Fusar‐Poli et al¹ comprehensively describe a series of developmental steps that lead to psychosis and related psychiatric illnesses. This perspective is especially relevant when considering prenatal intervention. Several pillars of evidence support the notion that the prenatal period is the first developmental step towards psychosis.

The first pillar is the epidemiological ev­idence derived from case‐control studies that point to maternal adversities, such as prenatal starvation and infection, as risk factors for later psychosis. The second pillar is the molecular evidence that a substantial group of genes associated with psychosis is expressed more robustly in the fetal brain before birth than in the brain after birth².

The third pillar is given by post‐mortem findings from psychotic individuals that provide information about the maturation of excitatory and inhibitory neurotransmission that normally occurs during gestation. In the brain of persons with schizophrenia, the normal pre‐term maturation of glutamate receptors – from lower affinity, slower acting NMDA‐type to higher affinity, fast acting kainate type – is incomplete³. The maturation of the chloride transporters that support GABA's inhibitory function from the less effective embryonic NKCC1 to the more effective mature KCCN2 is also incomplete⁴.

These failures in gestational maturation are thus apparently irreversible over an individual's life span. Furthermore, the likely functional result, brains that process information more slowly and less efficiently, is consistent with schizophrenia patients' life‐long deficits, including their well‐documented deficit in processing speed. The ability of neurocognitive and ­pharmacological interventions to ameliorate this deficit in patients after birth is limited. Prevention after birth is important, as outlined by Fusar‐Poli et al, but it is ­hampered by the need to compensate for failures in development before birth that are generally not reversible.

Despite the evidence supporting the po­tential value of prenatal intervention, its effectiveness for the prevention of psychosis is difficult to ascertain, because of its remoteness from the diagnostic emergence of psychosis in adulthood. An obvious remote aspect is temporal. The results of prenatal interventions instituted today cannot be judged until decades later, when the clinicians and investigators who designed and delivered the intervention may be long forgotten. A second remote aspect is the absence of a nosological link of the early childhood effects of prenatal intervention to psychosis. Because few babies in prenatal intervention studies are destined to develop schizophrenia, the outcome for most babies will be more general improvement in their behavior and cognition. No test in early childhood identifies babies who would have developed schizophrenia, had the intervention not occurred.

In our work, as cited in Fusar‐Poli et al's review, we observed a decreased prevalence of a physiological dysfunction in newborns of mothers who received phosphatidylcholine supplements compared to placebo. Phosphatidylcholine is the dietary source of the choline needed to activate fetal alpha 7‐nicotinic acetylcholine receptors. The receptor's gene CHRNA7 is associated with schizophrenia and related psychiatric illnesses, and is expressed more robustly in fetal brain than after birth². Alpha 7‐nicotinic receptor activation is a critical mechanism in the maturation of both glutamate receptors and GABA‐related chloride transporters⁵. The physiological dysfunction, a partial failure in inhibition of the cerebral evoked response to repeated auditory stimuli that occurs in many people with schizophrenia, indicates that these neurotransmitter mechanisms are not functioning optimally⁶. We observed normal inhibition of the evoked response more frequently among the newborns in the phosphatidylcholine‐supplemented group than in the placebo group.

Fusar‐Poli et al note that the positive effects of phosphatidylcholine on ­physiological dysfunctions found in schizophrenia and related mental illness are surrogate markers for prevention of psychosis. As childhood progresses, children whose mothers received phosphatidylcholine supplementation also have positive behavioral effects associated with their positive physiological response as newborns. These children have decreased problems with attention and social withdrawal, compared to children whose mothers received placebo⁶. Conversely, increased problems with attention and social withdrawal are rated retrospectively in children who later have developed schizophrenia, compared to those who did not⁷. Prenatal phosphatidylcholine supplementation appears to help children avoid a developmental pathway that is typical of many individuals who later develop schizophrenia as adults.

Improvement in attention and social function in early childhood is not prevention of mental illness, but neither is it merely a surrogate marker. Children with better attention and social behavior are benefitted in their future success in school, regardless of whether they were destined to become psychotic. Based on the low frequency of psychosis, these more general effects of phosphatidylcholine supplementation may be as important for population well‐being as any specific effect on psychosis. If preventive efforts in psychiatry are exclusively focused on prevention of mental illness, we may overlook broader benefits.

Unique aspects of psychotic illnesses, including the psychotic break in late adolescence, certainly merit investigation. However, individuals who do not convert to psychosis, the majority of patients in prodromal or attenuated symptom status, have problems with attention and other cognitive deficits that are similar to those who do convert to psychosis. These cognitive deficits are disabling regardless of whether an individual becomes psychotic or not⁸. Neuropsychological studies in schizophrenia patients find that attention and learning, rather than psychotic symptoms, are the major contributor to most patients' adverse outcomes. Current genome‐wide association studies, which now identify hundreds of genes in association with schizophrenia as well as with developmental problems, support the thesis that much of the molecular pathology of schizophrenia resides in general brain development that underlies social behavior, attention, and other brain functions.

These clinical and genetic findings suggest a broadened reconceptualization of schizophrenia as a general alteration of neu­rodevelopmental processes, rather than the outcome of a psychosis‐specific pathogenesis. This reconceptualization is congruent with a common characteristic of population‐wide primary prevention: beneficial effects on development that extend broadly beyond a narrow disease target. Folic acid, for example, has positive effects on cognition and behavior, in addition to its targeted use to prevent spina bifida and facial clefts. Vitamin D, included in prenatal vitamins to support bone development, appears to be helpful in the prevention of autism spectrum disorder and schizophrenia. Thus, folic acid, vitamin D, and now choline, along with other primary interventions to protect the uterine environment as part of good obstetrical care, have broad beneficial effects for the offspring, in addition to the possible prevention of later psychiatric illness. An example is the significant protective effects of prenatal choline on the development of attention in offspring of women who contract respiratory viruses in gestation⁹. These findings can provide guidance for treatment of pregnant women in the COVID‐19 pandemic, so that their children might not add another stone to the pillar of evidence link­ing prenatal infection to schizophrenia.

Most beneficial effects will appear in early childhood, long before preventive effects for psychosis and other psychiatric illnesses can be definitely ascertained. If expectant families are to see the benefit of improved childhood behavior and cognition with the eventual possible prevention of psychosis, psychiatry cannot be the only discipline to promulgate these prenatal interventions. Prenatal nutrients such as choline that have early beneficial childhood effects require widespread acceptance by obstetricians and maternal‐fetal medicine specialists, family medicine physicians, midwives and pediatricians. Working relationships with obstetricians for the assessment of perinatal depression is a model for what needs to happen to allow choline and other prenatal primary preventive interventions to become truly population‐wide.

Fusar-Poli等人¹全面描述了导致精神病和相关精神疾病的一系列发展步骤。在考虑进行产前干预时，这种观点尤其重要。几项证据支持以下观点：产前期是迈向精神病的第一步。

第一个支柱是从病例对照研究得出的流行病学证据，这些研究指出孕产妇患病（如产前饥饿和感染）是以后精神病的危险因素。第二个支柱是分子证据，表明与精神病相关的大量基因在出生前的胎儿脑中比在出生后的脑中^2的表达更为强烈。

第三个支柱是精神病患者的事后调查发现，这些发现提供了有关妊娠期通常发生的兴奋性和抑制性神经传递成熟的信息。在精神分裂症患者的大脑中，谷氨酸受体的正常早熟-从较低的亲和力，作用缓慢的NMDA型到较高的亲和力，快速作用的海藻酸盐型-不完全³。支持GABA抑制功能的氯化物转运蛋白从不太有效的胚胎NKCC1到更有效的成熟KCCN2的成熟也是不完全的⁴。

因此，这些妊娠成熟的失败在整个人的一生中显然是不可逆的。此外，可能的功能性结果是大脑处理信息的速度较慢且效率较低，这与精神分裂症患者的终生缺陷（包括他们已证明的加工速度缺陷）一致。神经认知和药理学干预措施可以减轻出生后患者的这种缺陷。正如Fusar-Poli等人所概述的那样，分娩后的预防很重要，但由于需要补偿通常无法逆转的分娩前发育失败，因此妨碍了预防。

尽管有证据支持产前干预的潜在价值，但由于其与成年期精神病诊断性诊断相距甚远，因此难以确定其预防精神病的有效性。一个明显的远程方面是暂时的。直到几十年后，才可以判断今天开始的产前干预措施的结果，那时设计和实施干预措施的临床医生和研究人员可能早已被人们遗忘了。第二个遥远的方面是，在产前干预对精神病的早期儿童影响方面，没有病因学联系。由于在产前干预研究中很少有婴儿注定会发展为精神分裂症，因此大多数婴儿的结局将是行为和认知的更全面改善。

在我们的工作中，正如Fusar-Poli等人的评论所述，与安慰剂相比，我们观察到在接受磷脂酰胆碱补充剂的母亲的新生儿中，生理功能障碍的患病率降低了。磷脂酰胆碱是激活胎儿α7-烟碱乙酰胆碱受体所需的胆碱的饮食来源。受体的基因CHRNA7与精神分裂症和相关的精神疾病有关，在胎儿脑中的表达比出生后更强²。α7烟碱样受体激活是谷氨酸受体和GABA相关氯转运蛋白成熟的关键机制⁵。生理功能障碍，即在许多精神分裂症患者中发生的对反复听觉刺激的脑诱发反应的抑制作用的部分失败，表明这些神经递质机制无法发挥最佳作用⁶。我们观察到，在补充磷脂酰胆碱的组中，新生儿对诱发反应的正常抑制作用比安慰剂组中的抑制作用更为频繁。

Fusar-Poli等人指出，磷脂酰胆碱对精神分裂症和相关精神疾病中发现的生理功能障碍的积极作用是预防精神病的替代指标。随着童年的发展，母亲接受磷脂酰胆碱补充的儿童也具有积极的行为效应，这些效应与他们对新生儿的积极生理反应有关。与母亲接受安慰剂的孩子相比，这些孩子的注意力和社交退缩问题有所减少⁶。相反，与那些没有精神分裂症的儿童相比，后来发展为精神分裂症的儿童的注意力和社交退缩问题增加了回顾性评估[ ^7]。。产前补充磷脂酰胆碱似乎可以帮助儿童避免发育障碍，这是后来成年后发展为精神分裂症的许多人的典型情况。

幼儿注意力和社会功能的改善不是预防精神疾病，也不只是替代指标。注意力和社交行为得到改善的孩子，无论他们是否注定要患精神病，都将从他们将来在学校的成功中受益。由于精神病发生率低，补充磷脂酰胆碱的这些更普遍的作用对人群的健康可能与对精神病的任何特定作用一样重要。如果精神病学的预防工作仅专注于精神疾病的预防，我们可能会忽略更广泛的好处。

精神病的独特方面，包括青春期晚期的精神病发作，无疑值得研究。但是，未转变为精神病的个体，大多数前驱或症状减轻状态的患者存在注意力和其他认知缺陷的问题，与那些转变为精神病的患者相似。无论个人是否患有精神病，这些认知缺陷都会致残⁸。对精神分裂症患者的神经心理学研究发现，注意力和学习而非精神病性症状是导致大多数患者不良结局的主要因素。当前的全基因组关联研究现已鉴定出数百种与精神分裂症以及与发育问题相关的基因，支持以下论点：精神分裂症的许多分子病理学都存在于一般的大脑发育中，而后者是社会行为，注意力和其他大脑的基础职能。

这些临床和遗传发现表明，精神分裂症作为神经发育过程的一般改变而不是精神病特有的发病机制的结果的广泛改变而重新概念化。这种重新概念化与全人群一级预防的共同特征是一致的：对发展的有益影响广泛地超出了狭窄的疾病目标。例如，叶酸除了有针对性地预防脊柱裂和面部裂口外，还对认知和行为有积极作用。产前维生素中支持骨骼发育的维生素D似乎有助于预防自闭症谱系障碍和精神分裂症。因此，叶酸，维生素D和现在的胆碱，作为良好的产科护理的一部分，以及其他保护子宫环境的主要干预措施，除了可能预防以后的精神疾病外，还对后代产生广泛的有益影响。一个例子是产前胆碱对妊娠期感染呼吸道病毒的妇女后代注意力发展的显着保护作用。⁹。这些发现可以为孕妇在COVID-19大流行中的治疗提供指导，从而使他们的孩子可能不会在将产前感染与精神分裂症联系起来的证据基础上再添一石。

最有益的作用将出现在儿童早期，很早就可以确定对精神病和其他精神疾病的预防作用。如果预期家庭希望从改善的儿童行为和认知中受益，并最终可能预防精神病，则精神病学不是颁布这些产前干预措施的唯一学科。产前营养对儿童具有有益的早期作用，例如胆碱，需要得到产科医生和母婴医学专家，家庭医学医生，助产士和儿科医生的广泛接受。与产科医师的工作关系以评估围产期抑郁症是一个模型，它说明了需要发生什么才能使胆碱和其他产前主要预防干预措施真正成为整个人群的模型。