Abstract
This cross-sectional study describes the national distribution of genetic testing for hereditary cancer risk, identifies disparities, and assesses whether a gender gap exists
Up to 10% of cancers are attributable to inherited gene mutations, such as BRCA1/2 and those associated with Lynch syndrome.1 Identifying mutation carriers is critical for treatment decisions, cancer prevention, and early detection.2 Despite the importance of equal access to testing, there is some evidence for disparities by education level, insurance status, and race/ethnicity.3 Although carrier rates are equivalent between men and women, there may also be a gender disparity in testing.4 To date, however, epidemiological studies have largely focused on genetic testing in women and hereditary breast and ovarian cancer (HBOC).5 This study describes the US national distribution of genetic testing for hereditary cancer risk, identifies disparities, and assesses whether a gender gap in testing exists.
Methods
The data source is the 2015 US National Health Interview Survey (NHIS), a cross-sectional in-person interview gathering self-reported health data for the US population. The study did not meet regulatory definition of a “human subject” study because the data are publically available. Therefore, neither certification of exemption from University of California Los Angeles institutional review board approval was not required.
The outcome of interest was receipt of genetic testing for cancer risk. Affirmative responders were asked 4 subsequent questions that were not mutually exclusive—if their test was for breast cancer, ovarian cancer, colon or rectal cancer, or other cancer. Answers did not have to be concordant with personal and family history.
Sociodemographic factors of individuals obtaining genetic testing were compared with the national sample, including gender, US Census region, race/ethnicity, insurance, citizenship, and education.
To assess the gender disparity in testing, we stratified by type of testing and cancer status. To evaluate potential underrecognition of family medical history leading to lower testing rates, we compared select family medical histories between unaffected men and women.
All estimates were adjusted for complex survey weights and used Taylor series standard errors. Bivariate comparisons were made using adjusted Wald tests and rate ratios (RRs) were generated via Poisson regression.
Results
In the 2015 NHIS, 378 adults (≥18 years) reported a history of genetic testing for cancer, representing 2 498 842 people. Sociodemographic differences in the testing subsample included a lower proportion of Hispanics (10% vs 16%; P = .002), the uninsured (2% vs 10%; P < .001), noncitizens (4% vs 8%; P < .001), and those with less education (high school or General Educational Development (GED) diploma; 30% vs 44%; P < .001). Almost 3 times as many women received testing as men (73% vs 27%; P < .001). This disparity persisted for unaffected men, testing at half the rate of unaffected women (RR, 0.51; 95% CI, 0.36-0.73) (Table 1).
Table 1. Geographic, Sociodemographic, and Cancer History for Individuals Who Have Undergone Genetic Testing for Cancer, Including Comparisons With the National Sample.
| Characteristic | % (SE)a,b | |
|---|---|---|
| Total Sample | Tested Subsample | |
| Unweighted, No. | 33 672 | 378 |
| Weighted, No. | 242 500 657 | 2 498 842 |
| Gender | ||
| Men | 48 (0.4) | 27 (3.4) |
| Women | 52 (0.4) | 73 (3.4) |
| Regionc | ||
| Northeast | 17 (0.4) | 20 (2.6) |
| Midwest | 22 (0.4) | 19 (2.9) |
| South | 37 (0.5) | 36 (3.2) |
| West | 23 (0.4) | 26 (2.9) |
| Race/ethnicity (self-reported) | ||
| Hispanic | 16 (0.3) | 10 (1.8) |
| Non-Hispanic white | 66 (0.4) | 71 (3.1) |
| Non-Hispanic black | 12 (0.3) | 12 (2.2) |
| Non-Hispanic Asian | 6 (0.2) | 4 (1.24)d |
| Other | 1 (0.1) | 3 (1.4)e |
| Insurancef | ||
| Private | 66 (0.4) | 71 (3.1) |
| Medicare | 21 (0.2) | 27 (2.7) |
| Medicaid/SCHIP | 12 (0.3) | 15 (2.7) |
| Other | 5 (0.2) | 7 (1.6) |
| Uninsured | 10 (0.3) | 2 (0.8)d |
| Citizenshipg | ||
| Citizen | 92 (0.3) | 96 (1.1) |
| Noncitizen | 8 (0.3) | 4 (1.1) |
| Highest educational achievement | ||
| Less than HS | 13 (0.3) | 13 (2.1) |
| HS, GED diploma, or some college | 44 (0.4) | 30 (3.1) |
| Associate’s degree | 11 (0.3) | 14 (2.3) |
| Bachelor’s degree | 20 (0.3) | 25 (3.0) |
| Graduate degree | 12 (0.3) | 17 (2.7) |
| History of any cancer | ||
| Yes (“affected”) | 9 (0.2) | 40 (3.2) |
| Proportion of affected, Menh | 44 (1.2) | 19 (5.0) |
| Proportion of affected, Womeni | 56 (1.2) | 81 (5.0) |
| No (“unaffected”) | 91 (0.2) | 60 (3.2) |
| Proportion of unaffected, Men | 49 (0.4) | 33 (4.0) |
| Proportion of unaffected, Women | 51 (0.4) | 67 (4.0) |
Abbreviations: GED, General Education Development diploma; HS, high school; IHS, Indian Health Service; SCHIP, State Children’s Health Insurance Program; SE, standard error.
Estimates adjusted for complex survey design and including weights. Data Source: NCHS, National Health Interview Survey.6
Proportions estimated for nonmissing.
Regions consistent with US Census definition (https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf).
Estimates are considered unreliable when the relative SE is greater than 30%.
The SE is greater than 50% and would not be included in summary statistics released by NCHS.
Insurance verified against an individual’s membership card, when available. Insurance proportions sum to greater than 100% owing to overlapping coverage (eg, dual Medicare/Medicaid eligible); Medicaid category includes SCHIP and other state-sponsored insurance programs; “Other” includes military coverage (eg, Tricare), IHS, and other governmental coverage.
Citizens are defined as individuals born in the United States, US-held territories, those born to US parent(s), and naturalized citizens.
Top 3 reported cancers among tested men were prostate (n = 5), colon (n = 4), and nonmelanoma skin (n = 4).
Top 3 reported cancers among tested women were breast (n = 80), ovarian (n = 14), and uterine (n = 9).
Three-quarters of genetic testing was for breast/ovarian cancer, 24% for colorectal cancer, and 22% for other cancers. Among the unaffected, men underwent testing for breast/ovarian cancer at one-tenth the rate of women (RR, 0.10; 95% CI, 0.05-0.23). There was no gender disparity for colorectal or other cancer testing (Table 2).
Table 2. Type of Cancer Genetic Testinga Including Stratification by Gender and Cancer Status.
| Characteristic | Breast/Ovarian | Colorectal | Other | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Unweighted, No. | Weighted | Unweighted, No. | Weighted | Unweighted, No. | Weighted | |||||
| No. | % (SE)b | No. | % (SE)b | No. | % (SE)b | |||||
| Total | 230 | 1 467 866 | 76 (3.7) | 101 | 704 115 | 24 (3.5) | 99 | 659 408 | 22 (3.9) | |
| Descriptive Data | ||||||||||
| Stratification by cancer status | ||||||||||
| Affected | 108 | 682 887 | 47 (4.3) | 30 | 226 756 | 32 (7.1) | 35 | 249 858 | 38 (7.1) | |
| Unaffected | 122 | 784 979 | 53 (4.3) | 71 | 477 359 | 68 (7.1) | 64 | 409 550 | 62 (7.1) | |
| Stratification by gender | ||||||||||
| Men | 12 | 123 629 | 8 (3.2)c | 44 | 352 544 | 50 (7.2) | 47 | 315 446 | 48 (7.2) | |
| Women | 218 | 1 344 237 | 92 (3.2) | 57 | 351 571 | 50 (7.2) | 52 | 343 962 | 52 (7.2) | |
| Stratification by cancer status and gender | ||||||||||
| Affected men | 3 | Too small to estimate | 11 | 114 503 | 16 (6.1)c | 12 | 109 307 | 17 (6.4)c | ||
| Unaffected men | 9 | 70 361 | 5 (1.8)c | 33 | 238 041 | 34 (6.7) | 35 | 206 139 | 31 (6.0) | |
| Affected women | 105 | 629 619 | 43 (4.4) | 19 | 112 253 | 16 (4.7) | 23 | 140 551 | 21 (5.7) | |
| Unaffected women | 113 | 714 618 | 49 (4.2) | 38 | 239 318 | 34 (6.6) | 29 | 203 411 | 31 (6.8) | |
| Rate Ratios (95% CI)d | ||||||||||
| Men vs Women | 0.10 (0.04-0.22) | 1.06 (0.60-1.87) | 0.97 (0.55-1.71) | |||||||
| Unaffected men vs unaffected women | 0.10 (0.05-0.23) | 1.04 (0.54-2.00) | 1.06 (0.54-2.08) | |||||||
Abbreviation: SE, standard error.
Self-reported type of genetic testing; answers are not mutually exclusive. Data Source: NCHS, National Health Interview Survey.6
Estimates adjusted for complex survey design and including weights.
Estimates are considered unreliable when the relative SE is greater than 30%.
Linearized 95% CIs were generated via Poisson regression.
There were no differences in reported family history between unaffected men and women (number of first-degree female relatives with breast cancer, breast cancer at ≤50 years, or ovarian cancer).
Discussion
Cancer genetic testing seems to reach a broad geographic and sociodemographic population in this national survey. However, there remain underrepresented groups, including Hispanics, the uninsured, noncitizens, and those with less education. Most strikingly, unaffected men underwent genetic testing at half the rate of unaffected women, owing to a 10 to 1 disparity in HBOC testing.
Previous theories4 for underutilization of HBOC testing in men include (1) lack of patient and clinician awareness on the importance of HBOC mutation status—despite the risks of male breast, pancreatic, melanoma, and aggressive prostate cancers,7 and (2) social roles of men vs women in health. The latter argument seems less likely given the lack of gender disparity in colorectal/other cancer testing. Limitations of this study include self-reported data, recall bias, and limited details regarding reason for testing. Large national efforts, including educational campaigns targeting male HBOC testing, must address this disparity to enable uniform opportunities for cancer prevention, early detection, and treatment for all at-risk individuals and their family members.
References
- 1.Garber JE, Offit K. Hereditary cancer predisposition syndromes. J Clin Oncol. 2005;23(2):276-292. [DOI] [PubMed] [Google Scholar]
- 2.Ludwig KK, Neuner J, Butler A, Geurts JL, Kong AL. Risk reduction and survival benefit of prophylactic surgery in BRCA mutation carriers, a systematic review. Am J Surg. 2016;212(4):660-669. [DOI] [PubMed] [Google Scholar]
- 3.Suther S, Kiros GE. Barriers to the use of genetic testing: a study of racial and ethnic disparities. Genet Med. 2009;11(9):655-662. [DOI] [PubMed] [Google Scholar]
- 4.Daly MB. The impact of social roles on the experience of men in BRCA1/2 families: implications for counseling. J Genet Couns. 2009;18(1):42-48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Childers CP, Childers KK, Maggard-Gibbons M, Macinko J. National estimates of genetic testing in women with a history of breast or ovarian cancer. J Clin Oncol. 2017;35(34):3800-3806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.National Center for Health Statistics. National Health Interview Survey, 2015. Public-use data file and documentation. http://www.cdc.gov/nchs/nhis/quest_data_related_1997_forward.htm. 2016. Accessed February 21, 2018.
- 7.Mersch J, Jackson MA, Park M, et al. . Cancers associated with BRCA1 and BRCA2 mutations other than breast and ovarian. Cancer. 2015;121(2):269-275. [DOI] [PMC free article] [PubMed] [Google Scholar]