Adapting social anxiety treatment: autism, teens, older adults, culture, depression and alcohol

The short answer

If a client is depressed, drinking heavily or having thoughts of self-harm, check risk first and share crisis line information.

Autistic clients

How common. In a review of autistic adults, 27% had a current anxiety disorder and 42% a lifetime one. Questionnaire-based estimates were higher, which matters: social anxiety scales ask about avoiding social situations, and an autistic person may avoid them for sensory or communication reasons rather than fear of judgement.

Assessment. Professionals in a focus-group study found it hard to separate social anxiety from core autistic features. They named diagnostic overshadowing (putting everything down to autism) and alexithymia (difficulty naming one’s own emotions). Practical points:

• Ask what the person fears will happen. Social anxiety centres on being judged or rejected. Avoiding a party because it is loud and unpredictable is not the same thing.
• Be careful with cut-offs on screening scales. They were not built for autistic people.
• Consider camouflaging (masking autistic traits to fit in). It can work like a safety behaviour, but it is also a reasonable response to stigma.

Treatment evidence.

• Children: in a US trial of 167 autistic children aged 7 to 13, CBT adapted for autism beat standard CBT and usual care on anxiety. Response was 92.4% with adapted CBT, 81.0% with standard CBT and 11.1% with usual care. This was anxiety in general, not social anxiety only.
• Adults: in an uncontrolled Australian study of 71 autistic adults, an 8-week modified group CBT for social anxiety reduced symptoms moderately and did not increase camouflaging.
• A pilot of MDMA-assisted therapy in 12 autistic adults is often quoted. It was tiny and experimental, and is not a treatment option.

Adapting CBT (drawn from these studies, not tested component by component): more structure and written or visual material; concrete behavioural experiments rather than abstract thought records; agree goals that reduce fear and distress rather than make the person look less autistic; involve family where the person wants it; plan for sensory load during exposure.

Adolescents

• CBT works for young people with anxiety. The Cochrane review of 87 trials found remission of 49.4% with CBT vs 17.8% on waiting list (number needed to treat 3), but no clear advantage over usual care.
• Social anxiety is the hardest anxiety disorder for generic CBT. In 1,519 children and adolescents, those with social anxiety as the main problem responded less well than those with generalized anxiety. This argues for treatment built for social anxiety.
• Social-anxiety-specific cognitive therapy, online. In a trial of 43 adolescents aged 14 to 18, 77% lost the diagnosis after internet cognitive therapy, vs 14% on waiting list, and 91% had lost it at 6 months. Small, from the developers’ group, but promising.
• Parents. Across 25 systematic reviews, 21 found no difference between ways of involving parents. Parent-only CBT tended to be weakest.
• NICE recommends individual CBT (8 to 12 sessions of 45 minutes) or group CBT (8 to 12 sessions of 90 minutes), with exposure, social skills training and parent involvement especially for younger children. It says not to offer medication routinely to children and young people.
• VR for teenagers has no completed randomized trial. One with 120 adolescents is under way.

Older adults

In a Canadian survey of 12,792 people aged 55 and over, lifetime prevalence of social phobia was 4.94% and 12-month prevalence 1.32%. Current social phobia declined with age. Social anxiety in later life is easy to miss because withdrawal is put down to retirement, bereavement, hearing loss or low mood.

No randomized trial of treatment designed for older adults with social anxiety disorder was found. Registry searches mostly return trials about loneliness, which is a different problem. In practice, standard CT-SAD or CBT is the reasonable default, adapted for hearing, mobility and any cognitive changes.

Culture and taijin kyofusho

Expression differs across cultures. In surveys across 28 countries, social anxiety disorder was least common in low and lower-middle income countries and in Africa and the Eastern Mediterranean. Asian samples report the lowest rates and Russian and US samples the highest. Part of this may be how the questions fit local norms.

Taijin kyofusho is a fear of offending or making others uncomfortable, for example through one’s gaze, body odour or blushing. In a comparison of US and Japanese students, it loaded on a separate factor from Western-style social phobia. DSM-5 describes it as a cultural concept related to social anxiety disorder, but evidence that it is a distinct syndrome is mixed. Cases have been described in India, outside Japan. Standard scales do not ask about fear of offending others, so ask directly. See the plain-language page on taijin kyofusho.

Treatment across cultures.

• Internet CT-SAD in Hong Kong (44 people) had a large advantage over waiting, with minimal cultural adaptation.
• A translated and culturally adapted iCT-SAD in Japan (single arm, 15 people) gave 60% remission at post-treatment and measured taijin kyofusho symptoms as a secondary outcome.
• In a Japanese trial of 42 people still symptomatic after antidepressants, adding 16 weeks of CBT to usual care gave remission of 47.6% vs 0.0%.

Trials from sub-Saharan Africa and South Asia are almost absent. Adapting to culture usually means changing examples, values about modesty and hierarchy, and the feared consequences (“I will shame my family” rather than “they’ll think I’m boring”), while keeping the core model.

Comorbid depression

• In a Munich study of 2,548 young people, social anxiety disorder predicted later depression (odds ratio 3.5). When both were present, depression was more likely to persist (OR 2.3) and suicide attempts were more likely (OR 6.1).
• Comorbid depression predicted poorer outcome in a trial of CBT and ACT, and comorbid depression or substance use worsened the long-term course in a 12-year clinical cohort.
• The BAP guideline notes that people with social anxiety often present with depression rather than with social fears, and advises asking about social anxiety when someone presents with depression.

In practice: assess risk; if depression is severe, treat it first or alongside; once mood allows, target the social anxiety directly, because it often came first and keeps the depression going through isolation.

Comorbid alcohol use

• In a US survey of 43,093 adults, social anxiety disorder was linked to alcohol dependence (odds ratio 2.8).
• In young people, a study combining 70 studies found social anxiety slightly negatively linked to how much people drank (r = −0.05) but positively linked to alcohol problems (r = 0.13). Socially anxious young people may not drink more, but they are more likely to run into problems with alcohol.
• BAP advises asking about social anxiety in people who misuse alcohol or cannabis.

Treatment evidence is thin.

The Randall result is a single trial, but it suggests that loading alcohol and social anxiety work into the same short course can overwhelm people. Sequence the work: stabilize drinking enough to engage in exposure, then treat the social anxiety. Drinking before a feared situation works as a safety behaviour, and dropping it is part of exposure. Avoid benzodiazepines in people who drink heavily. Any medicine should be prescribed and monitored by a doctor.

What we still don’t know

• Whether standard models of social anxiety fit autistic adults, and what treatment works best for them.
• Any trial in older adults.
• Trials in low- and middle-income countries, and measures that capture fear of offending others.
• The best order and combination of treatments for social anxiety with alcohol problems.