Physiological data of rat weight changes
A pregnant female rat was determined when it gave birth on day GD23 or earlier. The data ∆W
6 ~ ∆W
12 of pregnant rats and unpregnant rats were collected. In this way, 34 pregnant rats weight information and 49 unpregnant rats weight data were gathered. Among 76 female rats, 2 rats were confirmed as pregnant through caesarean section on GD20 and then they quit this scientific research, the other 74 rats gave birth naturally on GD23 without disturbance. The pregnancy course or unpregnant course of a female SD rat was recorded just for one time to assure the independency of each datum. In addition to this, the rats engaged in this study were of physical maturity at 12 weeks and older to present the normal condition.
∆
W6 diagnostic model for pregnancy
Both pregnant rats and unpregnant rats ∆W
6 data obeyed a normal distribution (Fig 1A). The mean value and standard de
viation value told the average of weight gains on day 6 of pregnant rats was 30.33 g while 9.38 g in unpregnant rats, which showed a statistical difference (Fig 1B). The area under the ROC curve reached to 0.9407, which demonstrated that the ∆W
6 could be a high-performance diagnostic model to help us recognize the pregnant SD rat on GD6 (Fig 1C). Besides, cut-off value helped us understand that a female rat owned 70% probability of pregnancy on GD6 when its ∆W
6 reached more than 25.50 g (Fig 1D). In other words, we could obtain the 70% power to identify the pregnant rats on GD6 by measuring out W6 and calculating the ∆W
6.
∆
W7 ~ ∆
W12 diagnostic models for pregnancy
Similar with ∆W
6, ∆W
7 ~ ∆W
12 data also followed Gaussian distribution (Fig 2A and 2B). The grouped comparisons by two-way ANOVA were applied to get the p value of each comparison with the 95% confidence interval (CI
95) of differences between means (Fig 2C and 2D) and statistical significances were showed in the differences between pregnant rats and unpregnant rats on means of ∆W
7 ~ ∆W
12 (DW7 32.87 g versus 10.15 g, DW8 34.91 g versus 11.00 g, DW9 36.70 g versus 10.91 g, ∆W10 41.23 g versus 10.09 g, ∆W
11 44.47 g versus 11.21 g, ∆W
12 46.94 g versus 12.37 g, all p<0.0001). The cut-off values of ∆W
7 ~ ∆W
12 on ROC curve were obtained, Specifically, cut-off value 23.00 g of ∆
W7 granted us 83.87% power to identify the pregnant SD rats on its GD7, cut-off value 22.50 g of ∆W
8 owned 93.75% identification power and ∆W
9 22.50 g owned 96.67%, ∆W
10 31.00 g conferred 86.67% power, cut-off value 27.00 g of ∆W
11 conferred 93.33% and ∆W
12 cut-off value 31.00 g gave 93.75%. Collectively, the models of ∆W
7 ~ ∆W
12 derived from pregnant rats could be in advantage of confirming whether the female rat is pregnant or not from GD7 to GD12.
The hypothesis is that pregnant SD rats may grow faster in weight than unpregnant SD rats on account of the embryo development in uterus. In this research, we confirm that there are statistical significances in differences of means of weight gains between pregnant SD rats and unpregnant SD rats from the 6
th day to 12
th day after mating. Subsequently, we get seven cut-off values respectively from ROC curve of ∆W
6 ~ ∆W
12 originated from pregnant SD rats and unpregnant SD rats. Each DW presents its high potential in distinguishing the pregnant rats and unpregnant rats with its satisfying sensitivity and specificity. Based on these findings, we now have a more efficient and convincing tool to help us pick out the pregnant SD rats at an early stage of gestation for further study about SD rats pregnancy characteristics. The calculated Time
2 (T
2, T
2=T
0+∆T/2 or T
2=T
1-∆T/2), another highlight of this research, is deemed as the time rat fertilized. All the 34 pregnant rats in our study gave birth after 22 or 23 days from Time
2, which means Time
2 is well tailored to ascertain the beginning of pregnancy.
In summary, this is the first study to utilize weight changes of female SD rats after mating to identify pregnancy. Findings demonstrate a pregnancy confirmation as early as GD6 and physiological data of weight changes during pregnancy of SD rats. Moreover, our findings indicate that SD female rats’ ∆W
6 ~ ∆W
12 own potential to perform diagnostic models identifying pregnant rats. This work will strengthen our capacity to obtain pregnant SD rat model for further studies.
Limitations of the study
The weight gains data in the research are collected from 34 pregnant SD rats and 49 unpregnant SD rats. Larger sample size is needed to verify the accuracy of cut-off value from current study. Furthermore, the data of pregnant rats in this study are all primiparous. Whether our study result is applicable for multiparous SD rats needs further study.
