Evaluation of hydration status and uid intake among low risk postpartum women in local community delivered in Hospital Ampang.

Background: Adequate uid intake is very crucial for mothers during postpartum as any dehydration that ensues during this period may predispose them to the risk of developing venous thromboembolism event. Yet, the sociocultural inuences may have deterred some of them from adequate uid intake especially those from Asian region. This study aims to determine hydration status of low risk postpartum women during postpartum period among Malaysian women. Methods: A prospective observationalstudy involving assessment of hydrational status at 3 points; after 24 hours, at the 3rd week and at the 6th week postpartum was conducted. Apart from hydrational status where patients were requested to recordon weekly basis of their input and output uid intake in a diary, they also underwent a series of biochemical parameters assessment. Results: 56 patients met the inclusion criteria. The sociodemographic data were analysed decriptively whereby all in the patients were in reproductive age group, with mean (SD) age of 27.96 (4.46) years.The majority were of Malay ethnicity 94.6% and 66.1% were multiparous. As many as 92% of them had at least secondary level of education and more than half were employed (55.4%). Subsequent analysis (n=15) at 3rd and 6th week following womenwho had completed all 3 follow ups showed; statistically signicant evidence of underhydration from the biochemical test results. Repeated analysisat the 3rd week and 6th week revealed that there were signicant differences using Friedman test in levels of urea (p=0.001 b ), creatinine(0.008 b ), as well as plasma osmolality (p<0.001 b ). Similarly with urine specic gravity(p=0.002) and urine osmolality (0.002 a ) calculated with repeated measures ANOVA from baseline. This correlated with the overall lack ouid intake taken by patients which recorded 1.78L/day at 3rd week and 2.13L/day by 6th week, values which were lower than recommended amount for nursing mother. Conclusion: The results suggested that lower uid intake among low risk postpartum women during the postpartum period may This study aimed to evaluate the hydration status among low risk post partum women during the six week of connement period within a Malaysian cultural setting . Despite a small cohort of samples there were statistically signicant evidence of underhydration among these local women from the biochemical parameters taken namely the urea, creatinine, urine specic gravity, plasma and urine and osmolality. This further supported by evidence of low uid intake measured perday as compare to other studies and for nursing mother during connement. Altough underhydrated, these women were not dehydrated and had normal electrolytes levels. However, this may predispose the low risk group to possible risk of developing venous thrombosis. Perhaps, a bigger cohort may provide better correlation between hydration and water intake to risk of developing venous thromboembolism. Nevertheless, this evidence is good to emphasize upon our low risk group of women to adequately consume uids during connement a ensure healthier mother, successful breast feeding and reduce risk of venous thrombosis.

Results: 56 patients met the inclusion criteria. The sociodemographic data were analysed decriptively whereby all in the patients were in reproductive age group, with mean (SD) age of 27.96 (4.46) years.The majority were of Malay ethnicity 94.6% and 66.1% were multiparous. As many as 92% of them had at least secondary level of education and more than half were employed (55.4%). Subsequent analysis (n=15) at 3rd and 6th week following womenwho had completed all 3 follow ups showed; statistically signi cant evidence of underhydration from the biochemical test results. Repeated analysisat the 3rd week and 6th week revealed that there were signi cant differences using Friedman test in levels of urea (p=0.001 b ), creatinine(0.008 b ), as well as plasma osmolality (p<0.001 b ). Similarly with urine speci c gravity(p=0.002) and urine osmolality (0.002 a ) calculated with repeated measures ANOVA from baseline.
This correlated with the overall lack o uid intake taken by patients which recorded 1.78L/day at 3rd week and 2.13L/day by 6th week, values which were lower than recommended amount for nursing mother.
Conclusion: The results suggested that lower uid intake among low risk postpartum women during the postpartum period may predisposes risk of venous thromboembolism. Emphasizing adequate amount of uid intake among low risk post partum women is crucial and will reduce possible risk of venous thromboembolism.

Background
Pregnancy is a hypercoagulable state where all three factors in Virchow's triad apply which include hypercoagulability, venous stasis and vascular damage. Although this state is a physiological process in pregnancy, it poses higher risk in some patients to thromboembolic events. The awareness of the risk thromboembolism and efforts in preventing it during antenatal and postpartum is increasing in the recent years. The United Kingdom maternal and mortality rate due to thromboembolism was 0.79 in 100 thousand livebirths (n= 18/155; 11.6%) between the year 2006-2008 (Lewis, 2011). Our local data revealed higher incidents with 1.5 in 100 thousand livebirths in 2007 (n= 7/136; 5.1%). The cases increased to 4.7/100K livebirths in the year 2008 (n= 23/133; 17.3%) (Krishnan, n.d.). Studies had also shown that incidences were 5 times higher during puerperium as compared to during pregnancy and up to 60 times higher as compared to a non pregnant woman. This is especially so, in rst three weeks of postpartum. (Reducing the Risk of Venous Thromboembolism during Pregnancy and the Puerperium, 2015) (POMP et al., 2008) (Heit et al., 2005).Many studies contributed to the formation of guidelines to further improve antenatal and post natal care. Unfortunately the existing risk scoring and screening emphasize only on patients with a known higher risk to develop venous thrombosis event (VTE). These patients receive the package of propercounseling , education tools and are started on anticoagulant with frequent monitoring. On the other hand, the lower risk group does not receive equal attention where their hydrational status and mobilization during their post-partum period are never objectively assessed. This is the window period which we may have missed as higher incidences of VTE happens within 6 weeks postpartum and up to 3 months (POMP et al., 2008) (Heit et al., 2005) (Kamel et al., 2014). This is also especially important in a society that has practices that affect both hydration as well as mobilization of a mother in her con nement period.
Most of Malaysian mothers were taken care by the senior family members according to their cultural practices during the 'con nement' or lying-in period post delivery. Although this cultural practice undoubtly provide good emotional and physical support, in various Asian cultures, there aresome similarities in terms of the caregivers limiting the mothers'types of food and uids they may consume.
These various cultural practices are also similar where most women will be advised to almost completely lie at home with restricted movements (Koon et al., 2005).Some extreme practices even for bid the women from washing their hair and taking bath because the culture believe that the mother needed to be kept warm to expedite the healing process (Koon et al., 2005). Unfortunately these of course exposes the women to risk of sepsis as lack of hygiene. Studies has revealed that higher risk of VTE occurs during the rst few weeks, especially rst to 3rd week of post partum, partly contributed by immobility and postnatal infections(Abdul Sultan et al., 2014).The culture may actually predisposing the mothers to the higher risk of developing VTE despite being a lower risk group.
Although many studies and guidelines emphasize on adequate uid intake during the postpartum period (Koon et al., 2005) (Krishnan, n.d.) there were no exact amount mentioned and written clearly to the patients or in the thrombosis guidelines. Furthermore, there were very few published studies on the uid intake and hydration status of postpartum women in speci cally among our Malaysian mothers adhering to the local customs. Unfortunately there is also not one single index of measurement able to accurately measure the amount of body uid (Armstrong et al., 2012) (Armstrong et al., 2013). This is because the process of water homeostasis is very dynamic, where the body uid balance throughout the day is determined by the uid intake and the physiological process of water absorption, reabsorption, ltration, secretion and excretion, as well as physical activity, posture and body temperature.
The body hydrational status requiresmore than one test to provide accurate assess. Plasma osmolality, P osm is one simple, safe, preciseand portable test recomended for this purpose (Armstrong et al., 2013).A few studies had shown that in subjects assigned to bed-rest, whether with or without supplementation; the plasma and urinary electrolytes such as sodium (Na), potassium (K), calcium (Ca) and magnesium (Mg) excretion levels were signi cantly increased (p≤0.05). The P osm , plasma protein level, whole blood hemoglobin (Hb) and hematocrit (Hct) level were also increased signi cantly (p≤0.05) (Zorbas et al., 2002) However many studies agreed that P osm was one of the test which is consider as the gold standard to assess hydrational status (Minton and Eberman, 2009), (Armstrong et al., 2012).
The aim of this study is to assess the uid intake and hydration status using several biochemical indices among the low risk postpartum women in Malaysia cultural setting, and to provide empirical evidences on the hydrational status among them to assess the risk of this group in developing VTE. It isa stepping stone in educating and counseling the lower risk women the importance of adequate uid intake during their con nement period as part of post partum care.

Methodology
This study is a prospective longitudinal study over six month period, involving subjects attending maternity wards in Hospital Ampang, Selangor state of Malaysia.The study protocol was approved by the National Malaysian Research Registry (NMRR-16-1361-29016). Inclusion criteria included low risk postpartum women who consented to the study, uncomplicated antenatal records of singleton pregnancy and mothers who delivered via spontaneous vertex delivery. The initial calculated sample size with the power of 80% and 95% con dence interval was 200. These women were recruited on day one post delivery. Their sociodemographic data, and baseline examination including measurements of biochemical analysis re ecting the hydrational status (plasma and urine osmolarity, plasma sodium, urea, creatinine and serum haemotocrit levels) were recorded. Patients were taught on how to selfmeasure their uid input and output throughout their con nement phase in a speci c diary. Subsequent follow up visit to reassess these parameters were made atthethird and sixth week post partum ( Figure 1) . Data were analysed with SPSS version 24.0. Demographic data were analysed descriptively while biochemical parameters were calculated statistically and repeated measures were analysed for signi cance (p<0.05). were working mothers while the remaining were housewives (44.6%). (Table 1) The biochemical indices which re ect hydrational status among 56 postpartum low risk women were also analysed where the level on day 1 postpartum was taken as the baseline (day 1).The mean haematocrit was35.26% (SD 4.09) which was normal (female normal range: 36 to 48%). Median urea level at baseline was 2.10mmol/L(SD 3.00)(normal range: 1.8 to 7.1 mmol/L)andmean creatinine level was 45.6μmol/L(SD 8.82)( normal range: 44 to 97 μmol/L). Themean P osm was 287.73 (SD 5.71) mOsm/kgand the measured urine osmolality, U osm was 377.93 (SD 154.13) mOsm/kg. Calculated mean Urine speci c gravity, U sg was 1.28 (SD 0.55). (Table 2 ) Subsequent measurements were taken and recorded at 3rd week and 6th week post partum. Only 15 respondent were able to complete all three visits with records of uid input an output monitoring as well as completing all the hydrational status biochemical assessment. Results showed that the haematocrit levels were elevated by the 3rd week of postpartum [mean 39.06% (SD 4.40)]and by6th weeks [mean 39.35% (SD 3.55]) with the median of 40.1% (SD 14.00). (Table 2) Mean ureaat3rd weekpostpartum was3.37 mmol/L(SD 1.09) while the median creatinine was59.00 μmol/L(SD 35.00).However these indices increased by 6th week postpartum with mean urea level at3.81 mmol/L(SD 0.97) and the mean creatinine level was; 59.93 μmol/L(SD 9.04).These patients had P osm of 297 mOsm/kg (SD 24.00)at 3rd week postpartum ,while the mean U osm was 652.46 mOsm/kg(SD 170.83) during the same period. The repeated measurement by6 th week post partum revealedmean P osm almost unchanged, 297.33 mOsm/kg (SD3.55). Yet, the calculated mean U osm seems to be lesser at 6 th week, 644.46 mOsm/kg ( SD 241.93). (Table 2) For the 15 women who completed the study, the measured U sg was1.28 (SD 0.55) at baseline and showed an increased pattern in the3rd week to2.18 (SD 0.59) and remained so by 6 th week postpartum, 2.17 (0.78). These women were also analysed on their cumulative uid intake and output through the 3rdand 6th week postpartum. The median cumulative amount for total uid intake at 3 week was much less than at 6 week; 1780ml/day per person as compared to 2130ml/day. The uid output also showed similar pattern, at3rd weeks it was 1083ml/day and increased a bit at 6 th week postpartum to1216ml/day perperson. (Table 2)The measurable output was only through the urine output. Although the patients did document the frequency of bowel opening and breast feeding in the diary, unfortunately they were not able to quantify it to be analysed. (Table 3) Friedman Test was used to compare the differences of haematocrit, P osm , urea and creatinine values between all 3 periods (baseline, 3rd week, 6th week). There was no signi cant changes on haematocrit values (p=0.057).Urine speci c gravity,U sg had signi cant increment (p<0.002 a ) from baseline to 3 rd week and 6 th week. (Table 3) On the other hand, as postpartum period progressed signi cant changes can be seen in P osm (p<0.001), urea (p=0.001) and creatinine (p=0.008) from baseline to 3 rd week and also from baseline to 6 th week postpartum. Post hoc Bonferroni analysis was done for urea level and showed signi cant increase level from baseline to3 rd week (p=0.009) and from baseline to 6 th week (p=0.001). Similarly for creatinine, post hoc Bonferroni analysis showed a signi cant increase level between baseline and week 3 (p=0.004) and between baseline and week 6 (p=0.004).P osm were also analysed and showed a signi cant increase level between baseline and week 3 (p=0.001), and between baseline and week 6 (p=0.001) and signi cant between week 3 and week 6 (p=0.008). (Table 3).

Discussion
Out of 960 women who were screened for the low risk post partum mothersduring the six months duration of study only 56 met the inclusion criteria. Low risk postpartum women in this study included singleton pregnancy with uneventful antenatal history; delivered vaginally and with uncomplicated post delivery. The small sample size in this study is explained by the fact that being in low risk and considering themselves healthy, the majority were reluctant to to commit to a strict uid monitoring as required in this study during the con nement period. They were also keen to undergo their con nement periodback in their hometown under the care of senior or close relatives care outside of the state. Only 15 women were able to complete the whole course of study till the end 6th week postpartum. Attritian rate was as high as 73%, contributed by reluctancy to come for follow up visits since they were considered low risk, with no medical complaints that needed any medical attention. This is very much in unced by the cultural beliefs that since there were no medical problems it is best to complete the con nement phase in bound at home. Few cited transportation di culties as reasons not to continue the study. Thus, due to these limitation, the baseline data comprised 56 patientswhile 15 of them completed all parameters of study measured at 3 different intervals.
The majority of our patients were of Malay ethnicity which re ects the demography composition in Malaysia where 61% of 32 million citizens are Malay (Statistic Department Malaysia). The majority of the Malay mothers still hold on to cultural practices during con nement period which include a strict regime of from food and uid intake, having series of traditional massages and abiding to ample rest that is translated into immobility and forbiddance to go out of home, for at least 40 to 60 days post partum, with the good intention to ensure good recovery period for both mothers and babies (Koon et al., 2005).
Regardless of the fact that the majority of mothers' mean age were within reproductive age and many were with multiple children (multiparous)indicating that they did have experience taking care of babies, the family and society support system usually will still ensure that a senior person take care of the mother and baby during this period. Malaysia is a modern developing country and the majority of women were well educated and have careers, yet, these do not stop them from adhering to the traditional practices as a way for recovery after birth (M, N and E, 2018). The only things that may have changed were that services like massage therapy and con nement nurse have been commercialisedin recent decade with trained masseurs and midwives using battery operated massaging tools as well as modern sauna apparatus. However the principles in restricting food, uid and mobilityare still highly observed.
The biochemical analyses for baseline group at day 1 post partum majority re ects normal values for hydrational status. The haematocrit were normal range as expected 35.26%± 4.09 as these women were not anaemic and there were no post partum haemorrhages post deliveries.Haematocrit is a haematological indices that can resemble the hydrational status with the relation to risk of VTE. There was a study that reveled a multivariable hazard ratios per 5% increment of hematocrit for the total population which adjusted for age, body mass index and smoking were 1.25 (95% CI: 1.08-1.44) for total venous thromboembolism which is a risk factor in a general population (Braekkan et al., 2009).The mean haematocrit in 3rd week and median in 6th weeks increased in as much as 4-5% which may indicate mild dehydation had ensued despite no signi cant history of uid loss recorded. Median was used instead of mean in some of the parameters due to skewness and not normally distributed in such a small sample. ANOVA did notshow anysigni cant increment of haematocrit seenat 3rd and 6th week(p = 0.057).
Blood urea and creatinine indicate the general status of renal function. Both are important in assessing the glomerular lteration rates which in turn is crucial in maintaining body uid and electrolytes. Elevated urea more than creatinine in ratio, may indicate acute dehydration although it needs to be con rmed with clinical correlation.. Increased blood urea and creatinine ratio (mmol/L)/(mg/dl) > 15 (odds ratio [OR] 8.75)is an independent risk factor of developing venous thromboembolism in post acute ischaemic stroke (Kim et al., 2017). However urea is sensitive to dietary intake as it is the end product of protein and to physiological changes as compared to creatinine which is more stable in re ecting the renal function. Though urea is sensitive yet not speci c, it is easily accessible and good in assessing general hydration status and renal function which explains its broad usage.. Our patients who were of low risk postpartum group, as expected showed normal hydration status and normal renal function at day 1 post delivery based on their urea and creatinine levels. However, repeated measurement analysis done at 3rd and 6th week post partum revealed signi cant increment of urea and creatinine level (p = 0.001 b and p = 0.008 b ).
Despite signi cant increment, the calculated ratio increased to not more than15. Thus, this indicated that our group of mothers a degree of less hydrated as compare to day 1 postpartum.
Plasma and urine osmolality are more reliable parameters in assessing hydrational status. They are subject to multiple factors especially when there are changes in the electrolytes and uid intake itself (Armstrong et al., 2013). Yet, our patients who were all from low risk uncomplicated post partum group had normal electrolytes values hence ensuring that the calculated values of both plasma and urine osmolality in this study were unlikely to be in uenced by electrolytes imbalance. Plasma osmolality, P osm for adult generally ranges between285-295 mOsm/kg H 2 O while random urine osmolality, U osm normally ranges between 50-1200 mOsm/kg H 2 O (Tuazon, 2020). A twelveto 14-hour uid restriction generally produces more than 850 mOsm/kg H 2 O (SI units) U osm which suggests dehydration (Wilczynski, 2019).
Plasma and urine osmolality are best compared together to get a better idea upon the hydrational status re ected by the kidney function.In our study ,there was statistically signi cant increment ofP osm and U osm between day 1 to 3rd week and 6th week post partum (Fig. 3). Although the U osm did not reach the dehydrated level > 850 mOsm/kg H 2 O (Wilczynski, 2019), P osm > 296 mOsm/kg is considered a hyperosmolar state and is associated with high risk of mortality among acute stroke (Bhalla et al., 2000).
It further support that these post partum women has been underhydartedduring the con nement phase.
Urine speci c gravity,U sg is commonly used as general tool for hydration status too, as the specimen is easily collected in a simpler clinical setting i.e outpatient clinic. It is known to be sensitive but not speci c.
It has been compared with P osm and U osm for more accurate measurement of hydration among athletes.
Yet, osmolality in both urine and plasma have better speci city in comparison. Nevertheless, our patient did have signi cant increment of U sg measurement > 1.010 which may suggest mild dehydration, and similar with other parameters throughout the duration of 3 weeks of con nement and relatively the same from 3rd week to 6th week.
Study by J Kelly et al has shown that dehydration itself is a strong independent association with VTE in patients with acute post cerebrovascular event. It revealed that signi cant risk of VTE with increament of P osm , urea and also urea creatinine ratio (Kelly et al., 2004).Another study revealed that uid, electrolyte and acid-base disorders had strong signi cant mutivariate analysis (P≪0.001) in risk of developing stroke and intracranial venous thrombosis post partum similar as cesarean delivery and hypertension during both peripartum and postpartum (Lanska and Kryscio, 2000).
Despite having a small cohort of sample, our study consistently revealed that the women with low postpartum risk, with limiting ambulation within their house but otherwise had normal dietry intake has been relatively dehydrated. There is no one speci c biochemical measurement that able to quantify the hydration status accurately due to the dynamic process of the kidney, post partum hormonal recovery and breast feeding among these women. Yet, almost all the indices measured suggested of less hydrationalstate among them. Although these women had been complying to the advice to maintain adequate oral hydration, the calculated amount of uid intake was rather low, slightly more than 1.5L/ day (1.78L/day at 3rd week and 2.13L/day) and the measured output were only 1L to 1.2 L/ day without calculating the insensible water loss ( Fig. 2).The recommended amount of uid intake for lactating mother is 3.1 liters equavalent to 13 cups perday, as compared to non-lactating mothers where the requirements are obviouslyless with 2.2 liters(9 cups perday) and 2.3 liters (10 cups perday)("Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate" at NAP.edu, 2004). Thus; our ndings suggested that our local low risk nursing mothers are most likely under hydrated during the three to six weeks post partum. A study among Mexican women revealed uid intakein between 2.62 L perday to 2.75 L perday during the rst week of con nement. A large study in Indonesia with almost similar Asian cultural background showed that the uid intake among the nursing mothers'during the rst week postpartum were 2.3L to 2.5 L perday. Although both study were studying the total of uid and calorie intake in the rst week post delivery among lactating mother with various socieeconomic background; both seems to have less uid intake then recommended. Both studies did not have any hydration biochemical parameters assessed. Our data showed that the mothers' uid intake was even lower which may predispose them to risk of dehydration and possible venous thromboembolism.

Limitation
It was rather challenging to recruit local women with low risk postpartum to commit with the prospective nature of this study that required series of visits. Despite the small sample size however, there were interesting statistically signi cant changes seen on the hydration parameters up to 3rd and 6th week post partum which correlated with the lower measured amount of uid input. Perhaps a larger cohort of samples, from multicentre data where shorter distance between the home and research centre may provide promising results as it will improve subjects compliance to protocol. A comparison study between a controlled uid intake and non control group may also provide better comparison for analysis.
Calculating the insensible water loss from passing motion and breast feeding may provide more accurate amount of water loss, which our study unable to quantify. Nevertheles this study is a stepping stone for a larger study in the impact of uid intake on the hydration status. It is a good evidence that more measures are required to educate and enhance post partum care and uid intake among post partum women in the local communities.

Conclusion
This study aimed to evaluate the hydration status among low risk post partum women during the six week of con nement period within a Malaysian cultural setting . Despite a small cohort of samples there were statistically signi cant evidence of underhydration among these local women from the biochemical parameters taken namely the urea, creatinine, urine speci c gravity, plasma and urine and osmolality. This was further supported by evidence of low uid intake measured perday as compare to other studies and recommendation for nursing mother during con nement. Altough underhydrated, these women were not dehydrated and had normal electrolytes levels. However, this may predispose the low risk group to possible risk of developing venous thrombosis. Perhaps, a bigger cohort may provide better correlation between hydration and water intake to risk of developing venous thromboembolism. Nevertheless, this evidence is good enough to emphasize upon our low risk group of women to adequately consume uids during con nement a ensure healthier mother, successful breast feeding and reduce risk of venous thrombosis. Availability of data and material

Abbreviations
The datasets generated and analysed during the current study are not publicly available due personal information and policy by the university but are available from the corresponding author on reasonable [online]