Journal Pre-proof Low-dose aspirin use in pregnancy and the risk of preterm birth: a Swedish register-based cohort study.

46 Background : Preterm birth is the leading cause of neonatal mortality and morbidity. Women 47 who have had a previous preterm birth are at increased risk of preterm birth in their subsequent 48 pregnancies. Low-dose aspirin reduces the risk of preterm birth among women at risk of 49 developing preeclampsia, however it is unclear whether low-dose aspirin may reduce the risk 50 of recurrent preterm birth. 51 Objectives : The purpose of this study was to investigate the association between low-dose 52 aspirin and preterm birth among women with a previous preterm birth. 53 Study Design : We conducted a Swedish register-based cohort study and included women who 54 had a first and second pregnancy between 2006 and 2019, where the first pregnancy resulted in 55 preterm birth (medically indicated or with spontaneous onset <37 weeks of gestation). The 56 association between low-dose aspirin use and preterm birth in the second pregnancy was 57 estimated via logistic regression via standardization and expressed as marginal relative risks 58 (mRR) with 95% confidence interval (CI). 59 Results: Among the study cohort (N=22,127), 3057 women (14%) were prescribed low-dose 60 aspirin in their second pregnancy and 3703 women (17%) gave birth preterm. Low-dose aspirin 61 use was associated with a reduced risk of preterm birth, (mRR 0.87 95% CI 0.77-0.99). There 62 were no statistically significant associations between low-dose aspirin use and an altered risk of moderate preterm birth, defined as birth 32-36 weeks’ gestation (mRR 0.90 95% CI 0.78- 64 1.03) or very preterm birth, defined as birth <32 weeks’ gestation (mRR 0.75 95% CI 0.54- 1.04). Regarding onset of preterm birth, low-dose aspirin use was associated with a reduced risk of spontaneous preterm birth (mRR 0.70 95% CI 0.57-0.86) but no reduced risk of medically indicated preterm birth (mRR 1.09 pregnancy, low dose aspirin use was associated with a reduced risk of spontaneous preterm 71 birth. Our results suggest that low-dose aspirin may be an effective prophylaxis for recurrent 72 preterm birth.


Word count 26 3144 27
Condensation: Low-dose aspirin reduces the risk of recurrent preterm birth. 28 Short title: Low-dose aspirin and recurrent preterm birth 29 AJOG at a glance 30 Why was this study conducted? 31 • Women who have had a previous preterm birth (medically indicated or spontaneous 32 onset) are at increased risk of preterm birth in their subsequent pregnancies. 33 • Low-dose aspirin reduces the risk of preterm birth among women at risk of developing 34 preeclampsia, but the effect on preterm birth among women with a previous preterm 35 birth is unknown. 36 B. What are the key findings? 37 • Low-dose aspirin use was associated with a reduced risk of preterm birth (any onset) 38 and preterm birth with a spontaneous onset. 39 C. What does this study add to what is already known? 40 who have had a previous preterm birth are at increased risk of preterm birth in their subsequent 48 pregnancies. Low-dose aspirin reduces the risk of preterm birth among women at risk of 49 developing preeclampsia, however it is unclear whether low-dose aspirin may reduce the risk 50 of recurrent preterm birth. 51 Objectives: The purpose of this study was to investigate the association between low-dose 52 aspirin and preterm birth among women with a previous preterm birth. 53 Study Design: We conducted a Swedish register-based cohort study and included women who 54 had a first and second pregnancy between 2006 and 2019, where the first pregnancy resulted in 55 preterm birth (medically indicated or with spontaneous onset <37 weeks of gestation). The 56 association between low-dose aspirin use and preterm birth in the second pregnancy was 57 estimated via logistic regression via standardization and expressed as marginal relative risks 58 (mRR) with 95% confidence interval (CI). 59 Results: Among the study cohort (N=22,127), 3057 women (14%) were prescribed low-dose 60 aspirin in their second pregnancy and 3703 women (17%) gave birth preterm. Low-dose aspirin 61 use was associated with a reduced risk of preterm birth, (mRR 0.87 95% CI 0.77-0.99). There 62 were no statistically significant associations between low-dose aspirin use and an altered risk 63 of moderate preterm birth, defined as birth 32-36 weeks' gestation (mRR 0.90 95% CI 0.78-64 Low-dose aspirin has been shown to reduce the risk of preeclampsia, a pregnancy condition 99 characterized by hypertension and organ injury. In addition, low-dose aspirin has been shown 100 to protect against preterm birth among women at risk of developing preeclampsia 5 . There is 101 also a growing body of evidence suggesting that low-dose aspirin use could be associated with 102 a reduced risk of preterm birth and in particular spontaneous preterm birth in women without 103 major risk factors for preeclampsia 6-8 . Still, there is insufficient evidence regarding the use of 104 low-dose aspirin in pregnant women with a previous preterm birth. The APRIL study, a 105 randomized controlled trial with 406 participants, reported a small but non-significant reduction 106 of preterm birth in women using low-dose aspirin with a previous spontaneous preterm birth 9 . 107 However, the study was only powered to detect a difference in preterm birth greater than 35% 108 between groups and the included population had a lower-than-expected preterm birth rate. A 109 larger study is needed to investigate whether low-dose aspirin can prevent recurrent preterm 110 birth 10 . Thus, we undertook a population-based study, including over 22 000 women with a 111 previous preterm birth, to investigate whether low-dose aspirin was associated with an altered 112 risk of preterm birth occurring before <37 weeks' gestation, moderate preterm birth 32-36 113 weeks' gestation, and very preterm birth <32 weeks' gestation, and the association between 114 low-dose aspirin use and the mode of onset of preterm birth. only women presenting with major risk factors for preeclampsia were prescribed aspirin at a 169 dosage of 75 mg/day. In a few hospitals in the country, the recommended dosage was 160 mg. 170 Low-dose aspirin is only available by prescription in Sweden and aspirin in higher doses that 171 are available over the counter are not recommended during pregnancy. 172

173
Outcomes 174 The primary outcome was preterm birth in the second pregnancy. Gestational age was based on 175 a first or early second-trimester ultrasound in 93% of the pregnancies. In the remaining 176 pregnancies, gestation age was based on date of the last menstrual period reported at the first 177 antenatal visit, of embryo transfer or a postnatal assessment (<1%) 12 . We further studied 178 preterm birth by severity and onset. Severity was categorized into moderate preterm birth (birth 179 between 36 -32 weeks' gestation) and severe preterm birth (birth <32 weeks' gestation). We analyzed rates of preterm birth in the second pregnancy among women with a previous 219 spontaneous preterm birth and a medically indicated preterm birth separately, adjusting for 220 the same covariates as above, except for onset of labor in the first pregnancy. 221

Sample characteristics 224
The study population consisted of 22 127 women who experienced a preterm birth in their first 225 pregnancy, of which 3057 women (14%) were prescribed low-dose aspirin in their second 226 pregnancy. In total, 3703 women (17 %) had a recurrent preterm birth and 547 of them (15%) 227 used low-dose aspirin during the second pregnancy. The population is presented in a flow chart 228 (Figure). 229

230
The women who used low-dose aspirin where in average 31.5 years old and had a BMI of 25.9, 231 compared to an average age of 30.6 years and an average BMI of 25.0 in the women who did 232 not use aspirin. Pregestational disorders and preeclampsia were common among women using 233 aspirin, 11.1 % of the women had at least one pregestational disorder in the second pregnancy 234 and 16.7 % were diagnosed with preeclampsia in the second pregnancy (Table 1). 235 236

Preterm birth by onset 252
Compared with women not using low-dose aspirin, women using low-dose aspirin had an 253 increased crude relative risk of 3.46 (95% CI 2.92-4.10) for medically indicated preterm birth. 254 However, this risk was reduced and no longer statistically significant after adjusting for 255 confounders (mRR 1.09 95% CI 0.91-1.30). Low-dose aspirin use was associated with a 256 reduced crude relative risk for spontaneous preterm birth of 0.42 (95% CI 0.34-0.53) that 257 remained after adjusting for confounders (mRR 0.70 95% CI 0.57-0.86) ( Table 2). 258 259 There was no statistically significant association between low-dose aspirin and onset of labor 260 among women with a first medically indicated preterm birth (Supplemental table 1

Principal findings 272
In this register-based cohort study on women with a previous preterm birth we found an 273 association between low-dose aspirin use and preterm birth occurring <37 weeks' gestation, but 274 no statistically significant association between low-dose aspirin use and moderate preterm birth 275 32-36 weeks' gestation or very preterm birth <32 weeks' gestation. When separating the risk 276 of spontaneous and medically indicated preterm births in the second pregnancy, we found that 277 low-dose aspirin use was associated with a reduced risk of spontaneous preterm birth <37 278 weeks' gestation. 279 280

Results in the context of what is known 281
The incidence of recurrent preterm birth in our study is lower than generally reported 2 , but in 282 line with the incidence of recurrent preterm birth in other Nordic countries 3,4 . 283 The ASPIRIN trial investigated low-dose aspirin treatment for the prevention of preterm birth 284 in healthy nulliparous pregnant women and reported that low-dose aspirin reduced the risk of 285 preterm birth <37 weeks' gestation (aRR 0.89 95% CI: 0.81-0.98) and preterm birth <34 weeks' 286 gestation (aRR 0.75 95% CI 0.61-0.93) 6 . Our point estimate for preterm birth <37 weeks' 287 gestation is similar although our study population had a higher incidence of preterm birth, which 288 may be attributed to our study population that included women with a previous preterm birth. 289 We did not find a statistically significant association between low-dose aspirin use and 290 moderate or severe preterm birth, but compared to a randomized controlled trial, it is less certain 291 that women had an actual aspirin intake in our study. 292 293 A secondary analysis of a randomized controlled trial has found a reduction of spontaneous 294 preterm birth <34 weeks' gestation (adjusted odds ratio 0.46 95% CI 0.23-0.89) but not <37 295 weeks' gestation in healthy nulliparous pregnant women using low-dose aspirin 7 , whereas a 296 meta-analysis reported a reduction of spontaneous preterm birth for both <34 weeks' gestation 297 (RR 0.86 95% CI 0.76-0.99) and <37 weeks' gestation (RR 0.93 95% CI 0.86-0.996) for low-298 dose aspirin treatment in women at high risk of developing preeclampsia 8 . 299 300 Similar to the APRIL study, which found no statistically significant association between low-301 dose aspirin and preterm birth among women with a previous spontaneus preterm birth (RR 302 0.83 95% CI 0.58-1.20) 9 , our subgroup analysis of the risk of recurrent spontaneous preterm 303 birth showed no significant association (mRR 0.75 95% CI 0.57-1.00). However, a small 304 protective effect cannot be ruled out and we believe that our findings could support the 305 hypothesis that low-dose aspirin might decrease the risk of recurrent spontaneous preterm birth, 306 even though we did not reach statistical significance. In our primary analysis, including women 307 with a first medically indicated or spontaneous preterm birth, low-dose aspirin reduced the risk 308 of both preterm birth and spontaneous preterm birth in the second pregnancy, which could be 309 attributed to a larger study population. 310

Clinical implications 312
Our data suggest that low-dose aspirin may be an effective prophylaxis for recurrent preterm 313 birth with an effect size similar to the preventative effect of low-dose aspirin on the 314 development of preeclampsia in women at high risk 5 . Though, low-dose aspirin has potential 315 side effects and has been associated with an increased risk of intrapartum and postpartum 316 hemorrhage 19 . This warrants caution and treatment of the right target population at high risk of 317 developing the condition at the lowest effective dose is important 5,20 . 318 319 320 J o u r n a l P r e -p r o o f

Research implications 321
A larger randomized controlled trial or an individual patient data meta-analysis of smaller 322 randomized controlled trials are warranted to confirm our findings of a protective effect of low-323 dose aspirin on the risk of recurrent preterm birth. In Sweden the recommended low-dose 324 aspirin dose during pregnancy is 75mg per day and the effect of higher doses, such as 150mg, 325 needs to be further investigated. 326 327

Strengths and limitations 328
The main strengths of our study are the population-based setting that facilitates generalizability 329 to the target population, and the large study sample. We utilized data from national registers 330 that capture data of >98% of the Swedish population 12,13 and contains information on maternal 331 characteristics, socio-demographic factors and pregnancy variables prospectively collected to 332 the outcome, and available for adjustments in the models. 333 334 Our study is limited by the baseline differences between women using low-dose aspirin and 335 those not, which we attempted to overcome in our adjusted modelling by adjusting for 336 gestational length first pregnancy and other covariates that affect the risk of preterm birth. Still, 337 there is a risk that our model might under-or overestimate the preventive effect of low-dose 338 aspirin due to residual confounding. To overcome this, we performed subgroup analyses by 339 spontaneous and medically indicated preterm births in the second pregnancy that showed an 340 association between low-dose aspirin use and a reduced risk of spontaneous preterm birth. 341 However, the low-dose aspirin group had a higher incidence of preeclampsia in the first 342 pregnancy and therefore a higher risk of preeclampsia and medically indicated preterm birth in 343 the second pregnancy. It is also possible that the women in the low-dose aspirin group were 344 more closely monitored and therefore had a medically indicated preterm birth in cases that 345 otherwise would have resulted in a spontaneous preterm birth. Hence, our model might have 346 overestimated the preventive effect of low-dose aspirin on spontaneous preterm birth. Although 347 our data on low-dose aspirin use is based on dispensed prescriptions, since low-dose aspirin is 348 not available over the counter in Sweden, we do not know the compliance and for which time-349 period in the pregnancy low-dose aspirin was used. 350 351

Conclusion 352
In this population register-based cohort study, we found that among women with a previous 353 preterm birth, low-dose aspirin use reduced the risk of recurrent preterm birth (any onset) and 354 specifically spontaneous onset. The use of low-dose aspirin in pregnant women with a previous 355 preterm birth could reduce the global burden of preterm birth, but further research is needed to 356 confirm our findings and investigate the optimal low-dose aspirin dosage and timing of use.     n=5925 and n=5474 included in the crude and adjusted models, respectively. The primary outcome was estimated in a binomial logistic regression and secondary outcomes in multinomial logistic regression.