# Anxiety modulates odour-linked brain connectivity and alcohol dependence risk

**Authors:** Khushbu Agarwal, Shefali Chaudhary, Valentina Parma, Yatan Pal Singh Balhara, Siddharth Sarkar

PMC · DOI: 10.1093/braincomms/fcag096 · Brain Communications · 2026-03-19

## TL;DR

This study shows that brain connectivity in smell-related circuits, influenced by anxiety, can predict the risk of alcohol dependence in young adults.

## Contribution

The study identifies a novel olfactory-limbic brain circuit and demonstrates how anxiety modulates its relationship with alcohol dependence.

## Key findings

- Connectivity between the precuneus–cingulate–accumbens and lateral occipital–hippocampal regions predicted lifetime alcohol dependence.
- Anxiety attenuated the risk effect of memory–reward circuitry and amplified risk via inferior frontal–parietal connectivity.
- Models showed strong performance in predicting alcohol-related outcomes using olfactory–reward connectivity features.

## Abstract

Alcohol use disorder (AUD) often emerges in adolescence and young adulthood, yet early neural predictors remain scarce. The olfactory system, through its connections with limbic and reward circuits, may provide a novel window into compulsive alcohol use. These brain regions are altered in AUD, but the role of olfactory pathways is poorly understood. Anxiety, a common comorbidity in AUD, may further modulate these circuits. This study investigates whether anxiety moderates the relationship between olfactory–reward brain connectivity and risky alcohol use and dependence. We analysed multimodal data from 1003 participants (22–36 years) in the Human Connectome Project. Resting-state functional MRI was used to compute functional connectivity among 41 regions implicated in olfaction, emotion, memory and reward. Odour identification scores were used to isolate connectivity features most correlated with olfactory function. These features served as predictors in ElasticNet and LASSO classification models for three alcohol-related outcomes: (i) lifetime Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV alcohol dependence, (ii) past-year risky drinking and (iii) past-week risky drinking. Models were trained on a Synthetic Minority Over-sampling Technique (SMOTE)-balanced subset (n = 704) and validated on a propensity score-matched holdout dataset (n = 83). Post hoc logistic regression examined anxiety (DSM scores) as a moderator of connectivity–outcome associations. Connectivity between the precuneus–cingulate–accumbens and lateral occipital–hippocampal regions predicted lifetime alcohol dependence (odds ratio = 1.12, P < 0.001). Models showed strong performance (area under the curve > 0.90 training; >0.80 validation). Anxiety moderated key connections: it attenuated the risk effect of memory–reward circuitry and amplified risk via inferior frontal–parietal connectivity. Odour identification scores did not differ across alcohol risk groups. Model performance replicated in a propensity-matched validation set. Our findings reveal a novel olfactory–limbic circuit predictive of alcohol dependence in young adults. Anxiety modulates these pathways, suggesting dynamic brain–behaviour interactions that may underlie individual vulnerability and resilience. This framework supports the integration of sensory, emotional and transdiagnostic features in neurobiological models of addiction.

Agarwal et al. report that brain connectivity within olfactory networks, together with anxiety, predicts vulnerability to alcohol dependence. Using functional MRI and behavioural data, they show that altered smell-related brain circuits, moderated by anxiety, can serve as potential markers for risky drinking and alcohol-related problems.

Graphical Abstract

## Linked entities

- **Diseases:** alcohol dependence (MONDO:0002046)

## Full-text entities

- **Diseases:** AUD (MESH:D000437), addiction (MESH:D019966), Anxiety (MESH:D001007)
- **Chemicals:** alcohol (MESH:D000438)
- **Species:** Homo sapiens (human, species) [taxon 9606]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC13023364/full.md

## Figures

2 figures with captions in the complete paper: https://tomesphere.com/paper/PMC13023364/full.md

## References

53 references — full list in the complete paper: https://tomesphere.com/paper/PMC13023364/full.md

---
Source: https://tomesphere.com/paper/PMC13023364