Individual differences in physiological flexibility predict spontaneous avoidance

Abstract

People often regulate their emotions by resorting to avoidance, a putatively maladaptive strategy. Prior work suggests that increased psychopathology symptoms predict greater spontaneous utilisation of this strategy. Extending this work, we examined whether heightened resting cardiac vagal tone (which reflects a general ability to regulate emotions in line with contextual demands) predicts decreased spontaneous avoidance. In Study 1, greater resting vagal tone was associated with reduced spontaneous avoidance in response to disgust-eliciting pictures, beyond anxiety and depression symptoms and emotional reactivity. In Study 2, resting vagal tone interacted with anxiety and depression symptoms to predict spontaneous avoidance in response to disgust-eliciting film clips. The positive association between symptoms and spontaneous avoidance was more pronounced among participants with reduced resting vagal tone. Thus, increased resting vagal tone might protect against the use of avoidance. Our findings highlight the importance of assessing both subjective and biological processes when studying individual differences in emotion regulation.

Keywords:

• Emotion regulation
• Spontaneous avoidance
• Vagal tone
• Psychopathology
• Context

Acknowledgements

The authors would like to thank Kara Christensen, Emily Dunn, Andre Plate and Anne Wilson for their insightful comments on previous versions of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The low-intensity pictures were: 9330, 9290, 1945, 9830, 9320, 9373, 7361, 7360, 3061 and 9490. The moderate-intensity pictures were: 9042, 7380, 3015, 8230, 9300, 9405, 9570, 2352.2, 3250 and 3140. The high-intensity pictures were: 3150, 3110, 3120, 3071, 3400, 3130, 3060, 3080, 3010 and 3000.

² See Footnote 1.

³ We also assessed state-level reappraisal and state-level perseverative cognition, but these were not the focus of this investigation.

⁴ Resting log RSA was not associated with intensity at any time point, all ps> .545. It was significantly associated with avoidance during the first block, r = − .195, p = .026 and marginally during the second and third blocks, r = − .157, p = .067, r = − .172, p = .052.

⁵ Analyses conducted using repeated measures multivariate analysis of variance (MANOVA) yielded similar findings. In Step 1, spontaneous avoidance was positively predicted by anxiety and depression symptoms, F [1, 125] = 8.727, p = .004, partial η² = .065 and negatively predicted by log RSA, F [1, 125] = 4.556, p = .035, partial η² = .035. In Step 2, anxiety and depression symptoms remained a significant predictor, F [1, 124] = 8.675, p = .004, partial η² = .065, log RSA became a marginal predictor, F [1, 124] = 3.437, p = .066, partial η² = .027, and the interaction was non-significant, F [1, 124] = .136, p = .713, partial η² = .001.

Of importance, demographic variables associated with vagal tone (age, Liao et al., 1995; gender, Thayer et al., 1998; ethnicity, Wang et al., 2005) did not affect the main findings. We did not control for the use selective serotonin re-uptake inhibitors (e.g., Licht, de Geus, van Dyck, & Penninx, 2010) because only two participants reported taking them. Although women reported spontaneously using avoidance to a greater extent than men, Wald’s χ² = 6.476, b = − .221, SE = .087, p = .011, the inclusion of gender in the model had no impact on the main effects of symptoms, p = .002, or resting log RSA, p = .011. Neither age nor ethnicity (White vs. other) were significant predictors in their respective models, ps = .979 and .500, respectively, and did not impact the pattern of findings.

To ensure these findings were indicative of cardiac vagal influence generally, we re-ran the GEE analyses with a composite score of the four measures identified by Allen et al. (2007) as indexing parasympathetic functioning: high-frequency HRV, mean square difference (MSD), percentage of the absolute differences between consecutive IBIs that are greater than 50 ms (pnn50) and cardiac vagal index (CVI) (α = .969). Spontaneous avoidance was predicted positively by anxiety and depression symptoms, Wald’s χ² = 9.654, b = .192, SE = .062, p = .002, and negatively by the parasympathetically based HRV composite, Wald’s χ² = 4.415, b = − .091, SE = .0435, p = .036. In Step 2, the interaction was non-significant (p = .925) and both symptoms and HRV composite remained significant predictors, ps = .002 and .040, respectively.

⁶ We conducted additional analyses examining whether baseline log RSA interacts with time/segment in the prediction of spontaneous avoidance. This interaction was not significant (p = .756).

⁷ We also assessed state-level reappraisal and state-level perseverative cognition, but these were not the focus of this investigation.

⁸ Resting log RSA was marginally associated with avoidance during the early film, r = − .250, p = .069 (all other ps > .569).

⁹ We re-ran the analyses using repeated measures MANOVA. In Step 1, spontaneous avoidance was positively predicted by anxiety and depression symptoms, F [1, 50] = 6.087 p = .017, partial η² = .109 and not predicted by log RSA, F [1, 50] = 2.508, p = .120, partial η² = .048. In Step 2, anxiety and depression symptoms remained significant, F [1, 49] = 6.873, p = .012, partial η² = .123, log RSA became marginally significant, F [1, 49] = 3.397, p = .071, partial η² = .065, and the interaction was significant, F [1, 49] = 4.078, p = .049, partial η² = .077.

Neither age, gender, nor ethnicity (White vs. not) were significant predictors in their respective models, ps = .384–.404, and their inclusion in the model did not impact the overall pattern of findings.

We also re-ran the GEE analyses with a composite score of the four measures identified by Allen et al. (2007) as indexing parasympathetic functioning: HRV, MSD, pnn50 and CVI. The internal reliability was excellent (α = .961). Spontaneous avoidance was positively predicted by anxiety and depression symptoms, Wald’s χ² = 7.563, b = .050, SE = .018, p = .006. It was not predicted by the HRV composite, Wald’s χ² = 2.698, b = − .027, SE = .018, p = .100. In Step 2, the interaction between symptoms and HRV composite was significant, Wald’s χ² = 4.218, b = − .033, SE = .016, p = .040. We did not control for the use selective serotonin re-uptake inhibitors (e.g., Licht et al., 2010) because only two participants reported taking them.

¹⁰ See Footnote 4.

¹¹ We conducted additional analyses examining whether baseline log RSA interacts with time/segment in the prediction of spontaneous avoidance. This interaction was significant (p = .046), such that, baseline log RSA had a marginally negative association with spontaneous avoidance only during segment 2 (i.e., early film clip, p = .069). This suggests that the very beginning of the stimulus presentation might be critical for the link between vagal tone and the enactment of avoidance. Given the marginal nature of this finding, however, this interpretation is highly speculative and further work is needed to properly test this idea.

¹² See Footnote 1.

¹³ See Footnote 3.