Despite advances in increasing survival rate, preterm birth is still the main cause of neonatal morbidity [1], representing one of the most leading risk factors for neurodevelopmental disabilities during childhood [2]. Critically, the lack of accurate prediction methods of preterm birth in TPL women is a matter of concern due to the potential iatrogenic effects of repeated antenatal corticosteroid on the future child’s neurodevelopment [3, 4], stressful and unnecessary hospitalizations, and elevated costs for public health system [5, 6]. One-third of hospitalized pregnant women suffer from a threatened preterm labor (TPL), but more than 50% do not progress to active labor [7] and only about half of preterm births are preceded by a known risk factor [8]. Different prevention programs have been implemented around the world aimed to reduce preterm birth [9] and social determinants are gaining consideration from scientific organizations [10]. Therefore, reliable methods to stratify TPL women into low and high-risk groups for preterm birth outcome are required.
A growing body of research has indicated that both chronic life stress prior to conception and stressful events during pregnancy may act as potential risk factors for preterm birth [11, 12]. When coping mechanisms are saturated due to chronic life stress, overexposure to neuroendocrine mediators (e.g., cortisol or α-amylase) that maintain the homeostasis of Hypothalamic-Pituitary-Adrenal (HPA) axis [13, 14] and the Sympathetic-Adrenal-Medullary (SAM) axis [14] may have a deleterious impact on both mother and fetus [15], increasing the risk of preterm birth. Among possible causes of chronic stress, a history of traumatic events and poor social or family functioning have usually been identified. In fact, it is well-documented that a history of traumatic life events prior to conception may increase the risk of preterm birth [16–20]. However, although social support may modulate the association between life stressful events and preterm birth, findings are inconclusive [12, 21–23]. Whereas a systematic review concluded null relationship between maternal social support and preterm birth [22], more recent studies pointed out that the lack of partner support rather than lack of general social support was associated with higher risk of preterm birth [21, 23, 24].
Regarding stressful events during pregnancy, they may alter normal balance of immune mediators, hormones, and neurotransmitters involved in timing of birth, increasing the risk of preterm birth [25, 26] as well as psychomotor impairments [27]. Noteworthy, TPL is considered a stressful prenatal event likely to trigger a biopsychological stress response [28a, 28b]. First, from a biological perspective, TPL event may dysregulate both the HPA axis [13, 14] and the SAM axis [14]. As for HPA axis biomarkers, research has revealed that cortisol levels at TPL diagnosis may predict birth 48 hours after TPL diagnosis [29]. Conversely, other study addressing SAM activity measured by α-amylase levels has showed inconclusive findings [14]. Whereas α-amylase dysregulation has been suggested as the underlying mechanism for the link between maternal depression and prematurity [30], no association between α-amylase levels and preterm birth has been found in non-depressed women [28a, 28b]. Second, from a psychological perspective, both gestational anxiety [31, 32] and depressive symptoms [33–35] which may be triggered by TPL diagnosis [36] can also be associated with preterm birth. In sum, women experiencing chronic life stress may need only another significant stressor during pregnancy such as TPL to reach the tipping point that leads to a preterm birth [31].
In spite of focused research efforts, it is still unclear which stress-related factors are most strongly associated with preterm birth, for several reasons. Firstly, although the impact of a stressful event during pregnancy can be modulated by a combination of biopsychosocial stress-related pathways (HPA or SAM stress biomarkers, anxious-depressive symptoms at TPL diagnosis, and/or previous traumatic events as well as social support), most studies have considered these pathways separately [37]. Secondly, the relationship between these biomarkers and self-reported psychosocial stress is not straightforward [38]. Moreover, the few studies that have simultaneously examined different stress-related outcomes have included asymptomatic pregnant women (i.e., without TPL), reporting inconclusive findings: whereas some studies not using self-reports found an association between preterm birth and stress biomarkers [39–40], others found that maternal self-reports may improve the biomarkers prediction [41]. Thirdly, prospective studies with symptomatic women usually conduct a follow-up until 48 hours after a TPL diagnosis instead of until birth [29]. In the present follow-up study, multiple stress-related outcomes are studied simultaneously in symptomatic women from TPL diagnosis to birth date.
This study aims to predict the birth date in TPL women by means of a combination of multiple stress-related factors: (i) cumulated life stressors (previous traumas, social support, and family functioning); and (ii) biopsychological response to TPL diagnosis (salivary cortisol and α-amylase as well as anxiety and depression symptoms). We expect that, considering studies that examine a combination of stress-related variables, biomarkers would be the strongest birth date predictors [39–40]. However, self-reports assessing chronic social stress and psychological stress response to TPL diagnosis may improve this prediction [41]. Also, the association between biomarkers and self-reports should not be constrained to be linear [38].