Table 2.

Descriptive data of population-mean cosinor parameters (by study group), respective model fit, and group comparison results (differential rhythmicity analyses)

	NSSI			HC			Cosinor model fita				Differential rhythmicity analysisb
							NSSI		HC		NSSI vs. controls
	N	Pop. Mean	Conf. Int	N	Pop. mean	Conf. int	F (DF1, DF2)	p-value	F (DF1, DF2)	p-value	F(DF1, DF2)	t(DF)	p-value	SMDc 95% CI
HR
MESOR	29	85.98	81.74–90.22	30	79.44	76.14–82.74	39.87 (2, 27)	< 0.001	89.85 (2, 28)	< 0.001	6.27 (1, 57)		0.015	0.64 [0.12, 1.16]
Amplitude	29	12.42	9.26–15.56	30	12.30	10.43–14.17						− 2.19 (57)	0.032
Acrophase	29	1.81	1.63–2.04	30	1.56	1.40–1.71					11.08 (1, 57)		0.002
HRV
MESOR	29	44.82	37.38–52.26	30	55.79	48.41–63.18	24.99 (2, 28)	< 0.001	10.35 (2, 27)	< 0.001	4.59 (1, 57)		0.036	− 0.56 [− 1.06; − 0.03]
Amplitude	29	13.90	7.68–20.11	30	14.45	10.30–18.60						1.86 (57)	0.034
Acrophase	29	5.02	− 4.75–5.33	30	4.73	4.57–4.90					5.69 (1, 57)		0.021

Bold font indicates statistical significance

HR heart rate, HRV heart rate variability (RMSSD, ms), MESOR midline estimating statistic of rhythm

^aBased on rhythm detection tests; significant results suggested that the respective cosinor models fit the present data well on a group level

^bTo test the hypothesis of significant group differences in diurnal variation of cardiac autonomic activity between NSSI group and HC, differential rhythmicity analysis was performed using the R-package ‘cosinor 2’ [22, 86, 96], including analysis of significant differences in amplitude and analysis of acrophase shift

^cHedge’s g reported