Table 5.

Associations between straightener, relaxer, or pressing products use in the 12 months prior to enrollment and uterine cancer in the Sister Study stratified by participant characteristics

Potential modifiera	Ever		Frequentb		P trend e
	Case	HR (95% CI)d	Case	HR (95% CI)d
	No. (%)c		No. (%)c
Race and ethnicityf
African American/Blackg	23 (79.3)	1.66 (0.67 to 4.09)	17 (58.6)	2.12 (0.83 to 5.39)	.09
Non-Hispanic White	12 (3.8)	1.94 (1.09 to 3.47)	7 (2.2)	2.66 (1.25 to 5.67)	.011
Pheterogeneityh		.77		.71
BMI
Nonobese (<30 kg/m2)	16 (7.7)	1.81 (1.04 to 3.17)	10 (5.1)	2.48 (1.22 to 5.03)	.010
Obese (≥30 kg/m2)	22 (12.6)	1.68 (0.90 to 3.14)	16 (9.2)	2.40 (1.19 to 4.85)	.022
Pheterogeneityh		.83		.94
Physical activityi
Low	22 (15.3)	2.59 (1.46 to 4.61)	15 (10.4)	3.72 (1.91 to 7.25)	<.001
High	16 (7.1)	1.30 (0.71 to 2.38)	11 (4.9)	1.86 (0.89 to 3.87)	.16
Pheterogeneityh		.05		.09

^aMissing: straighteners, relaxers, or pressing products (1 case, 98 overall). BMI = body mass index; CI = confidence interval; HR = hazard ratio.

^bDefinition of frequent use: >4 times in the past 12 months (permanent dyes; semipermanent dyes; and straighteners, relaxers, or pressing products).

^cParticipants with complete confounder information. Percentage of ever or frequent users among cases in each stratum.

^dCompared with never users. Accounted for age by using age as the timescale, where the participants were followed from age at baseline until age at end of follow-up. Adjusted for race and ethnicity (African American/Black, Hispanic/Latina non-Black, non-Hispanic White, other), education (high school or less, some college, college and above), BMI (restricted cubic spline, continuous, kg/m²), physical activity (metabolic equivalent [MET] hours per week, continuous), menopausal status at enrollment (premenopausal, postmenopausal), parity (0, 1, 2, ≥3), smoking (never, past, or current), alcohol (never or past, current <1 drink, current ≥1 drinks), oral contraceptive use duration (none, <2 years, 2 to <10 years, ≥10 years), hormone replacement therapy (none, estrogen alone, estrogen plus progestin but never estrogen alone), age at menarche (<13 years, ≥13 years old).

^eAssessed using Wald tests for continuous variables of hair product use frequency (never, infrequent, frequent use).

^fNon-Black Hispanic or other race and ethnicity were excluded because of small sample size and limited power.

^gIncluding Hispanic and Latina Black.

^hAssessed using Wald tests for hair product-by-modifier interaction terms.

ⁱLow and high physical activities were categorized by 33rd percentile 32.7 MET hours per week.