TABLE 5.

Predicted Probabilities of TGNC Students’ Health Status and Health Care Utilization, Stratified by Perceived Gender Expression and Birth-Assigned Sex (n = 2095)

	TGNC Assigned Male at Birth (n = 661)		TGNC Assigned Female at Birth (n = 1434)
	Responses (n)	Predicted Probability (%)	Responses (n)	Predicted Probability (%)
Health status
General health is poor, fair, or good	P = .003a		P = .001b
Very feminine	82	49.2	163	54.0
Somewhat feminine	90	50.8	314	69.5
Equally feminine and masculine	177	49.2	564	70.4
Somewhat masculine	123	57.5	231	71.7
Very masculine	113	32.1	85	68.6
Long-term physical disability or health problems	P = .048		P = .418
Very feminine	82	19.5	163	22.7
Somewhat feminine	90	35.8	314	24.9
Equally feminine and masculine	177	27.1	564	25.0
Somewhat masculine	123	29.0	231	25.5
Very masculine	113	18.8	85	33.9
Long-term mental health problems	P < .001c		P < .001d
Very feminine	82	40.7	163	55.4
Somewhat feminine	90	44.6	314	68.1
Equally feminine and masculine	177	45.7	564	69.8
Somewhat masculine	123	42.8	231	76.7
Very masculine	113	15.8	85	73.2
Stayed home sick (last 30 days)	P = .210		P = .265
Very feminine	82	50.9	163	48.3
Somewhat feminine	90	57.5	314	55.1
Equally feminine and masculine	177	48.6	564	55.4
Somewhat masculine	123	48.2	231	48.6
Very masculine	113	40.7	85	55.0
Care utilization
Nurse office visits (last 30 days)	P = .947		P = .688
Very feminine	81	34.0	163	47.4
Somewhat feminine	90	34.2	321	42.1
Equally feminine and masculine	178	34.5	571	42.5
Somewhat masculine	123	31.6	235	46.2
Very masculine	113	30.4	86	45.4
Preventive medical checkup	P = .558		P = .175
Very feminine	81	51.5	163	69.4
Somewhat feminine	90	54.0	321	61.8
Equally feminine and masculine	178	57.6	571	61.3
Somewhat masculine	123	62.4	235	57.2
Very masculine	113	59.6	86	64.3
Preventive dental checkup	P = .059		P = .225
Very feminine	81	63.8	163	74.7
Somewhat feminine	90	63.9	321	75.2
Equally feminine and masculine	178	64.9	571	71.9
Somewhat masculine	123	74.3	235	67.2
Very masculine	113	77.0	86	76.1

In these analyses, we controlled for free and/or reduced-price lunch, race and/or ethnicity, grade, and school location. α level set at .05. Post hoc tests used Bonferroni's correction to adjust α for all pairwise comparisons. Numbers do not sum to the sample size because of missing data on at least 1 variable.

^aPost hoc analysis indicated significant differences between equally feminine and masculine and very masculine perceived gender expressions as well as somewhat masculine and very masculine perceived gender expressions.

^bPost hoc analysis indicated significant differences between somewhat feminine and very feminine perceived gender expressions, equally feminine and masculine and very feminine perceived gender expressions, as well as somewhat masculine and very feminine perceived gender expressions.

^cPost hoc analysis indicated significant differences between very masculine and all other perceived gender expression groups.

^dPost hoc analysis indicated significant differences between very feminine and all other perceived gender expression groups.