Introduction

Family history of cancer is defined as having one or more first-degree relatives (FDR) (e.g., parent, sibling or child) diagnosed with cancer1,2. Individuals with a family history of breast, ovarian, uterine, and colorectal cancer, in particular, are known to be at increased risk of getting such cancers1,2. This risk increases with younger age at diagnosis of the relative1,3. Multiple FDR with a history of cancer affects an individual’s cancer risk1,3. While the complex interaction of genetic and environmental exposures contributing to the association between family history of cancer and cancer risk have been extensively studied4,5,6, exact pathways remain inconclusive7.

The US Preventive Task Force updated guidelines in 2020 for site-specific cancers such as colorectal, lung, breast, pancreas, cervix and ovary cancer8. Given that early detection is possible through effective screening methods, the U.S. guidelines highlight available screening programs for individuals with a family history1,8. In addition, clinical services also focused on prevention such as BRCA gene mutation testing or breast chemoprevention9. In the UK risk-reducing mastectomies are offered to women who have been identified at high risk10.

In Japan, the National Cancer Center (NCC) Japan publishes and updates cancer screening guidelines11,12,13,14,15. However, the recommendations do not specify physician assessment of family history. Family history is not always assessed and assessment methods are inconsistent though the rationale for obtaining this history is well established4. Several epidemiological studies exist on family history of cancer and various cancer outcomes in Japan4. However, a broad overview and update on the current prevalence of family history of total and site-specific cancer stratified by sex for age and birth cohorts are lacking. For example, no Japanese studies were included in the 2019 paper on trends by the Hereditary Breast Cancer Study Group10.

The objective of this study was to identify the prevalence of family history of cancer using cohorts participating in the Japanese National Center Cohort Collaborative for Advancing Population Health (NC-CCAPH).

Materials and methods

Study population

The Japan Health Research Promotion Bureau (JH Bureau) was launched in April 2020, to create innovation in research and development (R&D) and medical care by coordinating six national centers for advanced and specialized medicine in Japan (6NC)16. The six national centers are: NCC, the National Cerebral and Cardiovascular Center (NCVC), the National Center of Neurology and Psychiatry, the National Center for Global Health and Medicine (NCGM), the National Center for Child Health and Development, and the National Center for Geriatrics and Gerontology (NCGG). Additional details can be found on the JH bureau homepage16.

Study-specific data was pooled from seven of nine eligible independent cohorts (Supplementary file) participating in the National Center Cohort Collaborative for Advancing Population Health (NC-CCAPH) or 6NC for short. Four cohorts of the National Cancer Center contributed data: Japan Public Health Center Study I (JPHC I)17, Japan Public Health Center Study II (JPHC II)17, Japan Public Health Center NEXT Study (JPHC-NEXT)18, and the National Cancer Center Japan-Screening Cohort (JaSCo) Study19. The Japan Epidemiology Collaboration on Occupational Health Study (J-ECOH)20 of the NCGM joined and the National Center for Geriatrics and Gerontology-Study of Geriatric Syndromes (NCGG-SGS)21 and the Suita Study (Suita) of NCVC22. Participants in the JPHC II Study Osaka Suita Public Health Center area were excluded (n = 10,960) to avoid potential overlap with participants in the Suita Study. None of the other study areas overlapped, thus eliminating the possibility of duplication. The remaining two eligible cohorts participating in the consortium could not be included as they lacked data on family history of cancer. One of the cohorts would have added a younger birth cohort. This study was approved by the Institutional Review Board of the National Cancer Center Japan (approval number 2018–194). All methods were carried out in accordance with relevant guidelines and regulations. Informed consent was obtained from all subjects.

Analysis

Family history was assessed at baseline using questionnaires. While other questions such as the food frequency questionnaires (FFQ) have been validated, the family history questions have not been verified. A respondent was considered to have a family history of cancer if at least one family member (parent or sibling) had cancer7. We included respondents with complete information for age, birth year and who responded to relevant family history exposure questions. Individuals were counted once for total cancer even if they had multiple site-specific cancers. Prevalence estimates and respective 95% confidence intervals (CIs) were calculated by pooling the data and presented for total family history of cancer, parent and sibling separately, by sex and site-specific cancers for all cohorts except J-ECOH, for which only total cancer was available. Estimates were calculated using “ci proportion” in Stata to obtain the percent. Prevalence was presented in 10 year age categories from 40 to > 70 and birth cohorts < 1929 to 1979. All analyses were performed using Stata Statistical Software: Release 17, College Station, Texas.

Results

Over 343,000 participants with available data on family history of cancer and age from seven Japanese cohort studies were included in the study. Survey years ranged from 1989 to 2020 and age at baseline ranged from 15 to 104 (Table 1). Birth year ranged from 1906 to 2004 (Table 1). The total prevalence of family history of cancer varied between cohorts from 18.82% (J-ECOH) to 63.05% (JaSCo) (Table 1). J-ECOH represents a relatively young workers cohort whereas JPHC-NEXT, JaSCo, NCGG-SGS and Suita include older participants (Table 1).

Table 1 Overview of participating cohorts of the Japanese Six National Centers (6NC) consortium.
Full size table

Prevalence of a family history of cancer increased with age ranging from 10.51% (95% CI 10.22to 10.88) in the 15 to 39 year age category to 47.11% (95% CI 46.59 to 47.63) in 70-year-olds (Table 2). Overall prevalence increased in birth cohorts from ≤ 1929 until 1960 and decreased for the next two decades (Table 3). Gastric cancer (11.97%; 95% CI 11.84 to 12.10) was the most common site recorded for family members, followed by colorectal and lung (5.75%; 95% CI 5.66 to 5.84), prostate (4.37%; 95% CI 4.26 to 4.47), breast (3.43%; 95% CI 3.36 to 3.50) and liver (3.05%; 95% CI 2.99 to 3.12) cancer. Women consistently had a higher prevalence of family history of cancer (34.32%; 95% CI 34.08 to 34.56) versus men (28.75%; 95% CI 28.55 to 28.96) (Tables 2 and 3).

Table 2 Prevelance of family history of cancer and corresponding 95% confidence intervals by age in seven Japanese cohort studies.
Full size table
Table 3 Prevalance of family history of cancer and corresponding 95% confidence intervals by birth cohort in seven Japanese cohort studies.
Full size table

Discussion

One in three participants had a family history of cancer in this Japanese study, which was similar to the study finding reported in the United States in 20067. A consistently higher prevalence of family history of cancer among women versus men was also in line with the study conducted in the United States7. Gastric cancer was the most common family history of cancer site in this Japanese study, followed by lung, colorectal, prostate, breast and liver cancer. In comparison, the US-based study by Ramsey et al. found breast, followed by lung, colorectal, and prostate cancer, to be the most common cancer sites of those assessed. Even though breast cancer rates are lower in Japan, the incidence is increasing23. The US population is ethnically heterogeneous compared to a homogenous population in Japan. The proportion of some cancers which have strong genetic factors could differ by ethnicity24. In contrast to our study, a Swedish Cancer Registry study found a higher 20.15% proportion of family history of prostate cancer versus 4.37% in this Japanese study25.

A meta-analysis of case–control studies in Eastern and Central Europe found family history of lung cancer to be a risk factor for lung cancer, particularly in younger subjects under age 505. The prevalence of family history of lung cancer among Japanese increased by age group from 3.21% in participants in their 40 s to 9.89% in the ≥ 70 year olds. The prevalence was similar among men and women at 5.75% for total family history of cancer all age groups. In a large-scale population-based study in Utah, USA, 30% of individuals had a family history in at least one FDR, second-degree relative, or third-degree relative, however this percent decreased dramatically when only considering FDRs26.

In Japan, the prevalence of infection-related cancers such as a family history of gastric and liver cancer, 11.87% and 3.03%, was higher than those reported Europe and North America. In fact, the prevalence of gastric cancer family history was surprisingly high even though the incidence has been on a decreasing trend. Infection risk may be linked with birth cohorts. Specifically for Helicobacter pylori (H. pylori) the evidence is mixed whether the infection is the cause of gastric cancer27. General improvements in hygiene have contributed to the decline of the H.pylori infection in the Japanese population27 as well as scaling up targeted gastric cancer screening programs may contribute. At the political level in Japan, prevention of infection and detection of carriers, and antiviral treatment have contributed to a substantial reduction of liver cancer incidence.

The reasons why the proportion of people with a family history of cancer is increasing are complex. These trends are observed in other countries. Common reasons may include expanded longevity leading to more cancer cases and improved cancer detection rates due to the spread of screening and development of testing techniques28. These factors combined may complicate the interpretation of family history for each birth cohort. Family history in our study increased by birth cohort until 1970, participants born after this were younger at baseline with younger parents who may have not yet been diagnosed. On the other hand, it can be speculated that parents of participants in the earlier cohorts did not notice they had cancer or died before the occurrence of cancer because of shorter life expectancy. Additionally, an interaction between life expectancy, lifestyle such as diet and genetics must be considered29,30. Genetics within the Japanese population remained stable during the study period. The high percentage of affected persons highlights the importance of early and targeted cancer screening services.

The government develops guidelines for organized cancer screening and societies/associations create screening guidelines for opportunistic screening. However, some of the descriptions may not obtain consensus from all stakeholders. For example, professional medical societies such as the Japan Association of Breast Cancer Screening31, Japanese Association for Cancer Detection and Diagnosis, Japanese Breast Cancer Society (JBCS)32, The Japan Lung Cancer Society33, Japan Cancer Society34, Japanese Society for Cancer of the Colon and Rectum, and The Japan Urological Association35 publish their own site-specific statements. Among them, a family history of cancer is included only for opportunistic screening for prostate cancer.

The main strength of this study is that the data comes from several large-scale population-based cohorts in Japan, making the study representative. Age at baseline spanned almost 100 years (15–104) and birth cohorts cover nearly a century (1915–2004). This study may help predict the burden of screening on physicians, clinics, insurers, and systems. The self-reported information from the cohorts should be sufficiently reliable. While the family history questions were not validated, other parts of the questionnaires such as FFQ indicated relatively high quality. In the JaSCo study for example, consistency of responses to all items was inspected by research assistants in the presence of study subjects. Despite these strengths, several limitations warrant mentioning. First, two cohorts did not collect cancer site-specific data. Parent versus sibling family history was not differentiated in all cohorts. This information was presented separately for the whole family history of cancer. Second, this analysis did not consider additional details such as age of relative at diagnosis and socioeconomic factors. These disparities are smaller in Japan than in Europe and the US and thus likely contributed less to an individual’s risk36. Third, false reporting or misdiagnosis, especially in the earlier birth cohorts, cannot be ruled out. Cancer death may be equal to cancer incidence, leading to competing risk of dying from stroke instead of stomach cancer. Also, participants, especially in the early birth cohorts, may know of cancer death of a family member but not necessarily specifically which cancer site, which may result in a further under-estimation in the site proportion of family history of site-specific cancer. Memory may be ambiguous and people do not know accurate information about their parents’ and sibling. In the Twenty-First Century individuals may access and use digitalized personal health information and add their family history at any time. This information can be useful as a public health tool. Forth, the data of family history depended on self-report.

Albeit the limitations as mentioned above, this Japanese National Center Cohort Collaborative for Advancing Population Health study, including nearly 343,000 participants, provides important information on family history of cancer prevalence. Definition of “high risk” family history in cancer screening and prevention guidelines should be reviewed. Future research should adopt risk prediction models for Japan’s family history and site-specific cancers. Additionally, including multi-generational biobank data would help improve the accuracy and validity of family history of cancer and applying findings to clinical and prevention strategies.