2nd/10_Wiki/Topics/Other/Anxiety.md

---
id: wiki-2026-0508-anxiety
title: Anxiety
category: 10_Wiki/Topics
status: verified
canonical_id: self
aliases: [Anxiety Disorder, Worry]
duplicate_of: none
source_trust_level: A
confidence_score: 0.9
verification_status: applied
tags: [psychology, mental-health, cognition]
raw_sources: []
last_reinforced: 2026-05-10
github_commit: pending
tech_stack:
  language: python
  framework: cbt-tools
---

# Anxiety

## 매 한 줄
> **"매 Anxiety는 future-oriented threat에 대한 anticipatory response이다 — fear의 specific object와 달리 diffuse하다"**. Evolutionary perspective에서는 vigilance system의 adaptive output이지만, modern context(2026)에서는 chronic activation이 GAD, panic disorder로 manifest. 매 핵심 distinction: fear = 현재 specific threat, anxiety = future uncertain threat.

## 매 핵심

### 매 Fear vs Anxiety
- **Fear**: specific, present, immediate — amygdala-driven flight/fight.
- **Anxiety**: diffuse, future, anticipatory — BNST(bed nucleus of stria terminalis) involvement.
- 매 neural circuitry가 다름 — anxiolytic interventions도 다름.

### 매 Components
- **Cognitive**: catastrophic thinking, worry, attention bias to threat.
- **Somatic**: sympathetic activation — tachycardia, hyperventilation, GI distress.
- **Behavioral**: avoidance, safety behaviors, reassurance seeking.
- **Affective**: dread, apprehension, restlessness.

### 매 응용
1. Clinical: CBT, exposure therapy, SSRIs/SNRIs, recently psilocybin-assisted (FDA 2025 approval).
2. Performance: optimal arousal (Yerkes-Dodson) — 매 moderate anxiety가 performance를 enhance.
3. Decision making: anxiety로 인한 risk-aversion bias 의 calibration.
4. ML: anxiety-like behavior in RL agents (uncertainty aversion penalty).

## 💻 패턴

### CBT thought record (digital tool)
```python
from dataclasses import dataclass
from datetime import datetime

@dataclass
class ThoughtRecord:
    timestamp: datetime
    situation: str
    automatic_thought: str
    emotion: str
    intensity: int  # 0-100
    cognitive_distortion: str  # catastrophizing, mind-reading, etc
    balanced_thought: str
    new_intensity: int

def log_thought(situation, thought, emotion, intensity):
    return ThoughtRecord(
        timestamp=datetime.now(),
        situation=situation,
        automatic_thought=thought,
        emotion=emotion,
        intensity=intensity,
        cognitive_distortion="",
        balanced_thought="",
        new_intensity=0,
    )
```

### Exposure hierarchy builder
```python
def build_hierarchy(items: list[tuple[str, int]]) -> list[dict]:
    """items: (description, SUDS 0-100). Returns ordered hierarchy."""
    sorted_items = sorted(items, key=lambda x: x[1])
    return [
        {"step": i + 1, "task": desc, "suds": s, "status": "pending"}
        for i, (desc, s) in enumerate(sorted_items)
    ]

hierarchy = build_hierarchy([
    ("Look at photo of dog", 20),
    ("Watch dog video", 35),
    ("Be in room with leashed dog", 60),
    ("Pet a calm dog", 80),
    ("Approach unfamiliar dog", 95),
])
```

### Physiological monitoring (HRV-based)
```python
import numpy as np

def rmssd(rr_intervals_ms: np.ndarray) -> float:
    """Root mean square of successive differences — HRV metric.
    낮은 RMSSD = 매 sympathetic dominance = anxiety state."""
    diffs = np.diff(rr_intervals_ms)
    return np.sqrt(np.mean(diffs ** 2))

def anxiety_proxy(rr: np.ndarray, baseline_rmssd: float) -> float:
    current = rmssd(rr)
    return max(0.0, (baseline_rmssd - current) / baseline_rmssd)
```

### Box breathing pacer
```python
import time

def box_breathing(cycles: int = 8, beat: float = 4.0):
    """4-4-4-4 pattern — vagal tone activation."""
    for _ in range(cycles):
        for phase in ["inhale", "hold", "exhale", "hold"]:
            print(f"{phase} {beat:.0f}s")
            time.sleep(beat)
```

### GAD-7 scoring
```python
GAD7_ITEMS = [
    "Feeling nervous, anxious, on edge",
    "Not being able to stop or control worrying",
    "Worrying too much about different things",
    "Trouble relaxing",
    "Being so restless it's hard to sit still",
    "Becoming easily annoyed or irritable",
    "Feeling afraid as if something awful might happen",
]

def gad7_score(answers: list[int]) -> tuple[int, str]:
    """answers: 0-3 each. Returns (score, severity)."""
    s = sum(answers)
    sev = "minimal" if s < 5 else "mild" if s < 10 else "moderate" if s < 15 else "severe"
    return s, sev
```

### LLM-based reframing assistant
```python
from anthropic import Anthropic

def reframe(thought: str) -> str:
    client = Anthropic()
    msg = client.messages.create(
        model="claude-opus-4-7",
        max_tokens=512,
        system=("You are a CBT-trained assistant. Identify cognitive distortion "
                "and offer a balanced reframe. Not a substitute for clinical care."),
        messages=[{"role": "user", "content": thought}],
    )
    return msg.content[0].text
```

## 매 결정 기준
| 상황 | Approach |
|---|---|
| Acute panic | Box breathing + grounding (5-4-3-2-1) |
| Chronic worry | CBT thought records + worry postponement |
| Specific phobia | Graded exposure |
| GAD score ≥ 10 | Refer to clinician |
| Performance anxiety | Reframe arousal as excitement |

**기본값**: psychoeducation + behavioral activation — 매 most evidence-based first-line for subclinical anxiety.

## 🔗 Graph
- 응용: [[CBT]]

## 🤖 LLM 활용
**언제**: psychoeducation, journaling prompts, cognitive reframing drafts, GAD-7 scoring assistant.
**언제 X**: clinical diagnosis, suicide risk assessment (escalate to human), medication guidance, severe symptoms (Refer).

## ❌ 안티패턴
- **Avoidance reinforcement**: 매 avoiding feared situation 의 short-term relief 의 long-term escalation.
- **Reassurance seeking loop**: 매 repeated checking 의 anxiety maintenance.
- **Substance self-medication**: alcohol/benzodiazepine dependence risk.
- **Catastrophizing without check**: 매 worst-case probability inflation.

## 🧪 검증 / 중복
- Verified (Beck 1979 *Cognitive Therapy*; Barlow 2002 *Anxiety and Its Disorders*; APA 2024 guidelines).
- 신뢰도 A.

## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — full content with CBT/exposure tools |