Chronotype changes after sex hormone use: A prospective cohort study in transgender users of gender-affirming hormones

Figures & data

Figure 1. Flowchart in- and exclusions in the RESTED study, sample sizes per group and per measurement time point and drop-outs and missing measurements per time point and group.

Table 1. Characteristics of the transmasculine (TM) and transfeminine (TF) groups reported per measurement time point.

	TM participants		TF participants
Measurement point	Baseline (n = 48)	3 months (n = 38)	Baseline (n = 44)	3 months (n = 37)
Age (years; median, interquartile range)	22 (19.8; 24)	22 (19.5; 24)	26 (24; 31)	27 (24.8; 32.3)
Alcohol use Units per week (median, interquartile range) Missing	0 (0; 1) 1	0.5 (0; 0.5) 13	0.5 (0 to 2) 2	0.5 (0 to 2) 12
Smoking (n, %) Current smoker Not current smoker Missing	5 (10%) 42 (88%) 1 (2%)	3 (8%) 25 (66%) 10 (26%)	6 (14%) 38 (67%) 0 (0%)	1 (3%) 31 (84%) 5 (13%)
Psychotropic medication use (n, %)^a Any psychotropic medication Antidepressants Stimulants Antipsychotics	7 (15%) 5 (10%) 1 (2%) 2 (4%)	5 (13%) 4 (11%) 1 (3%) 1 (3%)	7 (16%) 5 (11%) 4 (9%) 2 (5%)	4 (11%) 3 (8%) 3 (8%) 1 (3%)
Cycle regulation use (n, %) Any cycle regulation Progestin-only forms Combined forms	21 (44%) 13 (27%) 8 (16%)	8 (21%) 7 (18%) 1 (3%)	– – –	– – –
Use of alarm clock on free days (n, %)	17 (35%)	14 (35%)	15 (34%)	12 (32%)
Work or school days per week (n, %) Less than 3 days of school or work 3 or more days of school or work	18 (37%)	11 (29%)	19 (43%)	19 (51%)
	30 (63%)	27 (71%)	25 (57%)	18 (48%)
Serum hormone levels (median, interquartile range)^b
Testosterone (nmol/L)	0.9 (0.6; 1.2)	19 (13.5; 24.5)	13 (10; 17.5)	0.55 (0.4; 0.8)
Estradiol (pmol/L)	114 (79; 208)	134 (104; 186)	71.5 (59.8; 89.5)	283 (180; 396.3)

^aParticipants could use multiple forms of psychotropic medication.

^bReference values for cisgender men: testosterone, 9 to 30 nmol/L, estradiol, 12 pmol/L to 126 pmol/L.

Reference values for premenopausal cisgender women: testosterone, 0.3 to 1.6 nmol/L, estradiol, 31 pmol/L to 2864 pmol/L (Amsterdam UMC Endocrinologisch Lab et al., 2023; Verdonk et al., 2019).

Figure 2. The mean sleep duration (line) and sleep debt-corrected midpoint of sleep on free days (point) before GAHT (baseline; displayed as the top orange bar) compared to after 3 months of GAHT (displayed as the bottom green bar). The Y-axis displays the groups, the X-axis displays the mean sleep duration (as a line) and mean sleep debt-corrected midpoint of sleep (as a point) in hours.

Table 2. Sleep duration and (sleep-corrected) Midpoint of Sleep on Free days (MSF_sc) obtained from the µMCTQ in the transmasculine (TM) group and the transfeminine (TF) group. All variables are reported in clock hours (hh:mm).

Outcome in hh:mm	Transmasculine group
	Predictor (n)	Mean (SD)	Unadjusted model Estimated change from baseline (95% CI: L; H) p-value)	Adjusted model Estimated change from baseline (95% CI: L; H) p-value)	Sensitivity model^a Estimated change from baseline (95% CI: L; H) p-value)
MSFsc	Baseline (48)	04:35 (01:26)	Reference	Reference	Reference
	3-month follow up (38)	04:54 (01:37)	00:20 (−00:02; 00:41) p = 0.084	00:24 (00:03; 00:45) p = 0.03	00:30 (00:03; 00:58) p = 0.039
Sleep duration	Baseline (48)	08:28 (01:14)	Reference	Reference	Reference
	3-month follow up (38)	08:42 (01:04)	00:13 (−00:08, 00:35) p = 0.22	00:14 (−00:08, 00:36) p = 0.22	00:27 (00:01; 00:55) p = 0.054
	Transfeminine group
Outcome in hh:mm	Time point, months after start GAHT (n)	Mean (SD) or median (IQR)^a	Unadjusted model Estimated change from baseline (95% CI: L; H) p-value)	Adjusted model Estimated change from baseline (95% CI: L; H) p-value)	Sensitivity model^a Estimated change from baseline (95% CI: L; H) p-value)
MSFsc	Baseline (44)	04:56 (01:27)	Reference	Reference
	3-month follow up (37)	04:39 (01:02)	−00:15 (−00:35; 00:05) p = 0.14	−00:21 (−00:38; −00:04) p = 0.023	−00:17 (−00:32; −00:03) p = 0.03
Sleep duration	Baseline (44)	08:26 (01:05)	Reference	Reference
	3-month follow up (37)	08:25 (00:54)	00:00 (−00:21; 00:21) p = 0.99	−00:02 (−00:22; 00:17) p = 0.83	−00:02 (−00:22; 00:17) p = 0.86

^aExcluding participants who use alarm clocks on free days. Sample sizes are n = 31 at baseline and n = 24 at 3 month-follow up for the TM group and n = 29 at baseline and n = 25 at 3 month-follow up for the TF group.

Figure 3. Sleep debt-corrected midpoint of sleep on free days (left) and sleep duration (right) per group and time point. Individual dots display every participants’ measurements, lines connect the participants’ measurements, the gray squares represent the means and the error bars represent the standard deviations per measurement time point and group.

Amsterdam UMC Endocrinologisch Lab, Hillebrand JJ, & Heijboer AC. 2023. Testosterone Reference Values. https://amc.getincontrol.eu/provision/overview/789ffb9e-2406-48e6-8601-e3f9b18d5b5c.

 Google Scholar

Verdonk SJE, Vesper HW, Martens F, Sluss PM, Hillebrand JJ, Heijboer AC. 2019. Estradiol reference intervals in women during the menstrual cycle, postmenopausal women and men using an LC-MS/MS method. Clin Chim Acta. 495:198–204. doi: 10.1016/j.cca.2019.04.062.

 PubMed Web of Science ®Google Scholar

Data availability statement

Data available on request