Objective
Women with twin pregnancies may have higher rates of isolated proteinuria than do those with singletons. We compared protein-to-creatinine (P:C) ratios longitudinally through gestation in uncomplicated twin and singleton pregnancies.
Study Design
P:C ratios were compared at 3 times points in 102 (51 twins, 51 singletons) healthy gravid patients who did not have preeclampsia develop, using linear and logistic regression techniques.
Results
P:C ratio increased significantly over gestation in all patients. This increase was significantly greater in twins than in singletons. The odds of P:C ratio >0.19 was 3.5 times higher in twins between 34 and 38 weeks.
Conclusion
Women with uncomplicated twin pregnancies have greater protein excretion as measured by P:C ratios than do those with singletons. In early pregnancy, protein excretion is similar, but it diverges significantly by the latter third trimester. We suggest that normal values for proteinuria in twins may differ from those in singletons, and warrant further evaluation.
Understanding normal protein excretion in pregnancy is essential in making pathologic diagnoses of renal dysfunction and preeclampsia. Many studies have evaluated protein excretion in both normal and preeclamptic pregnancies, and by consensus, physiologic proteinuria is defined as less than 300 mg of protein in 24 hours. Normal protein excretion in twin pregnancies is assumed to be identical to that in singleton gestations, but no study has evaluated twins independently, and most do not comment on whether their samples included multiple gestations. In addition, there is little understanding of how protein excretion changes over the course of uncomplicated pregnancy in either twins or singletons.
We have observed that patients with multiple gestations appear to exhibit proteinuria in the absence of hypertension more often than do singletons. We therefore sought to better understand protein excretion in uncomplicated twin pregnancies. Our primary aim was to compare protein with creatinine ratios in healthy twin and singleton pregnancies that did not go on to develop preeclampia, to assess whether norms for protein excretion in twins should be reevaluated. The secondary goal was to evaluate the change in protein excretion over the course of both twin and singleton pregnancies.
Materials and Methods
After institutional review board approval (2006P000395, 2009P000798), the Brigham and Women’s Hospital enrolled 1608 women, including 1444 with singletons, 107 with twins, and 57 with triplets, between October 2006 and August 2008, as part of a multicenter prospective cohort designed to evaluate the use of serum angiogenic markers for the early diagnosis of preeclampsia, in collaboration with Abbott Diagnostics. Women were eligible for inclusion if they were at least 18 years old, had a pregnancy with estimated gestational age of ≤15 weeks, were able to understand the study requirements and give informed consent. An ultrasound was performed before 20 weeks as part of standard clinical care to confirm the number of fetuses and pregnancy dating. Blood and urine samples were collected at 4 study visits: enrollment (<15 weeks), 16-20 weeks, 24-28 weeks, and 34-38 weeks. Samples were held for no more than 24 hours before processing and storage at −60° C. Maternal characteristics were collected at the first visit. Blood pressure, urinary protein, height, and weight were recorded at each study visit, as well as immediately postpartum. Additional outcomes were collected and documented as part of routine clinical care.
There were 51 twin pregnancies with 34-38 week samples who did not have medical comorbidities, gestational diabetes, intrauterine growth restriction or history of premature delivery, and that did not go on to develop preeclampsia. These were compared with 51 singletons that had been randomly chosen using “proc surveyselect” in SAS version 9.1, from the 956 singleton pregnancies that met the same criteria.
Urine protein/creatinine (P:C) ratio was calculated for each of 3 time points: 16-20 weeks, 24-28 weeks, and 34-38 weeks. Samples from the enrollment time point were not run as our interest was in the latter half of pregnancy, when most complications appear. All patients had 34-38 week samples, but some did not provide a sample at earlier time points, or their sample was insufficient to do analyses. This may lead to less precise estimates at each time point, but longitudinal regression is adequately robust to handle these missing data. All available data were included. Urine protein quantification was determined by the biuret reaction, and urinary creatinine was determined by the modified Jaffe reaction using an Olympus AU640 analyzer (Olympus America, Center Valley, PA). Protein excretion was compared in 2 ways, to determine whether mean P:C ratios differed between singletons and twins, and whether high ratios were more common in twins than in singletons. We defined a clinically relevant high ratio as 0.19.
P:C ratio was confirmed to be normally distributed. Demographic and health characteristics were assessed using χ 2 tests and Fisher’s exact tests where appropriate. Univariate analyses were conducted to examine mean P:C ratio by fetal number using t tests, and the distribution of high P:C ratio was examined in contingency tables with χ 2 tests.
To test the hypotheses on the associations between fetal number and P:C ratio over the course of pregnancy, linear models were fitted using “proc mixed” in SAS version 9.1 (SAS Institute, Cary, NC). The covariance for each participant at different follow-ups was modeled by a compound symmetry structure. In addition to examining them independently, the interaction term between fetal number and follow-up time was evaluated to determine whether twin status had a differing impact on P:C ratio at different time points. The interaction was not statistically significant, therefore only the independent effects were included in the final model. Variables considered potential confounders from the univariate analyses were included in the multivariate regression models. Models were selected using backwards elimination, and variables were removed at the P = .3 significance level, or if their elimination did not change the coefficient between fetal number and P:C ratio by 10% or more. Longitudinal multivariate linear regression models were used to evaluate the effect of fetal number on proteinuria levels, with estimates and standard deviations reported.
P:C ratio was also evaluated as a dichotomous variable, with a high ratio defined as >0.19. Logistic regressions were performed using “proc logistic,” and model selection was performed using the same parameters as the linear analysis. An appropriate longitudinal logistic regression model could not be fitted, so separate models were evaluated at each time point. Unadjusted odds ratios are reported, stratified by gestational age.
Results
As seen in Table 1 , women with twin pregnancies were, on average, older than those with singletons. Twins were more likely to be white and had a greater increase in weight from enrollment to 16-20 weeks. Gestational age at sample collection was significantly earlier in twins by an average of 0.82 weeks, and they were more likely to deliver before 37 weeks. A total of 80.4% of twin pregnancies were the result of assisted reproductive technology (ART), whereas only 3.9% of singleton pregnancies were. Groups were similar in body mass index and smoking status.
| Demographic | Singletons | Twins | P value |
|---|---|---|---|
| Maternal age, mean (SD) | 31.29 (4.65) | 35.29 (6.01) | .0003 |
| White, n (%) | 29 (64.44) | 39 (86.67) | .0100 |
| Hispanic, n (%) | 6 (11.76) | 6 (11.76) | NS |
| Ever smoker, n (%) | 12 (25.53) | 14 (29.79) | NS |
| Gestational age at collection, mean (SD) | |||
| 16-20 wk | 18.42 (1.57) | 17.75 (1.60) | .0449 |
| 24-28 wk | 26.36 (1.64) | 25.43 (1.29) | .0023 |
| 34-38 wk | 35.64 (1.55) | 34.79 (0.92) | .0010 |
| Assisted reproductive therapy, n (%) | 2 (3.92) | 41 (80.39) | < .0001 |
| Body mass index at baseline, mean (SD) | 25.87 (5.43) | 25.36 (6.17) | NS |
| Weight, kg, mean (SD) | |||
| 16-20 wk | 70.53 (16.26) | 73.13 (16.01) | NS |
| 24-28 wk | 75.89 (15.87) | 79.85 (17.43) | NS |
| 34-38 wk | 80.90 (15.20) | 85.75 (17.48) | NS |
| Weight change (kg), mean (SD) | |||
| Intake to 16-20 wk | 2.80 (2.10) | 3.83 (2.39) | .0307 |
| 16-20 wk to 24-28 wk | 4.16 (3.49) | 5.34 (3.24) | NS |
| 24-28 wk to 34-38 wk | 4.71 (4.80) | 5.82 (2.75) | NS |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
