Association between sleep duration, depression and breast cancer in the United States: a national health and nutrition examination survey analysis 2009–2018

Figures & data

Table 1. Characteristics of included samples by breast cancer.

Characteristics	Control (N = 1526)	Breast cancer (N = 263)	P for t or χ2
Age, year	65.6 ± 14.0	67.0 ± 11.4	0.077
PIR	2.69 (1.61%)	2.59 (1.64%)	0.401
BMI, kg/m^2			0.006
normal	400(26.2%)	59(22.4%)
overweight	578(37.8%)	128(48.6%)
obesity	548(35.9%)	76(29%)
Race			<0.001*
Mexican American	85(5.57%)	28(10.6%)
Other Hispanic	97(6.36%)	21(7.98%)
Non-Hispanic White	1019(66.8%)	140(53.2%)
Non-Hispanic Black	223(14.6%)	53(20.2%)
Other Race	102(6.68%)	21(7.98%)
Education			0.727
Less Than 9^th Grade	143(9.37%)	23(8.75%)
9-11^th Grade	188(12.3%)	31(11.8%)
High School	327(21.4%)	60(22.8%)
Some College	465(30.5%)	85(32.3%)
College Graduate or above	403(26.4%)	64(24.38%)
Marital Status			.
Married	829(54.3%)	109(41.4%)
Widowed	262(17.2%)	73(27.8%)
Divorced	212(13.8%)	45(17.1%)
Separated	47(3.08%)	13(4.94%)
Never married	112(7.34%)	17(6.46%)
Living with partner	58(3.8%)	5(1.9%)
Refused	6(0.39%)	1(0.38%)
Alcohol consumption			0.057
Yes	261(17.1%)	8(3.12%)
No	1265(82.9%)	255(96.9%)
Smoking			<0.001*
Yes	893(58.5%)	99(37.6%)
No	633(41.5%)	164(62.4%)
Hypertension			0.899
Yes	56(3.73%)	8(3.23%)
No	1470(96.3%)	255(96.8%)
Diabetes			0.211
Yes	330(21.6%)	61(23.2%)
No	1136(74.4%)	187(71.1%)
Borderline	60(3.94%)	15(5.7%)
Sleep quality			0.102
<7 h	818(53.6%)	159(60.5%)
7-9 h	476(31.2%)	68(26%)
>9 h	232(15.2%)	36(13.5%)
Depression			0.987
No	1235(80.9%)	212(80.5%)
Yes	291(19.1%)	51(19.5%)

Categorical variables are displayed as numbers (percentages); Normally distributed continuous variables are presented as means ± SD; PIR: Family income-to-poverty ratio.

Table 2. Logistic risk analysis for sleep duration, depression and breast cancer.

	Model 1	Model 2	Model 3	Model 4
Depression
No	1.00	1.00	1.00	1.00
Yes	1.13(0.64,2.01)	1.26(0.72,2.2)	1.43(0.84,2.84)	1.99(1.55,3.51)*
Sleep duration
<7h	1.68(0.69,1.79)	1.52(0.96,1.77)	1.45(0.82,1.63)	1.25(0.85,1.37)
7-9h	1.59(0.72,1.65)	1.43(0.75,1.65)	1.35(0.78,1.52)	1.18(0.88,1.25)
≥9h	1.48(0.76,1.58)	1.34(0.85,1.51)	1.23(0.76,1.46)	1.05(0.95,1.15)

Data are presented as OR (95% CI); OR: odds ratio; CI: confidence interval.

Model 1= No adjust;.

Model 2= sleep duration (When studying depression and breast cancer)/ depression (When studying sleep duration and breast cancer).

Model 3= Model 2 plus adjusted for age (years, continuous), education (less Than 9^th grade, 9-11^th grade, high school, some college, college graduate or above), race (non-Hispanic white, non-Hispanic black, Mexican American, other), PIR (continuous), BMI (<25, 25 to 30, and ≥30 kg/m2), Marital Status (married, widowed, divorced, separated, never married, living with partner, refused), hypertension (yes or no) and diabetes (yes, no, borderline).

Model 4 = Model 3 plus adjusted for smoking (yes or no) and alcohol consumption (yes or no).

*p < 0.05.

Figure 1. Receiver operating characteristic (ROC) curve of caffeine intake, PA and prostate cancer by using four machine learning algorithms.

Figure 2. Calibration curve of 6 machine learning algorithms.

Figure 3. Comparison of accuracy and kappa value among 4 machine learning algorithms.

Table 3. Detailed comparison of ROC curve related indicators of six machine-learning algorithms.

Item	Adaboost	ANN	RF	Boost tree	SVM	XGBoost
Sensitivity	0.970 95%CI: 0.94–0.98	0.972 95%CI: 0.94–0.99	0.924 95%CI: 0.88–0.95	0.953 95%CI: 0.91–0.97	0.91 95%CI: 0.86–0.94	0.915 95%CI: 0.87-0.95
Specificity	0.635 95%CI: 0.61–0.66	0.594 95%CI:0.57–0.62	0.663 95%CI: 0.64–0.69	0.667 95%CI: 0.64–0.69	0.574 95%CI: 0.55–0.60	0.599 95%CI: 0.57-0.63
Prediction	0.314 95%CI: 0.28–0.35	0.292 95%CI: 0.26–0.33	0.322 95%CI: 0.29–0.36	0.331 95%CI: 0.29–0.37	0.270 95%CI: 0.24–0.30	0.283 95%CI: 0.25-0.32
Negative prediction	0.992 95%CI: 0.98–1.00	0.992 95%CI: 0.98–1.00	0.981 95%CI: 0.97–0.99	0.988 95%CI: 0.98–0.99	0.974 95%CI: 0.96–0.98	0.976 95%CI: 0.96-0.98
AUC	0.84 95%CI: 0.81–0.87	0.83 95%CI: 0.79–0.87	0.84 95%CI: 0.81–0.87	0.83 95%CI: 0.79–0.87	0.80 95%CI: 0.76–0.84	0.82 95%CI: 0.78-0.86
Matthews correlation coefficient	0.43	0.401	0.422	0.444	0.343	0.364
False positive rate	0.365	0.406	0.337	0.333	0.426	0.401
F1 value	0.474	0.45	0.477	0.491	0.416	0.432
True Positive	255	205	195	201	192	193
False Positive	557	496	411	407	520	490
True Negative	969	725	810	814	701	731
False Negative	8	6	16	10	19	18

Data availability statement

All the data could be accessed from the website of the American Centers for Disease Control and Prevention (https://www.cdc.gov/nchs/nhanes).