Exploratory heart rate variability patterns across sleep stages following a 7-day multicomponent mantra-based cognitive intervention: A pilot randomized pre–post controlled study
Ankita Garg, Varun Dutt, Laxmidhar Behera
Acta Psychologica October 1, 2026 DOI: 10.1016/j.actpsy.2026.107984 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Randomized controlled trial Peer reviewed |
|---|---|
| Sample size | 34 |
| Population | Healthy males aged 18–30 years |
| Intervention | Hare Krishna mantra-based cognitive therapy (HMBCT) |
| Duration | 7-day intervention period; overnight polysomnography at baseline and after the 7-day intervention |
| Measures | SDNN, RMSSD, SD1, SD2, pNN50, CVI |
| Key findings | Nine heart rate variability interactions between group and time reached the uncorrected p < .05 threshold, four in REM and five in N2 sleep, with none in N3, but none survived Benjamini-Hochberg false discovery rate correction across 30 tests (smallest q = 0.148). The authors conclude the findings are exploratory and require confirmation. |
Abstract
Background: Cognitive-contemplative practices could potentially affect autonomic function, but existing research has focused on waking measures of heart rate variability (HRV). It remains unclear whether changes associated with brief interventions differ across sleep stages.
Objective: To determine whether differences in HRV according to sleep stage would differ in a group undergoing a 7-day, multicomponent, Hare Krishna mantra-based cognitive therapy (HMBCT) intervention compared to a group engaging in their normal routine activities.
Methods: Thirty-four healthy males aged 18–30 years were randomized to HMBCT ( n = 17) or a usual-routine control condition ( n = 17). Overnight polysomnography with continuous ECG was conducted at baseline and after the 7-day intervention period. HRV variables were calculated using time-domain and Poincaré/geometric HRV indices separately for REM, N2, and N3 sleep stages. Group × Time interactions were analyzed using repeated measures ANOVA, and then subjected to Benjamini-Hochberg false discovery rate (FDR) correction in 30 primary tests.
Results: Nine interactions reached the nominal uncorrected threshold ( p < .05): four in REM (SDNN, RMSSD, SD1, SD2) and five in N2 (SDNN, RMSSD, pNN50, SD1, CVI), while none did so in N3. Effect sizes of REM/N2 interaction were estimated as partial η 2 p = 0.120–0.173. However, none remained statistically significant after FDR correction (smallest q = 0.148).
Conclusions: Nominal interactions were more pronounced in REM and N2 sleep stages, but none of the primary interactions survived multiple testing correction. Thus the results are exploratory and should be confirmed in other studies.