Attention-Deficit/Hyperactivity Disorder (ADHD) is a neurodevelopmental disorder involving impaired attention, hyperactivity and impulsive behaviour; the condition affects over 3% of children worldwide.¹ ADHD not only poses potential life-threatening risks for individuals but also exerts a significant economic burden on society—estimated to cost $31.6 billion annually in the United States.² Identifying modifiable factors in the aetiology of neurodevelopmental disorders and psychiatric illness is paramount for risk prediction and the design of preventive interventions. Although our understanding of environmental origins for the development of ADHD is incomplete, it is recognised that many aetiologic determinants operate during foetal life and infancy, the earliest and most vulnerable stages of brain development. Given the absence of a cure for ADHD, identifying modifiable risk factors for ADHD remains critical.

In this issue of Paediatric and Perinatal Epidemiology, Nielsen and colleagues³ explore the role of multiple maternal exposures related to inflammation during pregnancy with ADHD in the offspring. The study was based on 908,770 children born from July 2001 to December 2011 in New South Wales, Australia, and followed up until December 2014. ADHD was identified in 16,297 children (incidence 3.5 per 1000 person-years) at a median age of 7 years at first treatment. Seven chronic maternal complications and risk factors (exposures) were identified from birth data and hospital admissions during pregnancy: autoimmune disease, asthma, hospitalisation for infection, mood or anxiety disorder, smoking, hypertension and diabetes. They found that each exposure was independently associated with an increased risk for ADHD (range of hazard ratios [HR] 1.19 to 1.89); the risks increased with cumulative exposures to inflammatory markers, suggesting dose–response relationships. Children born to women with asthma, infection, mood or anxiety disorder and smoking carried the greatest burden of ADHD risk (HR 6.12, 95% confidence interval [CI] 3.47, 10.70).

This study is unique in many ways. It provides a population-based characterisation of ADHD risks by maternal exposures to inflammation-related conditions; an examination of single and multiple conditions and, importantly, cumulative effects through a dose–response relationship. It also offers much-needed clarity in understanding how maternal inflammation may affect ADHD in children. In this commentary, we highlight how this study advances the field and points to areas where nuanced data are lacking but may provide important insights into understanding the role of inflammation in ADHD risk in children.

The findings in the study add to the accumulating evidence that shows increased risks of neurodevelopmental disorders in relation to maternal obesity, cardiovascular disease, chorioamnionitis and epilepsy.^4-6 These risks could be either caused by the chronic disease itself or by medication used to treat the chronic disease. The pathways that connect maternal illness and neurodevelopmental outcomes are complex and, for the most part, largely remain unknown; among them, maternal factors play an important role in shaping ADHD risks. Nielsen and colleagues³ emphasise that the maternal exposures considered are heterogeneous and several conditions have important non-inflammatory aspects that may influence the observed association.

It is important to acknowledge a limitation inherent in the heterogeneity of the maternal conditions studied, each possessing distinct biological pathways leading to offspring ADHD. While the study demonstrates an increased risk of ADHD with an escalating number of maternal exposures, the complexity of each condition poses a challenge in formulating a unified intervention strategy. For instance, with autoimmune disease, it is imperative to closely monitor and manage the condition and medication use during pregnancy. Conversely, addressing smoking habits requires targeted behavioural interventions. Optimal blood glucose monitoring is crucial for maternal diabetes management,⁷ while interventions for hypertension during pregnancy may include daily low-dose aspirin, antihypertensive medications, foetal monitoring and timed delivery.

This study begins to uncover the cumulative effects of maternal chronic conditions on ADHD risk. However, it prompts critical questions about how these conditions should be approached: as a constellation of interconnected conditions or as individual entities. Does smoking fall in the same group as a chronic condition? While extended periods of regular smoking qualify as chronic exposure, we argue against the inclusion of smoking as a chronic maternal illness. Instead, we regard smoking as a potential risk factor for ADHD. Recent studies have documented that the previously thought maternal smoking-ADHD risk in children is best described as an association that is strongly underscored by unmeasured confounding,⁸ and residual confounding. For instance, in this study, residual confounding by socioeconomic status (which is associated with inflammatory conditions in general) could potentially bias the associations. Although the authors adjusted for area-level average income as a proxy for the socio-economic status, the lack of adjustment for the specific markers of socio-economic status may have resulted in residual confounding by socio-economic status.

In our view, despite the shared goal of preventing ADHD, these differing approaches highlight the need for tailored strategies based on the unique characteristics of each maternal condition.

Future investigations may benefit from exploring each maternal chronic condition individually (i.e. diabetes, obesity, smoking) to disentangle the specific mechanisms underlying these associations to better understand the pathways through which maternal conditions may be causally associated with ADHD in children. This approach could provide insights into the causal pathways, facilitating the development of targeted interventions.

A second issue relates to how some pre-existing conditions and pregnancy complications are associated with preterm birth or SGA. Nielsen and colleagues³ found that children exposed to increasing maternal exposures were more likely to be born preterm, experiencing severe neonatal morbidity. This suggests that preterm birth or SGA may lie on the causal pathway between maternal inflammation manifesting as chronic conditions and ADHD. Assessing the extent to which these factors mediate the effects of maternal inflammation on offspring ADHD is crucial.⁹ A recent study revealed that 48% of the total effect of maternal chronic cardiovascular or metabolic disorders on offspring cerebral palsy was mediated by preterm delivery.¹⁰

In short, Nielsen et al.'s findings stress the complexity of maternal factors, and the study's diverse maternal conditions pose a challenge for a unified approach, emphasising the need to explore each condition individually. We recommend that an understanding of how specific maternal inflammatory mechanisms may affect ADHD risk in children, and how potential mediating factors shape these risks may be worthy of future investigations. Only when all the factors causing a condition are clearly understood can any effective attempts at prevention be targeted.

注意力缺陷/多动障碍 (ADHD) 是一种神经发育障碍，涉及注意力受损、多动和冲动行为；全球超过 3% 的儿童患有这种疾病。¹ ADHD 不仅给个人带来潜在的危及生命的风险，而且还给社会带来巨大的经济负担——在美国，估计每年造成 316 亿美元的损失。²确定神经发育障碍和精神疾病病因中的可改变因素对于风险预测和预防干预措施的设计至关重要。尽管我们对多动症发展的环境起源的理解并不完整，但我们认识到，许多病因决定因素在胎儿期和婴儿期（大脑发育的最早和最脆弱的阶段）发挥作用。鉴于 ADHD 尚无治愈方法，识别 ADHD 可改变的危险因素仍然至关重要。

在本期《儿科和围产期流行病学》中，Nielsen 及其同事³探讨了母亲多次接触与妊娠期间患有多动症的炎症相关的作用。该研究以 2001 年 7 月至 2011 年 12 月在澳大利亚新南威尔士州出生的 908,770 名儿童为对象，并进行随访直至 2014 年 12 月。在 16,297 名儿童中发现 ADHD（发病率为每 1000 人年 3.5 例），中位年龄为 7 岁第一次治疗时。根据怀孕期间的出生数据和入院情况确定了七种慢性孕产妇并发症和危险因素（暴露）：自身免疫性疾病、哮喘、感染住院、情绪或焦虑障碍、吸烟、高血压和糖尿病。他们发现每次接触都与 ADHD 风险增加独立相关（风险比范围 [HR] 1.19 至 1.89）；风险随着炎症标记物的累积暴露而增加，这表明存在剂量反应关系。患有哮喘、感染、情绪或焦虑障碍以及吸烟的女性所生的孩子患 ADHD 的风险最大（HR 6.12，95% 置信区间 [CI] 3.47，10.70）。

这项研究在很多方面都是独一无二的。它通过母亲接触炎症相关疾病来提供基于人群的 ADHD 风险特征；通过剂量反应关系检查单一和多种条件，以及重要的是累积效应。它还为理解母亲炎症如何影响儿童多动症提供了急需的清晰度。在这篇评论中，我们重点介绍了这项研究如何推进该领域的发展，并指出了缺乏细致数据的领域，但可能为了解炎症在儿童多动症风险中的作用提供重要见解。

越来越多的证据表明，与孕产妇肥胖、心血管疾病、绒毛膜羊膜炎和癫痫相关的神经发育障碍风险增加，该研究的发现进一步证明了这一点。^4-6这些风险可能是由慢性病本身引起的，也可能是由用于治疗慢性病的药物引起的。连接孕产妇疾病和神经发育结果的途径很复杂，而且在很大程度上仍然未知。其中，母亲因素在形成多动症风险方面发挥着重要作用。 Nielsen 及其同事³强调，所考虑的母体暴露是异质的，并且几种情况具有重要的非炎症方面，可能会影响观察到的关联。

重要的是要承认所研究的母体条件的异质性固有的局限性，每种条件都具有导致后代多动症的独特生物学途径。虽然研究表明，随着孕产妇暴露次数的增加，患多动症的风险也会增加，但每种情况的复杂性对制定统一的干预策略提出了挑战。例如，对于自身免疫性疾病，必须在怀孕期间密切监测和管理病情和药物使用。相反，解决吸烟习惯需要有针对性的行为干预。最佳血糖监测对于孕产妇糖尿病管理至关重要，⁷而妊娠期高血压干预措施可能包括每日小剂量阿司匹林、抗高血压药物、胎儿监测和定时分娩。

这项研究开始揭示孕产妇慢性疾病对多动症风险的累积影响。然而，它提出了关于如何处理这些条件的关键问题：作为一组相互关联的条件或作为单个实体。吸烟与慢性病属于同一类吗？虽然长期经常吸烟被视为慢性暴露，但我们反对将吸烟列为慢性孕产妇疾病。相反，我们认为吸烟是多动症的潜在危险因素。最近的研究证明，以前认为的母亲吸烟与儿童 ADHD 风险之间的关联最好被描述为一种关联，这种关联被未测量的混杂因素、⁸和残余混杂因素所强烈强调。例如，在这项研究中，社会经济地位（通常与炎症状况相关）的残余混杂因素可能会导致关联产生偏差。尽管作者调整了地区平均收入作为社会经济地位的代理，但缺乏对社会经济地位的具体标记的调整可能导致社会经济地位的残余混淆。

我们认为，尽管预防多动症的共同目标是，但这些不同的方法强调需要根据每种孕产妇状况的独特特征制定量身定制的策略。

未来的研究可能会受益于单独探索每种母亲的慢性疾病（即糖尿病、肥胖、吸烟），以理清这些关联背后的具体机制，从而更好地了解母亲的疾病可能与儿童多动症存在因果关系的途径。这种方法可以深入了解因果路径，促进有针对性的干预措施的制定。

第二个问题涉及一些原有疾病和妊娠并发症与早产或 SGA 的关系。 Nielsen 及其同事³发现，暴露于母亲接触量增加的儿童更有可能早产，从而出现严重的新生儿发病率。这表明早产或 SGA 可能是表现为慢性疾病的母体炎症与 ADHD 之间的因果关系。评估这些因素在多大程度上介导母体炎症对后代多动症的影响至关重要。⁹最近的一项研究表明，母亲慢性心血管或代谢紊乱对后代脑瘫的总影响中有 48% 是由早产介导的。¹⁰

简而言之，尼尔森等人的研究结果强调了母亲因素的复杂性，并且该研究中不同的母亲状况对统一方法提出了挑战，强调需要单独探索每种情况。我们建议了解特定的母体炎症机制如何影响儿童多动症风险，以及潜在的中介因素如何影响这些风险可能值得未来的研究。只有清楚地了解导致某种情况的所有因素，才能采取有效的预防措施。