Adolescent substance use is often associated with recognisable risk factors, including academic disengagement, peer substance use, family conflict and co-occurring emotional problems such as anxiety and depression. In clinical practice, these risk profiles shape who is asked detailed questions and who is screened. However, risk is not limited to overt social instability. In some high-achieving adolescents, less visible drivers, such as perfectionism, sustained performance pressure, fear of failure, untreated social anxiety and chronic pain, may increase vulnerability. These same factors may also reduce the likelihood of detection by families, teachers and clinicians. At the same time, access routes have changed. The expanding online availability of psychoactive substances can lower the practical barriers to initiating and sustaining use and can support concealed patterns (e.g., small-quantity purchases delivered to the home). International and Chinese reports have warned that digital drug markets and social media-mediated trafficking are becoming a major public health challenge.1-3 Routine screening in clinical settings often relies on self-report, limited toxicology panels and a traditional risk-profile heuristic. This approach may be insufficient when exposure involves multiple drug classes, novel psychoactive substances or adulterated products purchased online. Early symptoms may be misattributed to academic stress, and clinicians may underestimate substance exposure in adolescents who appear socially stable and academically successful. Here, we report the case of an academically successful late-adolescent male from China who developed anxiety, insomnia, suicidal ideation and autonomic instability after approximately 1 year of concealed online-acquired polysubstance use. This case highlights a striking diagnostic blind spot: substantial substance exposure may go unrecognised in adolescents who appear high-functioning, particularly when substances are obtained online. The patient was an 18-year-old male university student with ankylosing spondylitis (AS) diagnosed in 2013 (approximately 10 years prior to presentation in 2023). He had been receiving adalimumab for approximately 5 years. He had no history of tobacco or alcohol use. He had strong academic performance and came from a financially stable supportive family; because of his academic achievements, his parents tended to grant him substantial autonomy in daily life. In September 2023, shortly after the semester began, he experienced marked social anxiety during an orientation activity. The activity involved a traditional passing game in which the student holding a flower when the drumbeat stopped had to sing, dance or answer a question. Wanting to appear confident and to enhance his creativity, he took a single dose of modafinil, a prescription medication for narcolepsy, without medical supervision during that activity. According to his report, the dose was as high as 800 mg. Within hours of taking the drug, he developed a severe panic attack with suicidal ideation: he ran to the washroom and felt an urge to jump out of the window. Over the following week, he continued to have panic attacks 1–2 times per day, typically triggered by crowded settings. These attacks severely impaired his concentration, making it impossible for him to attend classes or maintain his daily self-care. His sleep also became disrupted during this period. He then informed his parents, requested sick leave and subsequently suspended his studies. After returning home in October 2023, his parents noticed within the first few days that he was persistently low in mood, crying nearly every day and sleeping only a few hours per night. He also complained of headaches and palpitations. Concerned, he and his parents began checking his blood pressure with a home cuff. To their alarm, the readings were repeatedly elevated, sometimes as high as 160/100 mmHg. About 1 week after his return, he underwent a single comprehensive physical examination at a nearby tertiary general hospital, which included a brain magnetic resonance imaging (MRI); all results, including the MRI, were normal. The internist who saw him suspected an underlying emotional disorder and suggested that he consult a psychiatrist. Following that advice, he presented to psychiatric outpatient care in late October 2023, approximately 2 weeks after coming home. There, he was diagnosed with depressive disorder and started on paroxetine 20 mg/day, trazodone 25 mg/day and lorazepam 0.5 mg/day. The medications brought little relief. His symptoms persisted, and his parents' concern only deepened his guilt over hiding his drug use. Unable to bear the secret any longer, in late November 2023, he finally told them he had not disclosed his substance use history to the clinicians. His parents were shocked but immediately accompanied him back to the clinic. There, he told the clinicians that he had used a variety of drugs for about 1 year. His drugs were purchased via overseas websites, domestic online forums and social media applications. The substances he described included opioid analgesics, sedatives/hypnotics, stimulants and performance-enhancing drugs, anabolic steroids, dextromethorphan derivatives, and other agents (table 1). Importantly, his exposure involved multiple pharmacological classes rather than a single substance, increasing both the complexity of presentation and the difficulty of clinical recognition. He was intensely afraid of both addiction and legal repercussions. ‘I've just turned 18’, he said, ‘and I want to save myself’. After his parents accompanied him back to the clinic and informed the psychiatrist of his substance use, he became intermittently impulsive and aggressive at home (e.g., shouting and throwing objects). The outpatient regimen was adjusted (paroxetine reduced; sodium valproate and low-dose quetiapine added), and the psychiatrist ordered serum and urine toxicology tests. Immediately afterwards, for a full month, his parents attempted home-based supervision: they searched the house for hidden drugs, intercepted home deliveries, restricted his finances and limited his access to his phone and computer. During the first 2 weeks of that month, the patient could not abstain from drugs and developed prominent withdrawal symptoms, including hypertension (up to 170/130 mmHg), diaphoresis and paraesthesias described as ‘ants crawling’ on the skin. He also reported increasing his benzodiazepine intake (lorazepam up to 10 mg/day; clonazepam up to 8 mg/day). His attempts at detoxification at home were distressing—causing severe agitation, frequent conflicts with his parents and physical discomfort—and were repeatedly followed by relapse. Over the next 2 weeks, his condition worsened, prompting referral for inpatient detoxification in late December 2023. Thus, from his initial disclosure of drug use to his parents and clinicians, 1 month had passed before inpatient admission was deemed necessary. On inpatient admission, he did not show any evidence of hallucinations, delusions or manic symptoms. He reported low mood and strong craving but denied a sustained period of elevated mood. He also gave several drivers for his initial drug exploration. First, he had long been curious about pharmacology and chemistry and felt drawn to experiment with drug effects. Second, he felt that he could not adequately control his AS-related pain, so he used opioids such as tramadol or morphine for analgesia. Third, given that he was ‘eager for quick success’, he used stimulants to enhance his concentration and reach a state of euphoria and anabolic agents to improve his fitness performance. In addition, he spent substantial amounts of money (tens of thousands of yuan) without his parents' awareness, mainly using his pocket money and holiday gift money. Biochemical tests showed blood levels of alanine aminotransferase 70.6 U/L (reference range 9–50), aspartate aminotransferase 45.4 U/L (15–40), γ-glutamyl transferase 78.6 U/L (10–60), total cholesterol 8.84 mmol/L (3.1–5.7), low-density lipoprotein cholesterol 5.3 mmol/L (2.1–3.7) and uric acid 450 μmol/L (142–416). However, routine blood tests, thyroid function, hormone levels and routine urine and stool analyses showed no abnormalities. Electrocardiography and electroencephalography were also normal. His 24-item Hamilton Depression Rating Scale (HAMD-24) and Hamilton Anxiety Rating Scale (HAMA) scores were 16 and 13, respectively, indicating mild-to-moderate depressive symptoms and mild anxiety symptoms. During his hospital stay, he continued to report persistent craving and experienced sweating and palpitations. Lorazepam was gradually tapered to 2 mg/day; quetiapine was increased to 400 mg/day and sodium valproate extended-release was increased to 750 mg/day. Behavioural dysregulation remained prominent during admission: he repeatedly walked about the ward in only his underwear, refused hospital meals whilst frequently ordering takeout, and consumed large amounts of caffeinated beverages. His mother was anxious about his condition but usually complied with his requests. Staff and family repeatedly encouraged him to remain hospitalised, but he left against medical advice 14 days after admission. At discharge, his HAMD-24 and HAMA scores had decreased to 4 and 2, respectively, both within the normal range, indicating marked improvement in affective symptoms despite ongoing behavioural dysregulation. Based on International Classification of Diseases, 10th Revision (ICD-10) criteria, the patient was diagnosed with mental and behavioural disorders due to multiple drug use, dependence syndrome (F19.2). This diagnosis corresponds to disorders due to multiple substance use in ICD-11 and is broadly comparable to polysubstance-related substance use disorder in Diagnostic and Statistical Manual of Mental Disorders, 5th Edition. During outpatient follow-up over the next 2 years (2024–2025), he remained actively engaged in treatment and developed a stable therapeutic alliance with the treatment team. Clinician assessments indicated generally improved mood and anxiety symptoms compared with baseline, although these symptoms fluctuated and did not remain as well controlled as at hospital discharge. Despite ongoing pharmacological treatment for insomnia and anxiety, together with psychological support, he continued to experience persistent cravings, anxiety and sleep disturbance. By the end of follow-up, he had returned to school but did not live on campus. Instead, he lived at home and relied on close parental supervision to reduce the risk of relapse. This case highlights a clinically important mismatch between outward functioning and actual risk: a high-achieving adolescent from a supportive family developed severe, concealed polysubstance use that remained unrecognised for months. Clinically, physicians often do not suspect drug use in adolescents who show academic success and come from supportive families. Similarly, families rarely suspect drug use when their child performs well academically; they may mistakenly see autonomy and independence as signs of safety. For this patient, easy digital access and family trust enabled him to escalate his use of multiple drug classes for about a year before he disclosed it. Notably, several risk factors existed but remained hidden. For example, the patient's relationship between social anxiety and stimulant use was difficult to untangle. He used modafinil to cope with social stress, but the stimulant exposure may also have triggered or worsened his anxiety. This use may have been associated with severe anxiety attacks and suicidal thoughts.4, 5 Moreover, his desire for quick success may have reinforced his repeated use of stimulants and anabolic steroids, whereas the chronic pain associated with AS likely contributed to his opioid use and may have intensified his distress during withdrawal. These factors help explain how a high-achieving profile can coexist with severe substance abuse. From a clinical perspective, this case highlights the importance of directly asking adolescents about nonmedical drug use when they present with fluctuating anxiety, insomnia, autonomic symptoms or suicidality that responds poorly to standard treatment. However, fear of punishment may contribute to concealment of substance use, particularly when adolescents worry that disclosure could result in serious school-related consequences. Disclosure may also raise legal and institutional concerns for clinicians, especially when online-acquired substances include novel psychoactive compounds not clearly covered by routine drug control frameworks. Early involvement of addiction medicine specialists, together with clear institutional protocols and a nonjudgemental clinical stance, may help clinicians navigate these challenges. Assessment of drug use is further limited by routine toxicology testing. Standard urine immunoassays may miss novel psychoactive substances, rare analogues, adulterants and some emerging synthetic opioids obtained online. When the history strongly suggests online procurement and polydrug use, expanded toxicology testing, such as liquid chromatography-mass spectrometry, may therefore be helpful, where available. However, access to such testing often depends on local laboratory capacity, resources and regulations.6 Complementary sample types may also add value. Oral fluid can help detect recent exposure, whereas hair testing may provide information on longer-term patterns of use. Together, these approaches may improve assessment in complex cases involving online-sourced polysubstance exposure, although their feasibility varies by setting.7, 8 Cases involving substances obtained online often require more than short-term detoxification alone. In this case, persistent symptoms and incomplete functional recovery remained evident throughout follow-up, despite reported abstinence from illicit drugs.9 Although school attendance resumed, close parental supervision at home remained necessary. These findings highlight the need for a multidisciplinary approach that integrates psychiatric care, addiction treatment, pain management and family intervention. Digital therapeutics may have an adjunctive role in the long-term management of addictive disorders and relapse prevention.10 Finally, this case carries important public health implications. Online drug access lowers the social visibility of substance use, broadens the range of exposure risks and raises the likelihood of adulterated or novel substances. Prevention efforts should address both clinician awareness (considering online drug procurement as a risk factor) and family education (teaching that autonomy does not eliminate risk). Public health efforts should also consider measures to limit the online availability of psychoactive substances. Nevertheless, this case report has several limitations, which also point to priorities for future care and research. The patient's exposure history depended largely on self-report and limited toxicology testing. This may have introduced recall and misclassification bias and reduced certainty regarding the specific agents involved and the timing of exposure. Moreover, the patient's premature discharge curtailed standardised inpatient observation, limiting the characterisation of withdrawal trajectories and treatment response. Although the follow-up period extended to 2 years, we relied mainly on patient interviews and family supervision for relapse monitoring; therefore, we cannot fully exclude intermittent use or drug substitution. Future work should incorporate more systematic multimatrix testing where feasible, as well as structured assessment of patient anxiety and family dynamics, to clarify directionality and targets for intervention. Despite these constraints, the case underscores how online access and performance pressures can enable concealed polysubstance use in high-achieving adolescents, and why early multidisciplinary management is essential. High-functioning adolescents can conceal polysubstance use; academic success and supportive families do not exclude the potential for substance use disorder. Online procurement increases drug exposure diversity and may involve adulteration or novel substances, which complicates detection and clinical management. Early multidisciplinary intervention and sustained follow-up matter greatly, particularly when chronic pain and anxiety drive substance initiation and relapse risk. Haifeng Zhang: Writing—original draft. Jialong Yi: Conceptualization; writing—review and editing. Xin Yu: Conceptualization; writing—review and editing; supervision. The authors are grateful to the patient for his consent to the writing of this case report. We thank Liwen Bianji (Edanz) (www.liwenbianji.cn) for editing the English text of a draft of this manuscript. This work was supported by the National Natural Science Foundation of China (82001150). The authors declare no conflicts of interest. Institutional approval was not required for a single anonymised case report in accordance with local policy. Written informed consent for publication of this case report was obtained from the patient. Haifeng Zhang, PhD, is an attending psychiatrist and assistant researcher at Peking University Sixth Hospital (Institute of Mental Health) in China. He obtained his PhD from Peking University in June 2019 and joined Peking University Sixth Hospital in 2019. He was a visiting trainee at University College London (UCL) in the United Kingdom during 2017–2018, supported by the China Scholarship Council. He completed standardised residency training in psychiatry in 2022 and obtained sub-specialty certification in geriatric psychiatry in 2025. His work integrates clinical practice and research in psychiatry, with expertise in neuropsychological assessment and neuroimaging approaches (e.g., functional MRI). He has participated in and/or led projects funded by the National Key R&D Program of China and the National Natural Science Foundation of China and has published multiple SCI-indexed papers as first author. He currently works in a general psychiatry inpatient ward at Peking University Sixth Hospital in China. All data relevant to the case report are included in the article. No additional data are available.
Zhang et al. (Wed,) studied this question.