This Research aims to investigate the impact of Vertebral Subluxation, or dysfunction, on stress levels. The study will utilise machine-learning models and time series analysis to predict stress based on subluxation severity and to analyse its evolution over time. Participants underwent chiropractic sessions, with stress levels gauged using Heart Rate Variation (HRV), sputum, and hair cortisol. Key outcomes will determine whether chiropractic care influences stress levels and elucidate the relationship between subluxation and stress. The integration of Artificial Intelligence into chiropractic research aims to provide personalised interventions and deepen the understanding of patient care.

Grant Value: $19,947.52 (funding for the first year)
Chief Investigator: Dr Imran Amjad & Dr Imran Khan Niazi  – New Zealand College of Chiropractic
Status: Complete

Clinical Applications:

This study examined the effects of chiropractic care on stress, using both physiological (cortisol) and psychological (DASS-21) measures in adults and children over 12 weeks. Linear mixed-effects models showed that baseline cortisol and DASS scores were stronger predictors of stress outcomes than subluxation scores or treatment groups, which was supported by the results from machine learning models (Random Forest, Gradient Boosting, and SVR). While cortisol levels increased post-intervention, DASS scores declined, suggesting differing physiological and psychological responses. However, the results for the children’s data were inconclusive. Overall results suggested that cortisol response may be more biologically driven and not easily altered through short-term chiropractic care, highlighting the need for longer-term studies and broader biomarker inclusion.

Researcher updates:

February 2026
Data analysis, machine learning, and AI modelling have been completed, and the results have been finalised.  A manuscript is being prepared for submission.
July 2025
The data analysis, machine learning, and AI modelling have been completed, and the manuscript is now being written with a completion date set for November 2025.

January 2025
Data preprocessing (extraction of Subluxation from clinical notes ) and conversion of the non-structured data into structured data has commenced and it should be finished by June 2025.  Once the data is in a structured format, the attention will be focussed on looking at the predictive AI-based modelling.