Étiopathogénie des délires de grossesse à partir d’une revue de la littérature : rôle de l’hyperprolactinémie et application de la théorie de l’inférence abductive

Abstract

Delusions of pregnancy are not well known. The delusion of pregnancy is defined as the belief of being pregnant despite factual evidence to the contrary. The clinical picture is heterogeneous (duration, mechanisms, topics and pre-existing psychiatric disorders). Several causes have been proposed to explain the occurrence of the delusions of pregnancy: cenesthetic theory, hyperprolactinemia, polydipsia and psychodynamic conflicts. Hyperprolactinemia is an interesting hypothesis (physiologic increase during pregnancy and similar manifestations in the course of gestation). The abductive inference theory is a probabilistic model that can clarify the role of hyperprolactinemia in the delusions of pregnancy. The purpose of this paper is to study the role of hyperprolactinemia in the delusions of pregnancy using a literature review. The abductive inference model is used to specify the etiopathogeny of this pathology. A research in Medline, Sudoc, BIUM and PSYLINK using the following key words "delusional pregnancy" or "delusion of pregnancy" and "hyperprolactinemia" was conducted. Three articles (case reports) about delusions of pregnancy associated with hyperprolactinemia were found. The cases have some similitudes. First of all, they have similar chronology: delusion appears at the same time as hyperprolactinemia and resolves with biological normalization. Secondly, hyperprolactinemia is always caused by a neuroleptic (haloperidol, olanzapine, risperidone). Concerning pre-existing disorders, a psychiatric pathology for each case was found (schizophrenia, schizo-affective disorder and bipolar disorder). Chronology, reproductivity and reversibility are strong arguments to involve hyperprolactinemia in the delusions of pregnancy (Bradford Hill criteria). Furthermore, this association is biologically plausible: physiologic increase during pregnancy (gestational signal), similar symptoms to those during pregnancy and the role in parental behavior (parental signal). Nevertheless, not everyone with hyperprolactinemia will develop a delusion of pregnancy; the interaction is more complex (non linear); the theory of abductive inference clarifies this relationship. Abductive inference is a probabilistic model whose goal is to explain the occurrence of delusional beliefs. The first factor is the abnormal data. The second factor is the cognitive process (abductive inference), which uses Bayes' theorem to select the most likely hypothesis to explain the abnormal data. In the delusion of pregnancy, abnormal data is the hyperprolactinemia, signal of gestation without pregnancy. Hypotheses in order to explain this signal are then produced (pregnancy or no pregnancy). In the second part, probabilities associated with each hypothesis, given the hyperprolactinemia, are compared. Since hyperprolactinemia is a gestational signal, the pregnancy hypothesis is most likely. Probabilities associated with each hypothesis without taking hyperprolactinemia into account are compared (prior probability). Since any element of reality indicates a pregnancy, the absence of pregnancy is most likely. In the last step, the posterior probability is calculated using the first two comparisons. The probability associated with the pregnancy hypothesis (taking into account hyperprolactinemia) is relatively higher than the probability associated with the no-pregnancy hypothesis (without taking into account hyperprolactinemia). So, the posterior probability associated with the pregnancy hypothesis is more likely than the posterior probability associated with the no-pregnancy hypothesis. Thus, the subject believes in a pregnancy. The research and the treatment of hyperprolactinemia must be conducted when faced with a delusion of pregnancy.