Pool walking may temporarily improve renal function by suppressing renin-angiotensin-aldosterone system in pregnant women

Background This study aimed to examine the effect of pool walking on renal function in pregnant women. Methods Fifteen pregnant women (mean gestational age, 37.8 weeks) walked in a pool (depth 1.3 m) for 1 h. A few days later, they walked on a street for 1 h. Within each activity, the starting and ending levels of plasma renin activity (PRA) and serum aldosterone (SA) were compared using paired t-test. Total urine volume, creatinine clearance, and change in PRA levels between each activity were compared by t-test. Regression coefficients for total urine volume and creatinine clearance during pool walking were estimated by linear regression and additionally controlled for the change in PRA levels. Land walking served as the reference group. Results Within each activity, the renin-angiotensin-aldosterone levels were suppressed during pool walking: the mean starting and ending values of PRA and SA were 6.8 vs. 5.5 ng/mL/h (p=0.002) and 654 vs. 473 pg/mL (p=0.02), respectively. Compared to land walking, the decrease in PRA level was more evident in pool walking (−1.27 vs. 0.81 ng/mL/h, p=0.004), resulting in higher total urine volume and creatinine clearance in pool walking (both p<0.05). In regression analysis, after controlling for the change in PRA levels, the significantly elevated regression coefficients for total urine volume and creatinine clearance in pool walking were attenuated. Conclusions Pool walking may temporarily improve renal function in pregnant women, partly through the suppressed renin-angiotensin-aldosterone system. Clinical Trial Registration URL: https://upload.umin.ac.jp/cgi-bin/ctr/ctr_view_reg.cgi?recptno=R000010618 Unique Identifier: UMIN000009051


INTRODUCTION 8
In this study, participants walked in the indoor pool with the head-out position for 1 hour 1 at their own pace. The dimensions of the pool having six lanes were as follows: length 25 2 m (27.3 yds), depth 1.3 m (4.3 ft). The water temperature was set at 31-32°C, and the 3 ambient room temperature at 29-30°C. A few days later (2 to 9 days), the same 4 participants walked on a street, i.e., a conventional prenatal activity, near the Kubota 5 Maternity Clinic for 1 hour at their own pace. We first planned to conduct this study in the 6 typical cross-over design by randomizing half of the participants to the reverse sequence. 7 However, due to a potential risk based on our empirical experience that land walking 8 might induce labor, 20 we finally decided not to choose the typical cross-over design to 9 prioritize the safety of our participants and their babies. Therefore, every participant 1 0 firstly did pool walking and then land walking. The weather conditions during land 1 1 walking were fine/cloudy, the temperature was ~23 to 26°C, and the humidity was ~ 40 to 1 2 50%. During each activity, the participants were allowed to drink water freely. Dates for 1 3 each activity were determined according to the schedule of the participants.

4
The eligible participants in this study were pregnant women (aged 20 to 39) with a 1 5 normal pregnancy (gestational age, 37 weeks and above), who had prenatal care at the 1 6 Kubota Maternity Clinic and had already participated in the pool walking program before 1 7 the study period. Participants were ineligible if they presented cardiovascular disease, 1 8 diabetes, hypertensive disorders of pregnancy, placenta previa/low-lying placenta, activity, all urine samples were collected, and the total amount of urine was recorded. 1 5 With the total urine sample at the end of each walking activity, urine creatinine was 1 6 measured. All laboratory measurements were performed at the Clinical Laboratory 1 7 Center of Fukuoka City Medical Association, Fukuoka, Japan. We calculated creatinine 1 8 clearance as follows: creatinine clearance = urine creatinine (mg/dL) × urine volume 1 9 (mL/min)/[(serum creatinine at the start + serum creatinine at the end)/2 (mg/dL)]. Statistical analysis 2 2 Between each activity, the starting levels of clinical and renal function parameters, as well 2 3 as the changes in these parameters, were compared by t-test. The total urine volume and 2 4 creatinine clearance were also compared by t-test. Within each activity, the starting and 1 ending levels of each parameter were compared by paired t-test. The analyses for total 2 urine volume and creatinine clearance were restricted to 14 participants since it was not 3 possible to collect a sample from one participant during land walking. 4 We estimated the regression coefficient (β) and 95% confidence interval (CI) of 5 pool walking for total urine volume and creatinine clearance, respectively. The land 6 walking activity served as a reference. Age, body mass index, and gestational age were 7 adjusted as confounding variables (model 1). In model 2, the decrease of PRA was 8 additionally controlled as a potential mediating variable, i.e., this variable would not 9 confound the association between walking activity and the renal function but might 1 0 instead help to explain the observed difference between activities. No other potential 1 1 covariates (systolic and diastolic blood pressures, pulse rate, and hematocrit) were 1 2 included due to the small sample size with non-significant results in prior analyses (data 1 3 not shown).

4
Alpha value was set at 0.05, and all p-values were two-sided. Data were analyzed 1 5 using JMP pro 12 (SAS Institute, Cary, NC, USA) and STATA/MP 13.1 (Stata-Corp, College Station, TX, USA). The baseline characteristics of the participants are shown in Table 1. The starting levels of 2 each parameter did not differ between pool and land walking activities ( Table 1). The 3 mean amount of water intake did not differ between activities (mean ± SD: 179 ± 76 4 versus 185 ± 56 mL, p=0.81). 5 Within each activity, the starting and ending levels of clinical and renal function 6 parameters are shown in Table 2. During pool walking, the ending levels of body-weight, 7 systolic and diastolic blood pressures, and hematocrit did not differ from the starting 8 levels. However, the ending levels of pulse rate, PRA, and SA were significantly 9 decreased ( Table 2 and Figure 2). The mean starting and ending levels of PRA were 6.8 ± 1 0 3.4 versus 5.5 ± 2.9 ng/mL/h (p=0.002), respectively; the mean starting and ending levels 1 1 of SA were 654 ± 518 versus 473 ± 304 pg/mL (p =0.02), respectively. By contrast, 1 2 during land walking, the starting and ending levels of the examined parameters did not 1 3 differ (although a tendency of decrease in body weight and pulse rate was observed, Table   1  4 2). 1 5 The results for the between-activity comparisons are shown in Table 2. Compared 1 6 to land walking, the changes in body weight, systolic and diastolic blood pressure, pulse 1 7 rate, and hematocrit did not differ in pool walking. However, the change in PRA differed marginally between pool and land walking activities ( Table 2). As a result, 2 1 compared to land walking, total urine volume and creatinine clearance were significantly 2 2 higher in pool walking (Table 2 and Figure 2) being the mean total urine volume in pool The regression coefficients for total urine volume and creatinine clearance were 3 significantly elevated in pool walking compared with land walking (model 1, renin-angiotensin-aldosterone levels. This study was the first to find that prenatal aquatic activity of head-out pool walking 2 appeared to improve renal function, in combination with the regulation of the Hydrostatic pressure might be mainly attributable to the advantage of our pool 1 2 walking activity. Previous studies suggested that hydrostatic pressure increased 1 3 perfusion and autotransfusion at capillary vessels, particularly in the lower extremities, 1 4 resulting in increased cardiac preload/output and decreased systemic vascular 1 5 resistance. 17,18,22,23 In addition, the renin-angiotensin-aldosterone system and 1 6 sympathetic activity appeared to be suppressed in head-out water immersion with 1 7 improved systemic/renal perfusion. 17,23,24 In our study, we observed increased urine our empirical experience, compared to swimming, pool walking tended to be more 1 2 comfortable for pregnant women since they preferred pool walking (the attendance rates 1 3 for swimming and pool walking activities were ~20% and ~80%, respectively). 20 In 1 4 addition, pool walking should be more "fun" compared to water immersion, in which 1 5 participants just stood/sat still in the water for several hours. 18

6
Our study had some limitations. First, in order to prioritize the safety of our 1 7 participants and their babies, we were not able to conduct our study in the typical 1 8 cross-over design, and every participant first walked in water and then walked on land.

9
This one-arm design might potentially carry-over the effect of pool walking to land 2 0 walking. However, our participants walked on land at least two days after pool walking, 2 1 and we confirmed that the baseline characteristics did not differ between the two 2 2 activities. In addition, previous studies stated that PRA levels recover to baseline levels 2 3 within a few hours after aquatic activities. 27 Therefore, this limitation might not affect our 2 4 conclusions. Second, in addition to our small sample size, patients with hypertensive 1 disorders in pregnancy were excluded, thereby limiting the generalizability of the results. 2 In addition, other circulatory parameters (arterial diameters and arterial blood flow) 18,28 3 and hormonal pathways (atrial natriuretic peptide) 24 were not available. However, our 4 empirical experience might support improved placental perfusion by pool walking 5 (Figure 1). 20 Third, although our data demonstrated temporal changes in the 6 renin-angiotensin-aldosterone levels and renal function, our study did not assess the 7 long-term and dose-response effects. Indeed, we did not observe the between-activity 8 differences in the baseline characteristics, which might suggest that only one session of 9 pool walking activity may not prevent preeclampsia risk along with other hypertensive 1 0 disorders of pregnancy. Therefore, future studies are needed to clarify the long-term 1 1 effect of pool walking with a randomized control design and to explore the potential 1 2 benefits on cardiovascular health among not only pregnant women but also those with 1 3 lifestyle-related diseases such as obesity.

4
In conclusion, this novel, pool walking activity may temporarily improve renal 1 5 function by suppressing the renin-angiotensin-aldosterone system in pregnant women.

6
Aquatic facilities are widely available at community-level (e.g., public indoor pools, as 1 7 well as swimming pools in school, hotel, and fitness club) at least in developed countries 1 8 such as USA and Japan. Hence, stakeholders including researchers, healthcare providers, and policymakers should focus further on the benefits of aquatic physical activities with 2 0 hydrostatic pressure to tackle the norm of hypertensive disorders of pregnancy. analysis and writing -review and editing. Toshiyuki Sasaguri: Conceptualization, 1 2 supervision, and writing -review and editing.  Conflicts of Interest: None.