, p = 0.005). Absolute birthweight was significantly lower in pregnancies after LKD (p = 0.014, Supplementary Appendix SC).

We demonstrated for the first time that post-LKD pregnancy does not lead to a change in mean eGFR. Interestingly, we did find a small significant difference in eGFR slope before and after post-LKD pregnancy. eGFR decreased with −0.19 mL/min/1.73 m² per year before pregnancy, and after pregnancy −0.23 mL/min/1.73 m² per year (p < 0.001). This might be explained by the fact that eGFR slowly rises the first 4–5 years after LKD, and the median time between LKD and first delivery was 5 years (IQR 4). This phenomenon was recently described in two large prospective studies [17, 18]. In line with the aforementioned findings, we also found that longer time between LKD and first delivery was associated with better eGFR (p < 0.001).

Furthermore, none of the women with post-LKD pregnancies developed a CVE. No difference was found in the in incidence of hypertension in the group with only pre-LKD pregnancies compared to the group with only post-LKD pregnancies. It is known that women with a history of preeclampsia are more at risk for hypertension and CVEs later in life [19, 20]. Recent studies demonstrated that the risk of CVE is significantly lower after gestational hypertension and late onset preeclampsia compared to early onset preeclampsia [21, 22]. Most of the preeclampsia in our post-LKD pregnancies were late onset, mild preeclampsia and none of the women who were pregnant after LKD had a CVE. Of note is that these women had a rather short follow-up time (median 11 years after pregnancy) and in literature CVEs occur at longer periods of time after preeclampsia [23]. Women with HDP before LKD had a longer follow-up time (median 22 years) after the first pregnancy. They did not experience more CVEs or hypertension after LKD compared to women who did not experience HDP. We hypothesize that these women also had mild HDP as they still were able to pass the LKD screening and donated their kidney. In an earlier study in the general population, a similar non-negative effect of HDP was shown on kidney function after pregnancy [24].

In line with earlier literature, we also found a higher risk of HDP in post-LKD pregnancies [10]. It is important to discuss risks of post-LKD pregnancy with women considering pregnancy before LKD. We report an absolute risk for mild near-term preeclampsia of approximately 7% (versus 2%–4% in the general population). More-over we found a median of 200 g lower birth weight in post-LKD pregnancies (absolute, not corrected for confounders), but no increased risk for birth weight <2,500 g. Comparing outcomes of studies remains difficult, especially since studies use different definitions for gestational hypertension and preeclampsia. This higher risk of HDP in pregnancies post-LKD can also be explained by the higher age of the mothers who were pregnant after LKD [22]. Although in our study pregnancies post-LKD were more often complicated by HDP, in line with previous studies no differences were found in adverse fetal outcomes [10]. Furthermore, no difference was observed in eGFR after LKD with or without HDP, in line with an earlier study in the general population [24].

Higher BMI and blood pressure at LKD were associated with adverse pregnancy outcomes. As was shown earlier, women with a high BMI have smaller rest capacity after LKD and therefore are at higher risk for hypertension and HDP [25]. In studies in the general LKD population, hypertensive donors had no increased risk for reduced eGFR, proteinuria or ESRD compared to donors without hypertension [26]. However, whether obese donors have an increased risk of CVE or ESRD in comparison to the general LKD population has not fully been elucidated yet [27–29]. More research is warranted for counseling overweight women with a future pregnancy wish who want to donate their kidney.

To date this is the first study that compared eGFR slopes after LKD in women with and without post—LKD pregnancies, and the first study that compared eGFR slopes before and after post-LKD pregnancy. Besides, we provide long-term follow-up data of pregnancies post-LKD up to 12 years after LKD, which has not been reported before. A few limitations should be taken into account, mainly due to the retrospective design of the study. There was some missing data and data on pregnancies was mainly self-reported, except for the complicated pregnancies after LKD. Furthermore, only a small group of women become pregnant after LKD. Due to these small numbers a prospective study on the effect of pregnancy on eGFR after LKD was not feasible. Therefore, the analysis to determine the impact of HDP on GFR decline and long-term CVE is probably not powered enough to detect small risks, but it does show that there is no large effect. There was a significant difference in age at LKD, eGFR before LKD, age at delivery, follow-up after delivery between the two groups, due to the study design. We adjusted for these differences in the model.

We conclude that pregnancy post-LKD does not affect kidney function of the mono-kidney at long-term follow-up, even when HDP occur. In line with existing literature, we report a higher risk of pregnancy complications, but this does not lead to an increase in adverse fetal outcomes. Therefore, our counseling advice is in line with British guidelines and KDIGO recommendations [30, 31]. We conclude that a pregnancy wish alone should not be a reason to exclude women for LKD. However, for women with high BMI or hypertension at the time of screening, a future pregnancy wish might be a reason to postpone or refrain from LKD.