一位精神科醫師的筆記，關於一種形塑生命、也牽動周遭每一個人生命的病症
同一間候診室裡的兩個男孩
他們在同一天下午前來，前後相隔一小時。我至今仍時常一起想起他們兩人，因為他們讓我明白了「光譜」這個詞真正的含義，比任何教科書都來得深刻。
第一位是T先生，七歲，被一位看起來比他還緊張的母親帶進我的診間。他坐下來，用力握了我的手，接著不待我開口，便說出了三個關於水循環的事實。他有眼神接觸，但那似乎讓他付出了某種代價，彷彿那是他在表演的一項技能，而非自然而然的舉動。他的老師轉介他來評估，並非因為他的課業落後（事實上恰恰相反），而是因為他總是不斷糾正同學，把自己的一舉一動大聲說出來，還曾在體育課上，因為有人未經預告就更改了遊戲規則，而整個人倒在地上。旁人或許只會說他「有點古怪」，或是「很獨特」。但他的母親用了另一個說法：她說,自己的兒子交不到一個朋友,無論他多麼渴望。
第二位是J先生，六歲。候診室裡的差異立刻顯現出來。他無法坐定,在耳邊快速地拍動雙手,腳跟不停搖晃,把三輛玩具車一遍又一遍地排成一條筆直的線,只要有人靠近就會尖叫。他沒有口語。當他的母親試圖引導他走向我的診間時,他倒在地上,用手掌拍打自己的頭側。當我越過他的頭頂與她對視時,她臉上沒有尷尬的神情,只有一種早已根深蒂固的疲憊。
這兩個男孩都被診斷為自閉症類群障礙。就臨床標準而言,他們同屬一個光譜。然而,他們的人生,以及他們家庭的人生,卻截然不同。這是關於自閉症最該先了解的一點:它沒有單一的樣貌,沒有任何一個孩子或成人,能夠代表所有其他人。
自閉症究竟是什麼
自閉症類群障礙,是大腦發展上的一種差異。它影響一個人如何溝通、如何與他人建立關係,以及如何感受聲音、光線、觸覺與不可預期性所構成的感官世界。它從幼年時期便已存在,並將伴隨一生。
它不是因為教養方式所造成的。它也不是因為疫苗所造成的——這個說法已被反覆研究、驗證無數次,如今持續追問這個問題,反而已成為對更有意義的問題的一種干擾。目前我們所理解的是,它源自於複雜的遺傳因素組合,在某些情況下也涉及環境因素,這些因素共同形塑了大腦在生命最初幾年裡的連結方式。
從臨床角度而言,《精神疾病診斷與統計手冊第五版》(DSM-5)將診斷建立在兩大核心類別之上,臨床醫師必須在兩者中都看到證據。第一項是持續存在的社交溝通與互動困難,又可細分為三個具體面向:在正常對話與情感交流的一來一往中出現困難;在與人連結的非語言層面上有困難(眼神接觸、臉部表情、肢體動作、解讀身體語言);以及在建立、維持並理解符合年齡的人際關係上有困難,從交朋友,到僅僅是依不同社交場合調整自身行為。
第二項類別則是侷限或重複的行為模式,患者必須至少表現出以下四種之一中的兩種:重複性的動作、言語或物品使用方式(拍手、搖晃身體、重複他人話語、排列玩具);對一致性與例行事務有強烈需求,即使是微小、未經預告的變化,也可能感覺像一場真正的危機;高度侷限且異常強烈或專注的興趣;以及對感官輸入有異常的反應方式,無論是主動尋求刺激,還是被聲音、光線、質地、溫度或疼痛所淹沒。
除了這兩大類別之外,這些特質必須在孩童發展的早期便已出現,即使當時未被察覺,並且必須對日常生活造成實質影響。臨床醫師還會依患者所需的支持程度,評定一到三級的嚴重程度,這也正是為何兩位擁有相同診斷的人,外顯樣貌可能截然不同的部分原因。
而這份臨床描述所遺漏的,是這些特質在個體之間的差異究竟有多大。有些孩子在學會綁鞋帶之前,就已經能說出完整的段落,並在未來獨立生活、開車、結婚、從事高要求的工作。另一些孩子則終身未發展出口語能力,需要協助處理最基本的日常事務,終其一生都需要支持。過去,醫師會將這些不同表現分別列為獨立的診斷。如今我們使用單一光譜的概念,因為大多數人都落在一個廣闊的中間地帶,即使是同一個人,其需求也可能隨著日子而有所變化。
尋找答案
診斷通常始於一個令人擔憂的觀察:嬰兒對自己的名字沒有反應、不會用手指向物品分享興趣,或是失去了原本已學會的詞彙。如今,兒科醫師會在十八個月與二十四個月左右的例行健檢中,為幼兒進行自閉症篩檢。若有需要,更完整的評估通常會涉及發展兒科醫師、兒童精神科醫師或心理師,以及經常包含語言治療師。
相關數字已有相當大的變化。根據美國疾病管制與預防中心於2025年發布的最新監測資料,美國每31名八歲兒童中,大約就有1人被確診自閉症,相較於2000年的每150人中1人,有明顯上升。多數研究者將此上升歸因於篩檢方式的改善與診斷標準的擴大,而非病例真正大幅增加,不過完整的全貌仍在釐清之中。男孩被診斷的比例是女孩的三倍以上,而且愈來愈多人意識到,自閉症女孩常常被漏診,部分原因在於她們的表現往往較為隱微,也更擅長掩飾自己的症狀,但這也讓她們的心理健康付出了真實的代價。
基因檢測如今已被視為完整評估流程中的標準項目。當找到特定的基因成因時,它能協助指引從癲癇監測,到讓一個家庭與其他擁有相同診斷的家庭建立連結等各種層面。然而在實務上,基因檢測的比例仍然偏低,大多數家庭永遠得不到一個乾脆俐落的「為什麼」。
一條路徑,而非一條直線
「病程進展」這個詞用在自閉症上並不恰當。它暗示著一種疾病沿著既定時間軸不斷推進,如同腫瘤生長或感染擴散一般。自閉症並非如此運作。孩子通常不會隨著年齡增長而「變得更自閉」。
真正改變的,是隨著世界對孩子的要求愈來愈多,自閉症變得愈來愈明顯可見。一個原本看起來只是稍微不同的幼兒,到了中學階段,可能與同儕顯得截然不同,這並非因為他本身改變了多少,而是因為他周遭的社交世界變得更加嚴苛。
在適當的支持下,許多自閉症孩子在溝通、彈性與獨立性方面,隨著時間會有真正的進步。若缺乏支持,或同時伴隨焦慮、癲癇或智能障礙等狀況(根據現有檢測資料,約有四成的自閉症兒童伴隨智能障礙),情況則可能艱難許多。青春期往往帶來另一種考驗,自閉症青少年會痛苦地意識到自己與同儕之間的差異,這也伴隨著真實的憂鬱與焦慮風險。接著便是成年期,而正如我稍後會談到的,那個支撐一個家庭長達十八年的支持系統,往往就這樣被撤走了。
沒有人看見的重擔
照顧一個需要高度支持的自閉症孩子,是一個人所能承擔的、最令人筋疲力盡、也最孤立無援的工作之一。而我們的體制,大多並未以此看待這件事。
研究結果與我在診間每週所聽到的相符。自閉症孩童的父母,回報的壓力、憂鬱與焦慮程度,都高於照顧其他慢性病孩童的父母,而這樣的壓力,經年累月下來,也會反映在他們自身的身體健康上。一項追蹤父母長達近三十年的研究發現,這些家庭的離婚風險,即使到孩子成年以後,依然居高不下,背後的原因包括財務壓力、睡眠不足,以及一種幾乎沒有真正喘息機會的照顧負擔。母親作為主要照顧者的多數群體,回報的心理健康狀況與負擔程度,都比父親更為嚴重,不過父親也絕非置身事外。
再來是經濟層面的問題。依據個人所需的支持程度不同,研究者估計,在美國,照顧一位自閉症人士的終身成本遠遠超過一百萬美元,涵蓋治療、教育、醫療照護、收入損失與長期住房支持等。過去二十年來,保險對自閉症相關服務的給付雖有改善,但各州之間仍差異極大,家長也經常必須為了孩子理應享有的權益,不斷抗爭。
還有一項多數人直到親身經歷之前,從未聽聞的安全風險。四到十七歲的自閉症孩童中,將近一半曾試圖從安全環境中「逃離」或「走失」,往往是被一種強烈卻難以理解的、對水的吸引力所驅使。溺水是自閉症孩童走失後死亡的首要原因,而據估計,自閉症孩童溺水身亡的機率,是同齡一般孩童的百倍以上。曾照顧有走失傾向孩子的家長,總是描述著同樣的情景:一扇永遠不能忘記上鎖的門,一種多數人只會聯想到戰區、而非尋常週二午後居家生活的高度警戒狀態。
新舊治療方式
自閉症目前最具實證基礎的治療方式,是早期、密集的行為介入,最常見的是應用行為分析(Applied Behavior Analysis,簡稱ABA)。ABA運用增強原則,協助建立溝通、日常生活與社交技能,數十年來的研究都支持其成效,尤其是及早開始且執行得當時效果更佳。
然而,這並非一項已有定論、毫無爭議的療法,這場辯論也值得在此獲得公平的呈現。愈來愈多、聲音也愈來愈響亮的自閉症成人族群——其中許多人小時候就曾接受過ABA治療——對這項療法過去的實施方式提出了嚴肅的質疑:過去那些僵化、以服從為導向的方案,每週可能長達三、四十小時;僅僅因為拍手等無害行為看起來「不尋常」就加以懲罰;讓部分成人在他們自己的敘述中,雖然學會了掩飾真實的自己,卻未曾學到真正的因應能力。有些人形容那段經歷更接近創傷,而非治療。這個領域已日益正視這些批評,並由此發展出真正的改革運動,把重心放在個案本身的目標與舒適度上,不再壓抑無害的自我刺激行為,轉而聚焦於真正能改善日常生活的技能,而非僅僅讓一個人看起來更「正常」的技能。在2026年的今天,這個領域正處於一場真正且必要的自我省思之中,家長也完全有理由,直接詢問任何一位服務提供者,他們所執行的方案究竟如何隨之演進。
除了ABA之外,可運用的工具還相當廣泛:語言治療、協助處理感覺與動作挑戰的職能治療、以遊戲為基礎的地板時間(Floortime)等取向,以及協助兒童與青少年建立友誼的社交技巧團體。目前沒有任何藥物能直接治療自閉症的核心特徵,但藥物確實能大幅協助處理共病的焦慮、易怒、注意力不足過動症或攻擊行為,而這往往正是我身為精神科醫師,最能發揮作用之處。
研究前沿正快速推進,不過以下這些都尚未成為常規臨床照護的一部分。針對Phelan-McDermid症候群這類與自閉症高度相關的基因疾病,基因治療試驗正在進行中,而天使人症候群(Angelman syndrome)與雷特氏症(Rett syndrome)等相關疾病的早期研究,也在過去這一兩年間,從動物模式邁入了人體試驗階段。此外,一項由美國國家衛生研究院資助、名為「自閉症生物標記聯盟」的重大計畫,近十年來致力於找出可測量的生物標記——例如腦部活動或眼球移動的特定模式——盼望有朝一日能協助更早、更客觀地進行診斷。這些都稱不上是「治癒」,而「治癒」這個詞,對許多自閉症成人而言,本身就是一個帶有爭議、令人反感的字眼。這些研究所提供的,是及早獲得答案、以及更精準對症支持的希望。
那個無法停止傷害自己的女孩
我想告訴大家一位我在住院精神科病房、而非門診診間所見的病患,因為她的案例,道出了我們在照護這個族群時,一些重要而令人不安的落差。
R小姐當年十四歲,沒有口語能力,是在數週逐漸加劇的自傷行為後,由父母帶到急診室的:嚴重到留下瘀青的撞頭行為,以及咬自己的雙手直到流血。她的父母帶來了一樣我從未在其他家庭身上見過的東西——一份仔細打字、護貝完成的員工提示單。兩頁的內容,詳細列出她主動尋求的感官與活動、她無法忍受的事物、她每日作息的樣貌,以及往往在危機發生前出現的具體觸發因子。這在某種意義上,是一份了不起的文件,凝聚了十四年來疲憊卻始終專注的愛。
住院期間,我們透過調整藥物,並與行為治療團隊密切合作,找出導致自傷行為的原因(在她的案例中,我們認為是感官超載,加上因無法表達需求而產生的深切挫折感共同造成),讓她逐步穩定下來。她的狀況穩定地改善。但隨著幾週過去,另一件事也逐漸浮現。
那對在提示單上如此細心的父母,在女兒住院後,卻變得出乎意料地難以聯繫。電話經常沒有回覆。出院規劃會議一再缺席。社工協助他們開始申請的住房與日間療育計畫,也擱置了數週未動。我們整個團隊漸漸察覺到一件不太自在、卻不得不說出口的事:這是R小姐十四年來第一次離開家中,而她父母的內心深處,某個部分——完全可以理解地,身為人、身為長期身心俱疲的照顧者——並不急著讓這段喘息時光結束。
我說這些,並非要譴責他們。我從未承受過他們所承受的重擔。但一個孩子不能因為她的照顧者出於完全可以同理的理由而停止推進,就無限期地滯留在醫院病床上。最終,我們的社工團隊向兒童保護服務機構提出了通報,並非因為任何人認為R小姐正遭受虐待,而是因為那已是促使她父母重新積極投入的唯一可行手段。這個做法奏效了。通報後短短兩週內,相關文件便完成了,一個專門的療育方案也找到了,R小姐出院前往一個真正能夠滿足她需求的地方。
我至今仍時常想起這個案例。兩位充滿愛、盡心盡力的父母,以及一個薄弱到必須動用州政府介入的威脅,才能讓他們得到他們始終在尋求的協助的體制。
當父母不在了以後
這是整個故事中,最令我夜不能寐的部分。
在孩子十八歲之前,體制保障了各項服務:特殊教育、治療,以及一份具法律效力的個別化教育計畫。但在多數州,孩子一到二十二歲生日前後,這份保障便戛然而止——家庭們貼切地稱之為「懸崖」。此後迎接他們的,是一套由州別各異、拼湊而成的Medicaid豁免計畫,而幾乎在每一個地方,這些計畫的資源都遠遠不足以應付需求。目前全美有數十萬名智能與發展障礙成人,正在等候名單上苦候日間療育、就業支持與居家照護服務。而住宿安置的等候名單——也就是許多成人未來勢必需要的團體家園——情況更糟:在部分州,等候時間長達七年、十年,甚至二十年。我曾聽過一位家長描述,在等候十五年之後,終於接到住宿名額釋出的電話。我也曾有其他家長直接問我,當他們自己有一天離世,他們的孩子將會何去何從。
這個問題沒有單一的解方。有些州確實做得比其他州更好。例如加州的《蘭特曼法案》(Lanterman Act),便保障了發展障礙人士的服務權利,並資助了一整個區域中心網絡,不過即使在加州,倡議人士也表示,需求依然遠遠超過現有的服務量能。相關資金主要來自Medicaid,透過所謂的「居家與社區服務」豁免計畫提供,部分地區則輔以州政府經費,而對於能夠及早規劃的家庭而言,則可運用特殊需求信託與ABLE帳戶等工具,讓家庭得以為身心障礙的家人存下款項,同時不影響其請領公共福利的資格。專家們一致建議、我如今也如此告訴我診間裡的每一個家庭:及早申請豁免計畫的等候名單,最好是在孩子滿十八歲的那一刻就開始。在情況尚未緊急之前,就先處理好監護權或輔助決策的安排。誠實地與手足討論,他們未來可能被要求扮演的角色。這一切都無法真正修補一個從未為如此龐大人口而設計的體制。但在此刻,這是一個家庭手中,唯一真正握有的槓桿。
診斷之外,延續一生的代價
我先前已經談過婚姻的壓力、經濟上的沉重負擔、溺水的風險,以及多年照顧所帶給父母身體上的損耗。以下還有幾項代價,值得我明白地說出來,因為這些家庭理應從一位醫師口中聽見這些,而不是獨自摸索發現。
有手足所承受的代價,他們往往被迫比同齡人更快長大,很早就學會將自身的需求,環繞著兄弟姊妹的危機來調整。有婚姻所承受的代價,離婚率的研究已清楚說明這一點,但在我的診間裡,它往往以更安靜的方式呈現:一對彼此相愛、卻已多年未曾有過一次不被打斷的對話的伴侶,只是因為心力交瘁而無法表現出愛意。也有法律層面的代價——監護權爭議、因診斷而變得更複雜的監護權訴訟,以及在極少數、最嚴重的案例中,一位被推向極限之外的照顧者,做出了災難性的決定。我在此不會詳述這些案例,但它們確有紀錄、真實存在,代表著一個要求一個人獨自承擔過多、過久的支持體制,所導致的最極端、最令人痛心的失能後果。此外還有攸關生命的代價:幼年時期因走失與溺水而升高的死亡率,以及有研究指出,自閉症成人——尤其是幾乎沒有口語能力者——比一般人口更早離世,而研究者認為,這樣的落差,源自於未被滿足的醫療需求,以及與醫療人員之間的溝通障礙,而非自閉症本身所致。
這一切,將我們帶向何方
在與這些家庭相處多年之後,有兩件事,我不斷回頭想起。
其一是,自閉症本身,並不是一件值得哀悼的事。我所認識的自閉症人士,無論有無口語能力,無論獨立自主或高度依賴他人,都只是單純地活著,以他們自己的方式,過著完整而獨特的人生。看著T先生成長為一位能言善道、能力出眾的年輕人,在一所終於學會與他合作、而非試圖糾正他的學校裡,我更加理解了,自閉症成人所說的——他們的神經特質是一種不同,而非一種缺陷,究竟是什麼意思。
第二件事是,那些正在支持有重大需求者的家庭,往往承擔了遠超過任何一個家庭合理所能負荷的重量,而這並非反映在他們身上的問題,而是反映出目前有多少支持體系,能夠一同分擔這份重量。J先生的母親,也就是我多年前在那間候診室裡遇見的那位母親,她愛她的兒子,那份熾烈的愛,我至今仍記得清清楚楚。像她、也像R小姐父母這樣的家庭,他們要的不是憐憫。他們要的,是及早出現、持續穩定,並且陪伴至真正不再需要為止的支持。而建立起這樣的支持,受益的將是每一個人:那些與自閉症共處的人們、愛著他們的家庭,以及他們所共同屬於的社群。
The Long Silence and the Loud World: Autism Spectrum Disorder
A psychiatrist’s notes on a condition that shapes a life, and every life around it
Two Boys in the Same Waiting Room
They came in on the same afternoon, an hour apart. I still think about the two of them together, because they showed me, better than any textbook could, what the word “spectrum” actually means.
The first was Mr. T, seven years old, marched into my office by a mother who looked more nervous than he did. He sat down, shook my hand a beat too firmly, and told me, unprompted, three facts about the water cycle. He made eye contact, though it seemed to cost him something, like a skill he was performing rather than a thing that came naturally. His teacher had referred him not because he was falling behind (he wasn’t) but because he corrected his classmates constantly, narrated his own actions out loud, and once dropped to the gym floor mid-game because someone had changed the rules without warning him. You might have called him “a little odd,” or “quirky.” His mother called him something else: her son who could not make a single friend, no matter how badly he wanted one.
The second was Mr. J, six years old. The difference in the waiting room was immediate. He would not sit. He flapped his hands near his ears in quick bursts, rocked on his heels, and lined up three toy cars into a perfectly straight row, over and over, screaming if anyone got close to them. He had no spoken words. When his mother tried to guide him toward my office, he dropped to the floor and hit the side of his own head with an open palm. Her face, when I caught her eyes over his head, wasn’t embarrassed. It was exhausted, in a way that had settled in long ago.
Both boys had autism spectrum disorder. Both were, by any clinical measure, on the same spectrum. And yet their lives, and their families’ lives, looked nothing alike. That’s the first thing worth knowing about autism: there is no single face of it, and no one child or adult who stands in for all the others.
What Autism Actually Is
Autism spectrum disorder is a difference in brain development. It shapes how a person communicates, relates to others, and experiences the sensory world of sound, light, touch, and unpredictability. It’s present from early childhood and lasts a lifetime.
It is not caused by parenting. It is not caused by vaccines, a claim studied so thoroughly and so often that it has become, at this point, a distraction from more useful questions. What we do understand is that it arises from a complex mix of genetic factors, and in some cases environmental ones, that shape how the brain wires itself in the earliest years of life.
Clinically, the DSM-5 organizes the diagnosis around two core categories, and a clinician has to see evidence in both. The first is persistent difficulty with social communication and interaction, which breaks down into three specific pieces: trouble with the normal back-and-forth of conversation and shared emotion, difficulty with the nonverbal side of connecting with people (eye contact, facial expressions, gestures, reading body language), and trouble developing, keeping, and understanding relationships appropriate to one’s age, from making friends to simply adjusting behavior to fit different social settings.
The second category is restricted or repetitive patterns of behavior, and a person needs to show at least two of four kinds: repetitive movements, speech, or use of objects (hand-flapping, rocking, echoing words, lining up toys); an intense need for sameness and routine, where even a small, unannounced change can feel like a genuine emergency; highly restricted interests that are unusual in their intensity or focus; and unusual reactivity to sensory input, either seeking it out or being overwhelmed by it, whether that’s sound, light, texture, temperature, or pain.
On top of both categories, these traits have to show up early in a child’s development, even if they aren’t recognized until later, and they have to meaningfully get in the way of daily life. Clinicians also assign a severity level, one through three, based on how much support a person needs, which is part of why two people with the same diagnosis can look so different from each other.
What that clinical description misses is how much these traits vary from person to person. Some children speak in full paragraphs before they can tie their shoes, and go on to live independently, drive, marry, hold demanding jobs. Others never develop spoken language, need help with basic daily tasks, and will need support for life. Doctors used to split these presentations into separate diagnoses. Now we use one spectrum, because most people fall somewhere in a wide middle, and even one person’s needs can shift from day to day.
Finding the Answer
Diagnosis usually starts with a worried observation: a baby who doesn’t respond to their name, doesn’t point to share something interesting, or loses words they once had. Pediatricians now screen for autism at routine checkups around eighteen and twenty-four months. From there, a fuller evaluation typically involves a developmental pediatrician, a child psychiatrist or psychologist, and often a speech-language pathologist.
The numbers have shifted a great deal. The CDC’s most recent surveillance, published in 2025, found that roughly 1 in 31 eight-year-olds in the United States has been identified with autism, up from 1 in 150 in the year 2000. Most researchers attribute that rise to better screening and broader diagnostic criteria, not a true explosion in cases, though the full picture is still being worked out. Boys are diagnosed more than three times as often as girls, and there’s growing recognition that autistic girls are frequently missed, in part because they tend to present more subtly and are often better at masking their symptoms, at real cost to their own mental health.
Genetic testing is now considered a standard part of a thorough workup. When a specific genetic cause is found, it can guide everything from seizure monitoring to connecting a family with others who share the same diagnosis. In practice, though, testing rates remain low, and most families never get a single, tidy answer to “why.”
A Path, Not a Straight Line
“Progression” is the wrong word for autism. It implies a disease marching forward on a set timeline, the way a tumor grows or an infection spreads. Autism doesn’t work that way. A child does not typically get “more autistic” with age.
What changes is how visible the autism becomes as the world asks more of a child. A toddler who seemed only mildly unusual can look starkly different from his peers by middle school, not because he’s changed so much as because the social world around him has gotten harder.
With the right support, many autistic children make real gains in communication, flexibility, and independence over time. Without it, or alongside conditions like anxiety, epilepsy, or intellectual disability (present in roughly forty percent of autistic children with available testing data), the picture can be much harder. Adolescence often brings its own reckoning, as autistic teenagers become painfully aware of how different they are from their peers, which carries a real risk of depression and anxiety. Then comes adulthood, when, as I’ll get to, the support that held a family up for eighteen years is often simply taken away.
The Weight Nobody Sees
Caring for a child with significant support needs is one of the most exhausting, isolating jobs a person can take on. Our systems mostly don’t treat it that way.
The research matches what I hear in my office every week. Parents of autistic children report higher rates of stress, depression, and anxiety than parents of children with other chronic conditions, and over years, that strain shows up in their physical health too. A study tracking parents over nearly three decades found that divorce risk in these families stays elevated well into their child’s adulthood, driven by financial pressure, lost sleep, and a caregiving load with no real breaks. Mothers, who remain the majority of primary caregivers, report worse mental health and higher burden than fathers, though fathers are far from spared.
Then there’s the money. Depending on a person’s support needs, researchers put the lifetime cost of supporting someone with autism in the United States well over a million dollars, covering therapy, education, medical care, lost income, and housing. Insurance coverage has improved over the past two decades, but it still varies wildly by state, and often requires parents to fight for what their child is supposed to be entitled to.
And there’s a safety risk most people never hear about until it touches their own family. Nearly half of autistic children between four and seventeen have tried to wander, or “elope,” from a safe environment at least once, often pulled by a strong and poorly understood attraction to water. Drowning is the leading cause of death among autistic kids who wander off, and autistic children are estimated to be over a hundred times more likely to die by drowning than their peers. Parents of kids prone to wandering describe the same thing, over and over: a door that can never be left unlocked, a level of vigilance most people would associate with a warzone rather than a Tuesday afternoon at home.
Therapies, Old and New
The best-established treatment for autism is early, intensive behavioral intervention, most often Applied Behavior Analysis, or ABA. ABA uses reinforcement to build communication, daily living, and social skills, and decades of research back its effectiveness, especially when it starts young and is delivered well.
It’s also not a settled, uncontroversial matter, and it deserves a fair hearing here. A growing, vocal movement of autistic adults, many of whom went through ABA as children, has raised real concerns about how it was historically practiced: rigid, compliance-heavy programs running thirty or forty hours a week, punishing harmless behaviors like hand-flapping simply because they looked unusual, and leaving some adults, by their own account, skilled at masking who they are but without genuine coping tools. Some describe the experience as closer to trauma than therapy. The field has increasingly taken this seriously, and a real reform movement has grown out of it, one that centers a person’s own goals and comfort, leaves harmless stimming alone, and focuses on skills that actually improve daily life rather than skills that just make someone look more typical. In 2026, this is a field in the middle of reckoning with itself, and it’s fair for a family to ask any provider directly how their particular program has evolved.
Beyond ABA, the toolkit is broad: speech and language therapy, occupational therapy for sensory and motor challenges, play-based approaches like Floortime, and social skills groups for kids and teens navigating friendship. No medication treats autism’s core features directly, but medication can help a great deal with co-occurring anxiety, irritability, ADHD, or aggression, and that’s often where I, as a psychiatrist, can be most useful.
The research frontier is moving fast, though none of it is routine clinical care yet. Gene therapy trials are underway for conditions like Phelan-McDermid syndrome, a genetic condition strongly linked to autism, and early studies in related conditions like Angelman and Rett syndromes have moved from animal models into human trials in just the past couple of years. Separately, a major NIH-funded effort called the Autism Biomarkers Consortium has spent nearly a decade trying to identify measurable biological markers, patterns in brain activity or eye movement, that could someday support earlier, more objective diagnosis. None of this is a cure, and “cure” is a loaded word many autistic adults reject outright. What it offers is the hope of earlier answers and better-targeted support.
The Girl Who Could Not Stop Hurting Herself
I want to tell you about a patient I saw in an inpatient psychiatric unit, not in my outpatient clinic, because her case says something important, and something uncomfortable, about the gaps in how we care for this population.
Ms. R was fourteen, nonverbal, and was brought to the emergency room by her parents after weeks of escalating self-injury: head-banging severe enough to leave bruising, biting her own hands until they bled. Her parents arrived with something I’d never seen a family bring before, a carefully typed, laminated tip sheet for staff. Two pages, listing exactly which sensations and activities she sought out, which ones she couldn’t tolerate, the shape of her daily routines, and the specific triggers that came before a crisis. It was, in its way, a remarkable document, built from fourteen years of exhausted, attentive love.
We stabilized her over the admission, adjusting medication and working with a behavioral team to figure out what was driving the self-injury (in her case, a mix of sensory overload and sheer frustration at not being able to communicate). She improved steadily. But something else became clear as the weeks passed.
Her parents, so meticulous in that tip sheet, grew strikingly hard to reach once she was admitted. Calls went unreturned. Discharge planning meetings got missed. Housing and day program applications our social worker had helped them start sat untouched for weeks. It was uncomfortable to name what our team had started to sense: that for the first time in fourteen years, Ms. R was out of the house, and some part of her parents, understandably human and understandably depleted, wasn’t in a hurry for that break to end.
I don’t say that to condemn them. I’ve never carried what they carried. But a child can’t stay in a hospital bed indefinitely because her caregivers, for entirely sympathetic reasons, have stopped pushing. Our social work team ultimately filed a report with child protective services, not because anyone thought Ms. R was being abused, but because it was the only lever left to get her parents actively engaged again. It worked. Within two weeks, the paperwork was done, a specialized program was found, and Ms. R was discharged somewhere built to actually meet her needs.
I still think about that case. Two loving, devoted parents, and a system so thin that it took the threat of state involvement to get them the help they’d been asking for all along.
When the Parents Are Gone
This is the part of the story that keeps me up at night.
For eighteen years, the system guarantees services: special education, therapy, a legally enforceable education plan. Around a child’s twenty-second birthday, in most states, that guarantee simply ends, a moment families call, accurately, “the cliff.” What follows is a patchwork of Medicaid waiver programs, and nearly everywhere, they’re underfunded relative to need. Several hundred thousand adults with intellectual and developmental disabilities are currently on waitlists nationwide for day programs, employment support, and in-home care. Waitlists for residential placement, the group homes many adults will eventually need, are worse: in some states, seven, ten, even twenty years. I’ve had a parent describe getting the call that a spot had opened after a fifteen-year wait. I’ve had others ask me, point blank, what happens to their child on the day they themselves die.
There’s no single fix for this. Some states do better than others. California’s Lanterman Act, for instance, guarantees services for people with developmental disabilities and funds a network of regional centers, though even there, advocates say demand outpaces capacity. Funding mostly comes through Medicaid, via Home and Community-Based Services waivers, sometimes supplemented by state funds and, for families who can plan ahead, tools like special needs trusts and ABLE accounts, which let a family save money for a disabled relative without losing public benefits. What experts consistently recommend, and what I now tell every family in my practice: get on waiver waitlists early, ideally the moment a child turns eighteen. Sort out guardianship or supported decision-making before it’s urgent. Talk honestly with siblings about what role they might be asked to play. None of that fixes a system that wasn’t built for this many people. But right now, it’s the only real lever a family has.
The Toll That Outlives the Diagnosis
I’ve already touched on the marital strain, the financial cost, the drowning risk, the physical toll of years of caregiving. A few more of these are worth naming plainly, because families deserve to hear them from a doctor rather than discover them alone.
There’s the toll on siblings, who often grow up faster than they should, learning early to fold their own needs around a brother or sister’s crisis. There’s the toll on marriages, which the divorce research makes plain, but which shows up in my office as something quieter: couples who haven’t had an uninterrupted conversation in years, who love each other and are simply too depleted to act like it. There’s a legal toll too, guardianship disputes, custody battles complicated by a diagnosis, and, in the rarest and most severe cases, a caregiver pushed past every reasonable limit who makes a catastrophic decision. I won’t detail those cases here, but they’re documented and real, and they represent the most extreme failure of a system that asked too much of one person for too long. And there’s a fatal toll: higher mortality from wandering and drowning in childhood, and research suggesting that autistic adults, especially those who are minimally verbal, die younger than the general population, a gap researchers link to unmet medical needs and communication barriers with providers rather than anything inherent to autism itself.
Where This Leaves Us
Two things seem clear to me after years of sitting with families like these, and I don’t think they contradict one another.
The first: autism itself is not something to grieve. The autistic people I’ve known, verbal and nonverbal, independent and deeply dependent on others, are simply people living full, distinct lives on their own terms. Watching Mr. T grow into a young man who’s articulate and capable, in a school that’s finally learned to work with him instead of around him, I understand a little better what autistic adults mean when they describe their neurology as different rather than broken.
The second: families supporting someone with significant needs are often carrying more than any one family reasonably should, and that’s not a reflection on them. It’s a reflection of how much support currently exists to share that load. Mr. J’s mother, the one I met in that waiting room years ago, loved her son with a ferocity I still remember. Families like hers, and like Miss R’s, aren’t asking for pity. They’re asking for support that shows up early, stays consistent, and lasts as long as it’s needed. Building that benefits everyone: the people living with autism, the families who love them, and the communities they’re all part of.
作者：Dr. Bob Lee, DO, MS, MBA
Chief Resident Physician, Department of Psychiatry & Behavioral Sciences
Child and Adolescent Psychiatry Fellow
Nassau University Medical Center
【Published by Chicago Chinese News/芝加哥時報】
