The selected SNPs and the two-sample MR analysis
We detected 16,359,424 SNPs at the age of first sexual intercourse in a European population. According to the above method, 196 SNPs were selected as instrumental variables in the population of the age of first sexual intercourse, which satisfies the local wide significance level (p<5*10-8)(Supplementary Table S1). Among this exposure datasets, there were 144-152 SNPs calculated for different cancer outcomes in the age of first sexual intercourse group(Supplementary Table S2). According to the IVW analysis results(Table 2), the age of first sexual intercourse has a significant causal effect on the risk of head and neck cancer(OR=0.997, 95%CI:0.9955-0.9984, p<0.0001) and lung cancer(OR=0.9928, 95%CI:0.9903-0.9954, p<0.0001). Therefore, the older the age of first sexual intercourse, the lower the risk of head and neck cancer and lung cancer. Similarly, there is a slight causal relationship between the age of first sexual intercourse and the liver cell cancer(OR=0.9993, 95%CI:0.9988-0.9999, p=0.0260) and cervical cancer(OR=0.9979, 95%CI:0.9960-0.9998, p=0.0366). Increasing the age of first sexual intercourse determined by genes can slightly reduce the risk of liver cell cancer and cervical cancer. At the same time, the age of first sexual intercourse has a slight causal effect on breast cancer(OR=1.005, 95%CI:1.0009-1.0092, p=0.0175) and prostate cancer(OR=1.1732, 95%CI:1.0118-1.3604, p=0.0344) . However, the older the age of first sexual intercourse, the higher the risk of breast cancer and prostate cancer(Supplementary Table S2).
Heterogeneity, pleiotropy, and leave-one-out
To evaluate the heterogeneity of the results, Cochrane’s Q test was applied . There was substantial heterogeneity determined in breast cancer(p=0.0011), lung cancer(p=0.0056), colorectal cancer(p=0.0067), prostate cancer(p=8.285E-21), non-melanoma skin cancer(p=1.4828E-136) and melanoma skin cancer(p=0.0006)(Supplementary Table S3). Therefore, the IVWMRE method was
alternatively performed to calculate the causal effect on these cancers (Supplementary Table S2). The significant causal effects disappeared among the age of first sexual intercourse and esophageal cancer (OR =0.9993, 95% CI: 0.9982-1.0005, p =
0.2733), bile duct cancer (OR = 0.9993, 95% CI:0.9985-1.0001, p = 0.0897), colorectal cancer (OR =1.0008, 95% CI: 0.9971–1.0045, p = 0.6656). Moreover, there was no horizontal pleiotropy detected in the exposure groups (p > 0.05), this indicates that there is no interference from mixed factors in the results, and the results are stable and reliable(Supplementary Table S4). The leave-one-out analysis was conducted for each cancer outcome to assess the stability of results in the exposure. Similarly, the results in the Leave-one-out analysis were consistent with that in the primary IVW analysis. Furthermore, the funnel plot was made to present the distribution balance of single SNP effects in the exposure (Supplementary Figures 2).
Table 2 result of IVWEF or IVWMRE analysis
Outcome
|
No. SNPs
|
Beta
|
Se
|
OR(95%CI)
|
P
|
Heterogeneity (P)
|
pleiotropy (P)
|
Head and neck
|
152
|
-0.003
|
0.0007
|
0.997(0.9955-0.9984)
|
<0.0001
|
0.3324
|
0.5036
|
esophageal
|
152
|
-0.0007
|
0.0006
|
0.9993(0.9982-1.0005)
|
0.2733
|
0.0935
|
0.2106
|
breast
|
151
|
0.0050
|
0.0021
|
1.0050(1.0009-1.0092)
|
0.0175
|
0.0012
|
0.9179
|
lung
|
152
|
-0.0072
|
0.0013
|
0.9928(0.9903-0.9954)
|
<0.0001
|
0.0056
|
0.5887
|
liver cell
|
144
|
-0.0007
|
0.0003
|
0.9993(0.9985-1.0001)
|
0.0260
|
0.8081
|
0.5574
|
bile duct
|
149
|
-0.0007
|
0.0004
|
0.9993(0.9985-1.0001)
|
0.0897
|
0.5265
|
0.3879
|
colorectal
|
152
|
0.0008
|
0.0019
|
1.0008(0.9971-1.0045)
|
0.6656
|
0.0067
|
0.1387
|
bladder
|
152
|
0.0001
|
0.0008
|
1.0001(0.9986-1.0017)
|
0.8847
|
0.4553
|
0.5101
|
prostate
|
152
|
0.1598
|
0.0755
|
1.1732(1.0118-1.3604)
|
0.0344
|
<0.0001
|
0.8562
|
cervical
|
151
|
-0.0021
|
0.0010
|
0.9979(0.9960-0.9999)
|
0.0366
|
0.6060
|
0.6281
|
ovarian
|
152
|
0.0014
|
0.0014
|
1.0014(0.9985-1.0042)
|
0.3420
|
0.1459
|
0.6201
|
non melanoma skin
|
152
|
0.0119
|
0.0083
|
1.0119(0.9957-1.0285)
|
0.1504
|
<0.0001
|
0.9724
|
melanoma skin
|
152
|
0.0011
|
0.0016
|
1.0011(0.9979-1.0042)
|
0.5045
|
0.0006
|
0.2912
|
MR: mendelian randomization; SNPs: single-nucleotide polymorphisms; SE: standard error; OR: odd ratio, CI:confidence interval. Bold values indicate statistical significance (p < 0.05).