The Association of age at Menarche and Adult Height With Mammographic Density in the International Consortium of Mammographic Density

Sarah V. Ward; Anya Burton; Rulla M. Tamimi; Ana Pereira; Maria Luisa Garmendia; Marina Pollan; Norman Boyd; Isabel dos-Santos-Silva; Gertraud Maskarinec; Beatriz Perez-Gomez; Celine Vachon; Hui Miao; Martín Lajous; Ruy López-Ridaura; Kimberly Bertrand; Ava Kwong; Giske Ursin; Eunjung Lee; Huiyan Ma; Sarah Vinnicombe; Sue Moss; Steve Allen; Rose Ndumia; Sudhir Vinayak; Soo-Hwang Teo; Shivaani Mariapun; Beata Peplonska; Agnieszka Bukowska-Damska; Chisato Nagata; John Hopper; Graham Giles; Vahit Ozmen; Mustafa Erkin Aribal; Joachim Schüz; Carla H. Van Gils; Johanna O. P. Wanders; Reza Sirous; Mehri Sirous; John Hipwell; Jisun Kim; Jong Won Lee; Caroline Dickens; Mikael Hartman; Kee-Seng Chia; Christopher Scott; Anna M. Chiarelli; Linda Linton; Anath Arzee Flugelman; Dorria Salem; Rasha Kamal; Valerie McCormack; Jennifer Stone

Disclosures

Breast Cancer Res. 2022;24(49)

Conclusions

In summary, we have shown in one of the largest international studies to date that later age at menarche and increased adult height are both positively associated with measures of MD. The associations observed for height are in line with previous findings for associations with breast cancer risk, but those for menarche are in the opposite direction. These findings suggest that the association of age at menarche with breast cancer risk is not likely mediated through MD. Whether the well-established positive association of height with breast cancer risk is driven in whole or in part by elevated MD warrants further study. These results suggest a complex relationship between growth and development, MD and breast cancer risk.

1 2 3 4 5 6

Next Section

Abbreviations
BMI: Body mass index; DA: Dense area; ICMD: International Consortium of Mammographic Density; IGF: Insulin-like growth factor; IGF-1: Insulin-like growth factor-1; MD: Mammographic density; PD: Percent density.

Acknowledgements
The Melbourne Collaborative Cohort Study acknowledges that their cases and their vital status were ascertained through the Victorian Cancer Registry (VCR) and the Australian Institute of Health and Welfare (AIHW), including the National Death Index and the Australian Cancer Database. Isfahan University acknowledges assistance from Dr. Vida Razavi and Dr. Shamila Razavi. Instituto Nacional de Salud Publica acknowledges the Ministry of Education of Mexico and the Institute for Social Security and Services for State Workers' Medical Directorate staff and regional office in Jalisco for technical and administrative support. Where authors are identified as personnel of the International Agency for Research on Cancer/World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy or views of these organizations.

Funding
This work was supported by the US National Cancer Institute at the National Institutes of Health [R03CA167771]; the International Agency for Research on Cancer; the University of Western Australia [Research Collaboration Award] and the Cancer Council Western Australia [Capacity Building and Collaboration Grant]. Original studies were supported, according to country by: Australia VicHealth; Cancer Council Victoria; Australian National Health and Medical Research Council [209057, 251,553 and 504711]; Australian National Breast Cancer Foundation [to JSt]; Canada the National Cancer Institute of Canada [to NFB]; Chile Fondecyt [11100238 to MLG, 1120326, 1130277, 3130532]; World Cancer Research Fund [2010/245]; Ellison Medical Foundation Grant [to AP]; Iran Isfahan University of Medical Sciences; Israel The Israel Cancer Association; Republic of Korea Asan Medical Center [2010-0811]; Malaysia Sime Darby LPGA Tournament; Ministry of Education University Malaya [High Impact Research Grant UM.C/HIR/MOHE/06]; University Malaya [Research Grant UMRG RP046B-15HTM]; Mexico National Council of Science and Technology (Mexico); the American Institute for Cancer Research [10A035]; Netherlands EPIC-NL-Europe against Cancer Programme of the European Commission (SANCO); Dutch Ministry of Health; Dutch Cancer Society; ZonMW the Netherlands Organisation for Health Research and Development; World Cancer Research Fund (WCRF); Poland Polish-Norwegian Research Programme [PNRF-243-AI-1/07]; Singapore National Medical Research Council [Clinician Scientist Award]; National University Cancer Institute Singapore (NCIS) Centre grant programme from National Medical Research Council; South Africa Pink Drive; Spain Spain's Health Research Fund (Fondo de Investigacion Santiaria) [PI060386 and PS09/0790]; Spanish Federation of Breast Cancer Patients (FECMA) [EPY1169-10]; Turkey-Roche Mustahzarlari San. A.S., Istanbul, Turkey; UK UK Engineering and Physical Sciences Research Council [EP/K020439/1 to JHi]; Breast Cancer Campaign [2007MayPR23], Cancer Research UK [G186/11 and C405/A14565]; Da Costa Foundation UK; USA National Cancer Institute [R01CA85265, R37 CA54281, R01 CA97396, P50 CA116201, R01 CA177150 and R01 CA140286]; Cancer Center Support Grant [CA15083; CA131332, CA124865, UM1 CA186107 and UM1 CA176726]; the Susan G. Komen Foundation.

Availability of data and materials
ICMD data cannot be deposited publicly as these data originate from 27 research institutions across 22 countries with different legal and ethical frameworks. Researchers seeking the analysis data set for this work are able to apply to the Environment and Lifestyle Epidemiology Branch at the International Agency for Research on Cancer for access (email: env@iarc.fr).

Declarations
Ethics approval and consent to participate
Ethics approvals for ICMD were obtained from the International Agency for Research on Cancer (IEC 12 ± 34). Each individual study obtained informed consent to participate from women included in the study and had received local ethical approval at the time of the original conduct of the study and again to contribute to the consortium.

Consent for publication
Not applicable.

Breast Cancer Res. 2022;24(49) © 2022 BioMed Central, Ltd.
Copyright to this article is held by the author(s), licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original citation.

Table 1. Characteristics of participants in the International Consortium on Mammographic Density
World Region^a:	All women (n = 10,681)	White (n = 4512)	Black (n = 938)	East Asian (n = 1837)	South Asian (n = 1133)	Mestizo (n = 333)	Hawaiian (n = 142)	Middle Eastern (n = 1786)
World Region^a:	n [column %] unless stated otherwise
Per cent density (cm²)
Median (IQR)	20 (11–31)	22 (16–35)	17 (10–26)	24 (16–35)	16 (8–25)	20 (13–27)	18 (9–32)	16 (8–24)
Dense area (cm²)
Median (IQR)	27 (16–41)	27 (17–42)	34 (20–49)	24 (15–35)	22 (12–35)	28 (18–42)	25 (12–38)	29 (16–43)
Age at mammogram (years)
Mean (SD)	52.7 (8.2)	53.8 (7.7)	52.4 (8.5)	52.1(8.5)	53.6 (7.2)	44.2 (6.4)	55.1 (7.5)	51.4 (8.5)
IQR	47–58	49–59	46–58	45–58	49–59	40–49	50–61	45–57
Age at menarche (yrs)^b
< 12.0	1470 [14]	838 [19]	80 [9]	166 [9]	153 [14]	77 [23]	20 [14]	136 [8]
12.0–12.9	2588 [24]	1031 [23]	116 [12]	437 [24]	300 [26]	87 [26]	56 [39]	561 [31]
13.0–13.9	2594 [24]	1185 [26]	167 [18]	350 [19]	283 [25]	73 [22]	0 [0]	536 [30)
14.0–14.9	2023 [19]	848 [19]	182 [19]	425 [23]	165 [15]	46 [14]	47 [33]	310 [17]
≥ 15.0	2006 [19]	610 [14]	393 [42]	459 [25]	232 [20]	50 [15]	19 [13]	243 [14]
Mean age (SD)	13.2 (1.7)	12.9 (1.6)	14.2 (2.1)	13.5 (1.8)	13.2 (1.7)	12.7 (1.7)	12.9 (1.8)	13.0 (1.4)
Height (cm)
< 155	2760 [26]	685 [15]	186 [20]	704 [38]	663 [59]	113 [34]	17 [12]	392 [22]
155–159.9	2609 [24]	905 [20]	211 [22]	604 [33]	268 [24]	125 [38]	15 [11]	481 [27]
160–164.9	2759 [26]	1299 [29]	287 [31]	395 [22]	136 [12]	70 [21]	42 [30]	530 [30]
> 165	2553 [24]	1623 [36]	254 [27]	134 [7]	66 [6]	25 [8]	68 [48]	383 [21]
Mean height (SD)	159.4 (7.2)	162.0 (6.9)	160.6 (7.6)	156.4 (5.6)	153.2 (6.8)	157.0 (5.3)	163.2 (6.0)	159.4 (6.3)
% of measured heights^c	66	57	39	57	85	98	0	100
BMI
≤ 20.0	642 (6)	208 (5)	9 (1)	284 (15)	95 (8)	3 (1)	2 (1)	41 (2)
20.1–25.0	3920 (37)	1787 (40)	171 (18)	1086 (59)	376 (33)	82 (25)	43 (30)	375 (21)
25.1–30.0	2342 (32)	1503 (33)	320 (34)	411 (22)	428 (38)	133 (40)	45 (32)	592 (33)
> 30.0	2687 (25)	1014 (22)	438 (47)	56 (3)	234 (21)	115 (35)	52 (37)	778 (44)
Mean BMI (SD)	27.0 (5.7)	26.7 (5.3)	30.3 (6.2)	23.2 (3.4)	26.3 (5.0)	28.7 (5.1)	28.9 (6.3)	29.6 (6.0)
Menopausal status
Pre-menopausal	4361 [41]	1761 [39]	328 [35]	799 [43]	324 [29]	261 [78]	35 [25]	853 [48]
Post-menopausal	6320 [59]	2751 [61]	610 [65]	1038 [57]	809 [71]	72 [22]	107 [75]	933 [52]
Ever used hormone therapy^d
Yes	2185 [25]	1387 [36]	133 [14]	384 [21]	82 [8]	27 [8]	78 [55]	94 [12]
No	6681 [75]	2456 [64]	802 [86]	1425 [79]	969 [92]	291 [92]	64 [45]	674 [88]
Parity
Nulliparous	1068 [10]	661 [15]	61 [7]	161 [9]	90 [8]	21 [6]	9 [6]	65 [4]
Parous	9613 [90]	3851 [85]	877 [94]	1,676 [91]	1,043 [92]	312 [94]	133 [94]	1721 [96]
Mean parity (SD)	2.6 (1.8)	2.4 (1.1)	3.3 (1.7)	2.5 (1.2)	3.9 (2.1)	2.9 (1.4)	3.3 (1.5)	3.6 (2.2)
Age at first birth (yrs)^b
Mean age (SD)	24.3 (5.1)	25.3 (4.7)	22.3 (4.8)	26.2 (4.2)	22.4 (6.2)	23.9 (5.0)	22.9 (4.1)	22.5 (5.2)
IQR	21–27.5	22–27.8	19–25	23–28.3	18–26	19.8–27	19–23	19–26

BMI body mass index, cm centimetres, IQR inter-quartile range, SD standard deviation, yrs years
^aPopulation groups and study source, by world region
White: White women in Australia (3 studies—Australian, Greek and Italian born), Canada, Netherlands, Norway, Poland, Spain, the UK (3 studies) and the USA (mainland: 3 studies, Hawaii: 1 study)
Black: Black women in Kenya, South Africa, the UK and the USA
East Asian: Chinese women in Hong Kong, Malaysia and Singapore; Japanese women in Japan, Korea and the USA (mainland: 1 study, Hawaii: 1 study)
South Asian: Malay women in Malaysia and Singapore; Indian women in India, Malaysia, Singapore and the UK
Mestizo: Mestizo women in Chile and Mexico
Hawaiian: Hawaiian women in USA (Hawaii: 1 study)
Middle Eastern: Egyptian women in Egypt; Arab women in Israel; Jewish women in Iran, Israel and Turkey
^bNote raw data recorded categorically by some studies and subsequently converted to an approximated numerical value
^cHeight data that were measured by the original study, as opposed to self-reported
^dAmong women with non-missing data. Missing data on hormone therapy were 17% overall, < 4% for black, Hawaiian, Mestizo and East Asian, 15% missing for white women and 57% for Middle Eastern. These women were included in a missing-hormone-therapy category in later regression models

Table 2. Association of per cent density and dense area with age at menarche in pooled ICMD studies
	No. of women	No. of measurements	Per cent density^a		Dense area^a,f
	No. of women	No. of measurements	Changes in square-root per cent density (SE)	P value	Changes in square-root dense area (SE)	P value
Age at menarche category (years)
< 12.0	1470	1722	0		0
12.0–12.9	2588	3036	0.006 (0.042)	0.889	− 0.032 (0.056)	0.575
13.0–13.9	2594	3067	0.064 (0.042)	0.130	0.053 (0.056)	0.346
14.0–14.9	2023	2397	0.057 (0.045)	0.207	0.034 (0.059)	0.569
≥ 15.0	2006	2364	0.128 (0.046)	0.005	0.166 (0.061)	0.007
Linear associations with a 1 year increase in age at menarche
All women—Model A^c
All, BMI and height adjusted	10,681	12,586	0.023 (0.008)	0.003	0.032 (0.010)	0.002
All, without BMI adjustment	10,681	12,586	0.057 (0.008)	< 0.001	0.036 (0.010)	0.001
All, without height adjustment	10,681	12,586	0.022 (0.008)	0.004	0.033 (0.010)	0.001
Subset by BMI group (kg/m²)
≤ 20.0	652	752	0.037 (0.032)	0.251		N/A^b
20.1–25.0	3915	4585	0.039 (0.013)	0.002	0.068 (0.014)	< 0.001
25.1–30.0	3427	4048	0.013 (0.014)	0.328	0.005 (0.018)	0.782
> 30.0	2687	3201	0.012 (0.016)	0.456	0.009 (0.023)	0.704
Pre-menopausal women—Model B^d
All	4349	5135	0.020 (0.013)	0.122	0.028 (0.017)	0.111
Subset by parity
Nulliparous women	483	558	− 0.044 (0.036)	0.226	0.009 (0.048)	0.847
Parous women	3866	4577	0.027 (0.013)	0.045	0.026 (0.018)	0.162
Parous women + age 1st birth adjustment	3866	4577	0.027 (0.013)	0.044	0.025 (0.018)	0.165
Post-menopausal women—Model C^d
All	5868	6902	0.022 (0.010)	0.029	0.029 (0.013)	0.029
All + age at menopause adjustment	5868	6902	0.022 (0.010)	0.031	0.029 (0.013)	0.031
Subset by parity
Nulliparous women	535	626	0.005 (0.035)	0.891	0.022 (0.044)	0.623
Parous women	5333	6276	0.023 (0.011)	0.029	0.029 (0.014)	0.040
Parous women + age at 1st birth adjustment	5333	6276	0.021 (0.011)	0.051	0.026 (0.014)	0.064
Parous women + age at menopause adjustment	5333	6276	0.023 (0.011)	0.030	0.029 (0.014)	0.041
Parous women + ages at 1st birth + age at menopause adjustments	5333	6276	0.021 (0.011)	0.049	0.026 (0.014)	0.063
Nulliparous women—Model D^e
All	1068	1238	− 0.016 (0.025)	0.517	0.013 (0.033)	0.696
Parous women—Model E^c
All	9561	11,283	0.027 (0.008)	0.001	0.033 (0.011)	0.002
All + age at 1st birth adjustment	9561	11,283	0.026 (0.008)	0.001	0.033 (0.011)	0.003

BMI body mass index, ICMD International Consortium of Mammographic Density, SE standard error
^aA multi-level linear regression model was used to estimate the difference in square-root per cent density and dense area associated with each per year increase in age at menarche. Models specified correlations of women (level 1) within population groups (level 2)
^bBMI ≤ 20 category combined with BMI 20.1–25.0 categories due to lack of convergence
^cModels A (all women) and E (parous women): adjusted for age at mammogram³, menopausal status, age at mammogram*menopausal status interaction term, BMI³, use of HRT, mammogram view, calibration, reader, parity, height
^dModels B (pre-menopausal women) and C (post-menopausal women): adjusted for age at mammogram³, BMI³, use of HRT, mammogram view, calibration, reader, parity, height
^eModel D (nulliparous women): adjusted for age at mammogram³, menopausal status, age at mammogram* menopausal status interaction term, BMI³, use of HRT, mammogram view, calibration, reader, height
^fDense area models: BMI² adjusted for instead of BMI³

Table 3. Association of per cent density and dense area with adult height in pooled ICMD studies
	No. of women	No. of measurements	Per cent density^a		Dense area^a,e
	No. of women	No. of measurements	Changes in square-root per cent density (SE)	P value	Changes in square-root dense area (SE)	P value
Height (categorical) (cm)
< 155	2718	3240	0		0
155–159.9	2551	3027	− 0.041 (0.037)	0.259	0.048 (0.048)	0.319
160–164.9	2683	3151	− 0.056 (0.038)	0.141	0.096 (0.050)	0.057
≥ 165	2480	2900	− 0.075 (0.042)	0.076	0.118 (0.055)	0.034
Linear associations with a 10 cm increase in adult height
All women—Model A^b
All, without BMI adjustment	10,432	12,318	0.044 (0.023)	0.054	0.069 (0.028)	0.012
All, without age at menarche adjustment	10,432	12,318	− 0.057 (0.021)	0.007	0.062 (0.028)	0.025
All, with BMI and menarche adjustment	10,432	12,318	− 0.060 (0.021)	0.005	0.059 (0.028)	0.034
By BMI strata: (kg/m²)
≤ 20.0	652	752	0.042 (0.075)	0.569	0.184 (0.073)	0.011
20.1–25.0	3882	4551	− 0.008 (0.034)	0.811	0.112 (0.041)	0.006
25.1–30.0	3328	3937	− 0.043 (0.037)	0.241	0.121 (0.050)	0.015
> 30.0	2570	3078	− 0.138 (0.040)	0.001	− 0.026 (0.060)	0.670
Subset by height measurement method
Measured	7062	8389	− 0.040 (0.026)	0.125	0.084 (0.034)	0.015
Self-reported	3370	3929	− 0.096 (0.036)	0.007	0.020 (0.047)	0.676
Pre-menopausal women—Model B^c
All	4221	5000	− 0.049 (0.032)	0.130	0.074 (0.044)	0.092
Subset by parity
Nulliparous women	478	553	− 0.201 (0.089)	0.024	− 0.012 (0.117)	0.922
Parous women	3743	4447	− 0.023 (0.034)	0.503	0.099 (0.047)	0.034
Parous women + age 1st birth adjustment	3743	4447	− 0.024 (0.034)	0.490	0.099 (0.047)	0.033
Post-menopausal women—Model C^c
All	5747	6769	− 0.067 (0.028)	0.017	0.047 (0.036)	0.201
All + age at menopause adjustment	5747	6769	− 0.070 (0.028)	0.013	0.043 (0.036)	0.239
Subset by parity
Nulliparous women	531	622	− 0.144 (0.081)	0.076	− 0.084 (0.098)	0.389
Parous women	5216	6147	− 0.041 (0.029)	0.163	0.090 (0.039)	0.019
Parous women + age at 1st birth adjustment	5216	6147	− 0.045 (0.029)	0.121	0.085 (0.038)	0.027
Parous women + age at menopause adjustment	5216	6147	− 0.043 (0.029)	0.139	0.087 (0.038)	0.023
Parous women + age at 1st birth + age at menopause adjustments	5216	6147	− 0.047 (0.029)	0.107	0.083 (0.038)	0.031
Nulliparous women—Model D^d
All	1059	1229	− 0.197 (0.060)	0.001	− 0.043 (0.075)	0.563
Parous women—Model E^b
All	9321	11,024	− 0.036 (0.022)	0.111	0.090 (0.030)	0.002
All + age at 1st birth adjustment	9321	11,024	− 0.038 (0.022)	0.090	0.088 (0.030)	0.003

BMI body mass index, ICMD International Consortium of Mammographic Density, SE standard error
^aA multi-level linear regression model was used to estimate the difference in square-root per cent density and dense area associated with each 10 cm increase in height. Models specified correlations of women (level 1) within population groups (level 2)
^bModels A (all women) and E (parous women): adjusted for age at mammogram³, menopausal status, age at mammogram* menopausal status interaction term, BMI³, use of HRT, mammogram view, calibration, reader, parity, age at menarche
^cModels B (pre-menopausal women) and C (post-menopausal women): adjusted for age at mammogram³, BMI³, use of HRT, mammogram view, calibration, reader, parity, age at menarche
^dModel D (nulliparous women): adjusted for age at mammogram³, menopausal status, age at mammogram* menopausal status interaction term, BMI³, use of HRT, mammogram view, calibration, reader, age at menarche
^eDense area models: BMI² adjusted for instead of BMI³

Authors and Disclosures

Sarah V. Ward¹, Anya Burton^2,3, Rulla M. Tamimi⁴, Ana Pereira⁵, Maria Luisa Garmendia⁵, Marina Pollan^6,7, Norman Boyd⁸, Isabel dos-Santos-Silva⁹, Gertraud Maskarinec¹⁰, Beatriz Perez-Gomez^6,7, Celine Vachon¹¹, Hui Miao¹², Martín Lajous¹³, Ruy López-Ridaura¹³, Kimberly Bertrand¹⁴, Ava Kwong^15–17, Giske Ursin^18–20, Eunjung Lee²¹, Huiyan Ma²², Sarah Vinnicombe²³, Sue Moss²⁴, Steve Allen²⁵, Rose Ndumia²⁶, Sudhir Vinayak²⁶, Soo-Hwang Teo^27,28, Shivaani Mariapun²⁸, Beata Peplonska²⁹, Agnieszka Bukowska-Damska³⁰, Chisato Nagata³¹, John Hopper³², Graham Giles^32,33, Vahit Ozmen³⁴, Mustafa Erkin Aribal³⁵, Joachim Schüz¹, Carla H. Van Gils³⁶, Johanna O. P. Wanders³⁶, Reza Sirous³⁷, Mehri Sirous³⁸, John Hipwell³⁹, Jisun Kim⁴⁰, Jong Won Lee⁴⁰, Caroline Dickens⁴¹, Mikael Hartman^12,42, Kee-Seng Chia⁴³, Christopher Scott¹¹, Anna M. Chiarelli⁴⁴, Linda Linton⁸, Anath Arzee Flugelman⁴⁵, Dorria Salem⁴⁶, Rasha Kamal⁴⁶, Valerie McCormack²* and Jennifer Stone¹

¹School of Population and Global Health, The University of Western Australia, Perth, Australia. ²Environment and Lifestyle Epidemiology Branch, International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France. ³Translation Health Sciences, University of Bristol, Bristol, UK. ⁴Population Health Sciences, Weill Cornell Medical College, Cornell University, New York, USA. ⁵Institute of Nutrition and Food Technology, University of Chile, Santiago, Chile. ⁶Cancer and Environmental Epidemiology Unit, Instituto de Salud Carlos III, Madrid, Spain. ⁷Consortium for Biomedical Research in Epidemiology and Public Health (CIBERESP), Madrid, Spain. ⁸Princess Margaret Cancer Centre, Toronto, ON, Canada. ⁹Department of Non-Communicable Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK. ¹⁰University of Hawaii Cancer Center, Honolulu, HI, USA. ¹¹Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA. ¹²Saw Swee Hock School of Public Health, National University of Singapore, Singapore City, Singapore. ¹³Instituto Nacional de Salud Pública, Cuernavaca, Mexico. ¹⁴Slone Epidemiology Center at, Boston University, Boston, MA, USA. ¹⁵Division of Breast Surgery, Faculty of Medicine, University of Hong Kong, Pok Fu Lam, Hong Kong, China. ¹⁶Department of Surgery and Cancer Genetics Center, Hong Kong Sanatorium and Hospital, Pok Fu Lam, Hong Kong, China. ¹⁷Hong Kong Hereditary Breast Cancer Family Registry, Pok Fu Lam, Hong Kong, China. ¹⁸Cancer Registry of Norway, Oslo, Norway. ¹⁹Department of Nutrition, Institute of Basic Medical Sciences, University of Oslo, Oslo, Norway. ²⁰Department of Preventive Medicine, University of Southern California, Los Angeles, CA, USA. ²¹Department of Population and Public Health Sciences, Keck School of Medicine, University of Southern California, Los Angeles, USA. ²²Department of Population Sciences, City of Hope National Medical Center, Duarte, CA, USA. ²³Division of Cancer Research, Ninewells Hospital and Medical School, University of Dundee, Dundee, Scotland, UK. ²⁴Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, UK. ²⁵Department of Imaging, Royal Marsden NHS Foundation Trust, London, UK. ²⁶Aga Khan University Hospital, Nairobi, Kenya. ²⁷Breast Cancer Research Group, University Malaya Medical Centre, University Malaya, Kuala Lumpur, Malaysia. ²⁸Cancer Research Malaysia, Subang Jaya, Malaysia. ²⁹Department of Environmental Epidemiology, Nofer Institute of Occupational Medicine, Łodź, Poland. ³⁰Department of Physiology, Pathophysiology and Clinical Immunology,, Medical University of Lodz., Łodź, Poland. ³¹Department of Epidemiology and Preventive Medicine, Graduate School of Medicine, Gifu University, Gifu, Japan. ³²Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, University of Melbourne, Melbourne, VIC, Australia. ³³Cancer Epidemiology Centre, Cancer Council Victoria, Melbourne, VIC, Australia. ³⁴Faculty of Medicine, Istanbul University, Istanbul, Turkey. ³⁵Department of Radiology, School of Medicine, Marmara University, Istanbul, Turkey. ³⁶Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands. ³⁷Radiology Department, George Washington University Hospital, Washington, DC, USA. ³⁸Radiology Department, Isfahan University of Medical Sciences, Isfahan, Iran. ³⁹Centre for Medical Image Computing, University College London, London, UK. ⁴⁰Asan Medical Center, Seoul, Republic of Korea. ⁴¹Department of Internal Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa. ⁴²Department of Surgery, Yong Loo Lin School of Medicine, National University of Singapore, Singapore City, Singapore. ⁴³NUS Graduate School for Integrative Sciences and Engineering, National University of Singapore, Singapore, Singapore. ⁴⁴Ontario Breast Screening Program, Cancer Care Ontario, Toronto, ON, Canada. ⁴⁵National Cancer Control Center, Lady Davis Carmel Medical Center, Faculty of Medicine, Technion-Israel Institute Technology, Haifa, Israel. ⁴⁶Woman Imaging Unit, Radiodiagnosis Department, Kasr El Aini, Cairo University Hospitals, Cairo, Egypt.

*Correspondence
McCormackV@iarc.fr

Author Contributions
The study was conceptualized and designed by VM, JSt and ABur and supervised by VM and JSt. Data were curated by GM, BPG, CV, HMi, ML, RLR, AP, MLG, RMT, KB, AK, GU, EL, HMa, SVinn, SMo, SA, RN, SVina, SHT, SMa, BP, AB-D, CN, JSt, JHo, GG, VO, MEA, JSc, CVG, JOPW, RS, MS, JHi, JK, JWL, CD, MH, KSC, CS, AMC, LL, MP, AAF, DS, RK, NB, IdSS and VM. Centralized breast density readings were performed by VM, IdSS and NB, and data harmonization was conducted by ABur. Primary data analyses were performed by SW, with additional analyses and input from VM and ABur. Results were interpreted by SW, JSt, VM, ABur, GM, AP, MLG, KB, BP, MP, NB and IdSS. The manuscript was drafted by SW, JSt, VM and ABur and revised by SW and JSt. All authors reviewed the final manuscript. All authors read and approved the final manuscript.

Competing interests
ML declares a non-restricted investigator-initiated grant from AstraZeneca and minor support from Swiss Re.