Abstract
Subclinical liver impairment due to fibrosis could influence the development and detectability of prostate cancer. To investigate the association between liver fibrosis and prostate cancer incidence and mortality, we included 5,284 men (mean age: 57.6 years, 20.1% Black) without cancer or liver disease at Visit 2 in Atherosclerosis Risk in Communities study. Liver fibrosis was assessed using the aspartate aminotransferase to platelet ratio index (APRI), fibrosis 4 index (FIB-4), and non-alcoholic fatty liver disease fibrosis score (NFS). Over 25 years, 215 Black and 511 White men were diagnosed prostate cancer, and 26 Black and 51 White men died from the disease. We estimated hazard ratios (HRs) for total and fatal prostate cancer using Cox regression. FIB-4 (quintile 5 versus 1: HR=0.47, 95%CI:0.29–0.77, p-trend=0.004) and NFS (HR=0.56, 95%CI:0.33–0.97, p-trend=0.03) were inversely associated with prostate cancer risk in Black men. Compared with no abnormal score, men with ≥1 abnormal score had a lower prostate cancer risk if they were Black (HR=0.46, 95%CI:0.24–0.89), but not White (HR=1.04, 95% CI:0.69–1.58). Liver fibrosis scores did not appear to be associated with fatal prostate cancer in Black or White men. Among men without a clinical diagnosis of liver disease, higher liver fibrosis scores were associated with lower incidence of prostate cancer in Black men, but not in White men, and not with fatal prostate cancer in either race. Further research is needed to understand the influence of subclinical liver disease on prostate cancer development versus detectability and the racial differences observed.
Keywords: liver fibrosis, prostate cancer, risk, cohort study
Introduction
Lower serum prostate-specific antigen (PSA) concentration in men with liver disease, as shown in our previous research and others, may adversely impact the accuracy for early detection of prostate cancer. (1–4) PSA screening has been the primary tool for the early detection of prostate cancer in the US for 30 years(5), and the burden of liver disease has been increasing over the same period(6). Further, the higher burden of liver diseases in certain subgroups, e.g., the higher prevalence of viral hepatitis in Black men,(7) might help partially explain, the racial disparity observed in prostate cancer mortality. Given that early detection may be beneficial for prostate cancer treatment and survival, a delayed diagnosis of a potentially aggressive prostate cancer could contribute to the higher prostate cancer mortality in US Black men.(8) We are not suggesting that men with a substantially reduced life expectancy due to chronic liver disease should be screened for prostate cancer. However, men with early liver fibrosis, well before clinically apparent cirrhosis, may not have markedly shortened life expectancy, and thus, may have prostate cancer as a competing risk of death. These latter men, especially in populations with a higher prostate cancer mortality, might benefit from the early detection of prostate cancer.
Besides affecting the accuracy of PSA screening, liver impairment resulting from fibrosis may play a complex role in prostate cancer development and progression (biology), and may additionally affect prostate cancer detection through effects on prostate size. The liver is involved in regulating androgen levels and produces sex hormone binding globulin (which carries androgen in circulation); and prostate cancer(9) and prostate(10) size are both, in part, dependent on androgens, albeit in complex ways.(11,12) Any influence of androgens on prostate volume is important for prostate cancer diagnosis following a positive screen, as biopsy is more sensitive in detecting prostate cancer in men with smaller prostate volumes.(13) Liver conditions also affect the regulation of other pathways that in turn may influence prostate cancer risk. For example, NAFLD is considered to be a hepatic manifestation of insulin resistance(14) and studies have observed that men with insulin resistance (but not frank diabetes) have increased prostate cancer risk.(15,16)
No firm link has been established between liver disease and prostate cancer incidence or mortality. With respect to risk, the prevalence of prostate cancer was higher in men who were hepatitis C antibody positive compared with negative in a retrospective cohort of US men who underwent prostate biopsy.(17) A prospective cohort study using the UK Biobank found that men with higher serum aspartate aminotransferase (AST, a biomarker for liver function) had a lower prostate cancer risk.(18) With respect to outcomes, NAFLD status was inversely associated with biochemical recurrence (BCR) in Korean men surgically treated for prostate cancer.(19) In contrast, AST (also called serum glutamic-oxaloacetic transaminase, SGOT) and alanine aminotransferase (ALT; also called serum glutamate pyruvate transaminase, SGPT; another liver function biomarker) were not associated with BCR among surgically treated patients in Shared Equal Access Regional Cancer Hospital (SEARCH) database.(12) Whether these observed associations may be explained by the influence of liver impairment on the development of prostate cancer, detection bias, another non-causal explanation, or a combination of these effects is unclear.
Thus, we aimed to evaluate the association between liver fibrosis, as measured by three non-invasive scores, and total and fatal prostate cancer risk in Black and White men without a clinical diagnosis of liver disease at the start of follow-up in the Atherosclerosis Risk in Communities (ARIC) study. We hypothesized that if subclinical liver fibrosis only reduces the sensitivity of PSA-based prostate cancer screening, we would observe an inverse association with prostate cancer risk and a positive association with fatal prostate cancer due to delayed detection and treatment.
Materials and Methods
Study population
The ARIC study is a US prospective cohort with 15,792 participants aged 45–66 years old when recruited in 1987–1989 from four communities: Forsyth County, NC; Jackson, MS; Minneapolis, MN; and Washington County, MD. Details are described elsewhere.(20,21) Participants were interviewed and examined at baseline, and then returned for follow-up study visits (2–5: 1990–1992, 1993–1995, 1996–1998, 2011–2013; and with additional visits after the end of follow up for cancer). Local institutional review boards approved the ARIC protocol and written informed consent was obtained from all ARIC participants. The research was conducted under the U.S. Common Rule. The majority (99.7%) of participants gave approval for follow-up for non-cardiovascular diseases.
We used Visit 2, when liver enzymes were first measured, as the start of follow up, and included men without a cancer diagnosis by Visit 2 who had complete data to calculate liver fibrosis scores and key covariates. We excluded men with a clinical diagnosis of liver disease based on response to the question “Has a doctor ever told you that you have cirrhosis or another chronic liver disease” at Visit 3. For men who missed Visit 3, routinely collected hospital discharge summaries for liver reasons (ICD-9: 571 or ICD-10: K74) was used to determine chronic liver disease status if they had a history of hospitalization. Since most ARIC participants were Black (27.0%) or White (72.7%), we also excluded men with a self-reported race other than Black or White, and Black men from the Minneapolis and Washington County field centers, which had very few Black participants.
Exposures
Serum liver enzymes – AST, ALT – platelet count and albumin were measured at Visit 2 as part of the ARIC study protocol. To assess the presence and extent of liver fibrosis, we calculated three non-invasive fibrosis scores for each man as we did previously(4): AST/platelet ratio index (APRI), fibrosis 4 index (FIB-4) and NAFLD fibrosis score (NFS). Specifically, we used Visit 2 liver enzyme concentrations and Visit 1 platelet counts (were not measured for all field centers at Visit 2). We used 33 U/L as upper-limit of normal of AST as we did previously.(22) Visit 2 body mass index (BMI) was calculated as weight (kg) divided by height squared (m2). Men were classified as having impaired fasting glucose or diabetes if they had fasting glucose >100 mg/dL (if not fasting: >140 mg/dL), self-reported a physician’s diagnosis of diabetes, or used a diabetes medication. We defined abnormal fibrosis score as APRI>1, FIB-4>2.67 or NFS>0.676.(23,24)
Outcomes
Outcomes were first primary incident, lethal (metastasis to any organ or death from prostate cancer as the underlying cause), and fatal prostate cancer (death from prostate cancer as the underlying cause) that occurred after Visit 2 through 12/31/2015. Prostate cancer status, including stage at diagnosis, was ascertained through linkage with the MN, NC, MD, and MS state cancer registries. Additional cases were identified by the review of medical records following a cancer-specific telephone call, archived hospital discharge codes, and death certificates.(21) The median follow-up time in the study was 20.1 years, with a range of 0.01 to 25.9 years. Through 2015, 726 (511 White and 215 Black) men were diagnosed with prostate cancer, and of these, 83 (55 White and 28 Black) had lethal prostate cancer, and 77 (51 White and 26 Black) died of prostate cancer in the analytic cohort.
Covariates
Other variables considered were (Visit 2 unless otherwise specified): age, race and field center (Visit 1), diabetes, BMI, waist-hip ratio, cigarette smoking status (never, quit<10 years, current/quit within 10 years), alcohol consumption (never, former, current), family history of prostate cancer (Visit 3), and statin and aspirin use; these are known or purported risk or protective factors for total or lethal prostate cancer.(25)
Socioeconomic status (SES) and its correlates, such as access to and uptake of healthcare, may confound the association between liver fibrosis and prostate cancer. Generally, persons with poorer liver function tend to have lower SES, which is in turn linked with reduced access to care, and may limit opportunities for cancer screening.(26) We generated a propensity score (see Statistical analysis) that included childhood, early adulthood, and later adulthood SES each calculated using Visit 4 data as previously done in ARIC(27); US Census tract data on neighborhood income for the year 1990(28); typical frequency of routine medical examinations at Visits 1–3; health insurance status at Visit 1; type of health insurance at Visit 2; usual type of medical care at Visit 2.
Statistical analysis
Cox proportional hazards regression was used to estimate adjusted hazard ratios (HR) for the association between liver fibrosis scores and (1) prostate cancer incidence, (2) lethal prostate cancer incidence, and (3) prostate cancer mortality. Person-years at risk were calculated from the date of Visit 2 until the date of prostate cancer diagnosis, diagnosis of a different cancer, death from another cause, or administrative censoring, whichever occurred first. For fatal prostate cancer, follow-up was through date of prostate cancer death, death from another cause, or administrative censoring, whichever occurred first. We expressed the liver fibrosis scores as continuous (per standard deviation) or in quintiles. We tested the linear trend for each score by modeling the median of each quintile as a continuous term. We ran three models: Model 1 adjusted for age and the combination of race and field center; Model 2 further adjusted for BMI, waist circumference, diabetes/impaired fasting glucose status, alcohol drinking, cigarette smoking status, family history of prostate cancer, and statin or aspirin use; and Model 3 further adjusted for a propensity score for SES and correlated variables (fully-adjusted model). To generate the propensity score, we modeled the association between liver fibrosis scores (normal vs. abnormal) and SES at each of the 3 points in life and the correlated variables using logistic regression and predicted the propensity score for each participant.
All analyses were conducted in the overall population, and stratified by race given the racial difference in prostate cancer risk. In addition, since NFS calculation includes BMI and diabetes, we ran a Model 4 for NFS in which we adjusted the variables in Model 3 but without these two factors to avoid potential over-adjustment. We also cross-categorized the men with respect to all three cores (APRI, FIB-4, NFS) to improve the accuracy of the classification of the absence of liver fibrosis (0 abnormal scores), and examined the association between the cross-categorized score and prostate cancer incidence; the number of lethal and fatal prostate cancers was too sparse to conduct this analysis.
Statistical tests were two-sided and P<0.05 was considered as statistically significant. All analyses were performed using STATA 14.2.
Data availability
The data analyzed in this study are available from the Atherosclerosis Risk in Communities (ARIC) study. Detailed information about data access and sharing is available online (https://sites.cscc.unc.edu/aric/sites/default/files/public/listings/ARIC%20data%20sharing.pdf).
Results
The analysis included 5,284 men with a mean age of 57.6 years. Table 1 shows age-adjusted characteristics by liver fibrosis status. Abnormal liver fibrosis scores were present in 0.68% (APRI), 3.20% (FIB-4), and 4.83% (NFS) of men. These percentages were higher in Black (APRI, FIB-4, NFS: 1.12%, 5.42%, 8.31%) than White (0.57%, 2.63%, 3.94%) men. Men with abnormal liver fibrosis scores were more likely to be diabetic, obese (BMI ≥30 kg/m2), current smokers, less likely to have health insurance, and have a family history of prostate cancer. Alcohol drinking status was similar between those with and without abnormal liver fibrosis scores in these participants without diagnosed liver disease.
Table 1.
Age-adjusted Characteristics of Participants by Liver Fibrosis Scores at Visit 2, Men in the Atherosclerosis Risk in Communities Study, 1990–1992
| Overall | APRI | FIB-4 | NFS | ||||
|---|---|---|---|---|---|---|---|
| Factor | Normal | Abnormala | Normal | Abnormala | Normal | Abnormala | |
| N | 5284 (100%) |
5248 (99.32%) |
36 (0.68%) |
5115 (96.80%) |
169 (3.20%) |
5029 (95.17%) |
255 (4.83%) |
| Age in year, mean (SD) | 57.6 (5.7) |
57.6 (5.7) |
58.1 (5.8) |
57.5 (5.7) |
60.8 (5.3) |
57.5 (5.7) |
60.4 (5.5) |
| Race, % (95% CI) b | |||||||
| White | 79.9 (78.8–81.0) |
80.0 (78.9–81.1) |
66.5 (50.9–82.0) |
80.5 (79.4–81.6) |
63.0 (55.5–70.5) |
80.8 (79.7–81.9) |
62.7 (56.6–68.8) |
| Black | 20.1 (19.0–21.2) |
20.0 (18.9–21.1) |
33.5 (18.0–49.1) |
19.5 (18.4–20.6) |
37.0 (29.5–44.5) |
19.2 (18.1–20.3) |
37.3 (31.2–43.4) |
| Field center, % (95% CI) | |||||||
| Forsyth, North Carolina | 26.2 (25.0–27.4) |
26.3 (25.1–27.5) |
13.9 (2.6–25.2) |
26.2 (25.0–27.4) |
25.7 (19.1–32.4) |
26.2 (25.0–27.4) |
27.0 (21.5–32.4) |
| Jackson, Mississippi | 17.6 (16.6–18.7) |
17.5 (16.5–18.6) |
30.7 (15.6–45.9) |
17.1 (16.1–18.2) |
32.8 (25.5–40.2) |
16.9 (15.8–17.9) |
32.5 (26.6–38.5) |
| Minneapolis, Minnesota | 29.0 (27.8–30.3) |
29.2 (27.9–30.4) |
11.1 (0.8–21.3) |
29.5 (28.3–30.8) |
14.3 (9.1–19.6) |
29.8 (28.5–31.0) |
14.5 (10.2–18.8) |
| Washington County, Maryland | 26.8 (25.6–28.0) |
26.7 (25.5–27.9) |
44.1 (27.9–60.4) |
26.7 (25.5–28.0) |
28.0 (21.3–34.7) |
26.8 (25.6–28.0) |
26.8 (21.4–32.2) |
| Diabetes status, % (95% CI) | 11.1 (10.2–11.9) |
11.0 (10.1–11.8) |
27.2 (12.7–41.7) |
10.9 (10.1–11.8) |
15.4 (10.1–20.6) |
9.2 (8.4–10.0) |
50.5 (44.3–56.8) |
| BMI (BMI>=30), % (95% CI) | 25.3 (24.1–26.5) |
25.3 (24.1–26.5) |
27.9 (13.2–42.6) |
25.3 (24.1–26.5) |
26.6 (19.8–33.4) |
23.8 (22.7–25.0) |
53.6 (47.4–59.8) |
| Smoking status, % (95% CI) | |||||||
| Never smoker | 29.1 (27.8–30.3) |
29.1 (27.9–30.4) |
16.7 (4.5–29.0) |
29.0 (27.8–30.3) |
30.5 (23.4–37.6) |
28.9 (27.6–30.1) |
32.4 (26.5–38.2) |
| Current smoker or former smoker that quit <10 years ago | 39.5 (38.2–40.8) |
39.4 (38.0–40.7) |
64.2 (48.5–79.9) |
39.3 (38.0–40.6) |
46.3 (38.7–53.9) |
39.6 (38.3–41.0) |
37.1 (31.1–43.2) |
| Former smoker that quit >=10 years ago | 31.0 (29.7–32.2) |
31.0 (29.8–32.3) |
18.7 (6.1–31.3) |
31.2 (29.9–32.5) |
23.3 (17.1–29.4) |
31.0 (29.7–32.3) |
30.0 (24.5–35.6) |
| Drinking status, % (95% CI) | |||||||
| Never drinker | 11.1 (10.3–12.0) |
11.1 (10.3–12.0) |
8.2 (−0.7–17.1) |
11.1 (10.3–12.0) |
9.9 (5.5–14.3) |
11.1 (10.2–12.0) |
11.1 (7.3–14.9) |
| Current drinker | 64.5 (63.2–65.8) |
64.4 (63.1–65.7) |
72.6 (58.1–87.1) |
64.4 (63.0–65.7) |
67.8 (60.8–74.8) |
64.9 (63.6–66.2) |
56.0 (49.9–62.1) |
| Former drinker | 24.4 (23.2–25.5) |
24.4 (23.2–25.6) |
19.2 (6.4–32.0) |
24.4 (23.3–25.6) |
22.3 (16.1–28.4) |
23.9 (22.8–25.1) |
32.7 (27.0–38.5) |
| Statin and aspirin use, % (95% CI) | 48.6 (47.2–49.9) |
48.6 (47.3–50.0) |
47.0 (30.7–63.3) |
48.7 (47.3–50.0) |
46.9 (39.3–54.4) |
48.6 (47.3–50.0) |
47.6 (41.4–53.7) |
| Family history of prostate cancer, % (95% CI) | 6.0 (5.4–6.6) |
6.0 (5.3–6.6) |
11.0 (0.8–21.2) |
5.9 (5.3–6.6) |
7.9 (3.9–11.9) |
5.9 (5.2–6.5) |
8.7 (5.2–12.1) |
| Frequency of routine medical examinations, % (95% CI) | |||||||
| Do not have routine physical examinations | 33.8 (32.5–35.1) |
33.7 (32.4–35.0) |
44.8 (28.4–61.1) |
33.8 (32.5–35.1) |
33.5 (26.2–40.8) |
33.7 (32.4–35.0) |
35.7 (29.7–41.7) |
| At least once a year | 30.6 (29.4–31.9) |
30.6 (29.4–31.9) |
32.5 (17.1–47.8) |
30.5 (29.2–31.8) |
35.0 (27.9–42.1) |
30.2 (28.9–31.5) |
38.8 (32.8–44.8) |
| At least once every five years | 29.8 (28.6–31.1) |
30.0 (28.7–31.2) |
11.2 (0.8–21.5) |
30.0 (28.7–31.2) |
25.2 (18.5–31.8) |
30.3 (29.0–31.5) |
21.0 (16.0–26.1) |
| Less than once every five years | 5.2 (4.6–5.8) |
5.2 (4.6–5.8) |
11.1 (0.8–21.4) |
5.2 (4.6–5.8) |
5.5 (2.0–9.0) |
5.3 (4.7–5.9) |
3.6 (1.3–5.9) |
| Health insurance status, % (95% CI) | 92.2 (91.5–92.9) |
92.3 (91.5–93.0) |
80.6 (67.6–93.5) |
92.4 (91.6–93.1) |
86.9 (81.8–92.1) |
92.4 (91.7–93.2) |
87.0 (82.8–91.2) |
| Type of medical care, % (95% CI) | |||||||
| Private physician | 72.8 (71.6–74.0) |
72.8 (71.6–74.0) |
74.9 (60.8–89.1) |
72.8 (71.6–74.0) |
72.9 (66.1–79.7) |
72.7 (71.5–73.9) |
74.5 (69.1–79.9) |
| HMO | 5.8 (5.2–6.5) |
5.9 (5.2–6.5) |
2.8 (−2.6–8.2) |
5.9 (5.2–6.5) |
4.2 (1.2–7.3) |
6.0 (5.3–6.6) |
3.6 (1.3–5.9) |
| Walk-in clinic | 1.6 (1.2–1.9) |
1.6 (1.2–1.9) |
2.7 (−2.5–7.9) |
1.6 (1.2–1.9) |
2.7 (0.1–5.4) |
1.6 (1.2–1.9) |
2.2 (0.3–4.1) |
| Regular clinic | 12.7 (11.8–13.6) |
12.7 (11.8–13.6) |
13.8 (2.6–25.0) |
12.8 (11.8–13.7) |
12.5 (7.5–17.4) |
12.7 (11.8–13.6) |
13.2 (9.1–17.3) |
| Hospital ER | 3.5 (3.0–4.0) |
3.5 (3.0–4.0) |
2.7 (−2.6–8.0) |
3.5 (3.0–4.0) |
3.4 (0.5–6.2) |
3.6 (3.0–4.1) |
2.6 (0.5–4.7) |
| Other | 3.2 (2.8–3.7) |
3.2 (2.8–3.7) |
2.8 (−2.6–8.1) |
3.2 (2.7–3.7) |
4.1 (1.1–7.1) |
3.2 (2.7–3.7) |
3.5 (1.2–5.7) |
Cut point for abnormal liver fibrosis score equates to advanced fibrosis (APRI >1, FIB-4 >2.67 and NFS >0.676);
Results were centered by mean age (57.6 years).
As shown in Table 2, in each model, including the fully-adjusted model (Model 3, and Model 4 for NFS), men with a higher APRI, FIB-4, and NFS score had a lower risk of prostate cancer, although neither the associations nor the linear trends were statistically significant (all P>0.05). After stratifying by race, these associations appeared to be null in White men (Table 3). However, in Black men, prostate cancer risk was significantly lower comparing the highest to lowest quintile of FIB-4 (HR=0.47, 95%CI: 0.29–0.77; P-trend=0.004) and NFS (HR=0.56, 95%CI: 0.33–0.97; P-trend=0.03; Table 4). In Black men, the HR for APRI was modestly below the null, but was not statistically significant (HR= 0.91, 95%CI: 0.54–1.55). A one-standard deviation increase in the APRI, FIB-4, and NFS scores was associated with an 11%, 23%, and 28%, respectively, decrease in the HR of prostate cancer incidence in Black men.
Table 2.
Hazard Ratios and 95% CIs for the Association between Fibrosis Scores and Total Prostate Cancer Incidence, Atherosclerosis Risk in Communities Study, 1990–1992
| Quintile of Fibrosis Scores | P value for Trende | Per SD increase in fibrosis scoref | ||||||
|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | HR (95%CI) | P value | ||
| APRI | ||||||||
| Events, No. | 138 | 154 | 144 | 150 | 140 | 726 | ||
| Person-years | 17689 | 18811 | 18888 | 18624 | 18325 | 92338 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 1.05 (0.83,1.32) | 0.97 (0.77,1.23) | 1.00 (0.79,1.26) | 0.91 (0.72,1.16) | 0.33 | 0.98 (0.90,1.06) | 0.57 |
| Model 2 HR (95%CI)b | 1 [Reference] | 1.01 (0.80,1.27) | 0.95 (0.75,1.21) | 0.95 (0.75,1.20) | 0.85 (0.67,1.07) | 0.11 | 0.96 (0.87,1.05) | 0.34 |
| Model 3 HR (95%CI)c | 1 [Reference] | 1.03 (0.81,1.31) | 0.96 (0.75,1.22) | 0.95 (0.75,1.21) | 0.84 (0.65,1.07) | 0.08 | 0.91 (0.82,1.01) | 0.09 |
| FIB-4 | ||||||||
| Events, No. | 134 | 147 | 149 | 140 | 156 | 726 | ||
| Person-years | 19526 | 19291 | 18542 | 18462 | 16516 | 92338 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 1.01 (0.80,1.28) | 0.99 (0.78,1.26) | 0.87 (0.68,1.11) | 0.97 (0.76,1.25) | 0.69 | 0.99 (0.91,1.07) | 0.81 |
| Model 2 HR (95%CI)b | 1 [Reference] | 0.99 (0.78,1.26) | 0.94 (0.74,1.19) | 0.83 (0.64,1.06) | 0.91 (0.71,1.17) | 0.40 | 0.97 (0.89,1.06) | 0.54 |
| Model 3 HR (95%CI)c | 1 [Reference] | 0.99 (0.78,1.26) | 0.89 (0.70,1.15) | 0.81 (0.63,1.04) | 0.86 (0.66,1.11) | 0.19 | 0.92 (0.83,1.02) | 0.12 |
| NFS | ||||||||
| Events, No. | 144 | 143 | 155 | 151 | 133 | 726 | ||
| Person-years | 19709 | 19771 | 18990 | 17927 | 15941 | 92338 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 0.92 (0.73,1.16) | 0.96 (0.76,1.21) | 0.93 (0.73,1.17) | 0.82 (0.64,1.06) | 0.15 | 0.95 (0.89,1.01) | 0.13 |
| Model 2 HR (95%CI)b | 1 [Reference] | 0.92 (0.73,1.17) | 0.94 (0.74,1.18) | 0.93 (0.73,1.18) | 0.88 (0.67,1.15) | 0.39 | 0.96 (0.88,1.04) | 0.28 |
| Model 3 HR (95%CI)c | 1 [Reference] | 0.92 (0.73,1.17) | 0.87 (0.69,1.11) | 0.92 (0.71,1.17) | 0.83 (0.63,1.10) | 0.22 | 0.91 (0.81,1.01) | 0.08 |
| Model 4 HR (95%CI)d | 1 [Reference] | 0.92 (0.73,1.17) | 0.87 (0.68,1.11) | 0.90 (0.71,1.15) | 0.77 (0.59,1.00) | 0.06 | 0.88 (0.80,0.98) | 0.02 |
Model 1 adjusted for age and race*field center;
Model 2 adjusted for the age, race*field center, obesity, waist-hip ratio, diabetes status, alcohol drinking status, smoking status and family history of prostate cancer;
Model 3 adjusted for the variables in model 2 and the propensity score of socio-economic status and access to and uptake of medical care;
Model 4 adjusted for the variables in model 3 but without BMI or diabetes status;
P value for trend was calculated by modeling the median of each quintile as a continuous term;
The standard deviations (SD) for APRI, FIB-4, and NFS are 0.19, 0.68, and 1.41, respectively.
Table 3.
Hazard Ratios and 95% CIs for the Association between Fibrosis Scores and Total Prostate Cancer Incidence Among White Men, Atherosclerosis Risk in Communities Study, 1990–1992
| Quintile of Fibrosis Scores | P value for Trende | Per SD increase in fibrosis scoref | ||||||
|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | HR (95%CI) | P value | ||
| APRI | ||||||||
| Events, No. | 105 | 104 | 108 | 102 | 92 | 511 | ||
| Person-years | 14929 | 15826 | 15487 | 15087 | 13747 | 75076 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 0.93 (0.71,1.22) | 1.00 (0.76,1.31) | 0.94 (0.72,1.24) | 0.95 (0.72,1.26) | 0.82 | 1.00 (0.90,1.10) | 0.94 |
| Model 2 HR (95%CI)b | 1 [Reference] | 0.90 (0.69,1.19) | 0.98 (0.74,1.28) | 0.90 (0.69,1.19) | 0.90 (0.68,1.20) | 0.58 | 0.99 (0.89,1.09) | 0.81 |
| Model 3 HR (95%CI)c | 1 [Reference] | 0.93 (0.71,1.24) | 0.98 (0.74,1.30) | 0.93 (0.70,1.23) | 0.90 (0.67,1.21) | 0.51 | 0.97 (0.86,1.09) | 0.59 |
| FIB-4 | ||||||||
| Events, No. | 95 | 107 | 109 | 91 | 109 | 511 | ||
| Person-years | 16892 | 15946 | 15407 | 14629 | 12201 | 75076 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 1.10 (0.83,1.45) | 1.08 (0.81,1.43) | 0.89 (0.66,1.21) | 1.18 (0.88,1.60) | 0.42 | 1.02 (0.94,1.11) | 0.62 |
| Model 2 HR (95%CI)b | 1 [Reference] | 1.07 (0.81,1.42) | 1.02 (0.76,1.35) | 0.85 (0.63,1.15) | 1.13 (0.83,1.52) | 0.58 | 1.01 (0.92,1.10) | 0.84 |
| Model 3 HR (95%CI)c | 1 [Reference] | 1.06 (0.80,1.42) | 0.96 (0.71,1.28) | 0.87 (0.64,1.19) | 1.09 (0.80,1.49) | 0.65 | 1.00 (0.90,1.10) | 0.93 |
| NFS | ||||||||
| Events, No. | 109 | 108 | 108 | 101 | 85 | 511 | ||
| Person-years | 17142 | 16871 | 15499 | 14133 | 11432 | 75076 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 0.94 (0.72,1.22) | 0.95 (0.72,1.25) | 0.93 (0.70,1.23) | 0.91 (0.67,1.23) | 0.54 | 0.99 (0.89,1.09) | 0.78 |
| Model 2 HR (95%CI)b | 1 [Reference] | 0.94 (0.72,1.23) | 0.93 (0.71,1.23) | 0.96 (0.72,1.28) | 1.02 (0.74,1.40) | 0.92 | 1.02 (0.90,1.15) | 0.77 |
| Model 3 HR (95%CI)c | 1 [Reference] | 0.94 (0.71,1.23) | 0.88 (0.66,1.17) | 0.96 (0.72,1.29) | 0.99 (0.72,1.38) | 0.97 | 1.01 (0.88,1.15) | 0.94 |
| Model 4 HR (95%CI)d | 1 [Reference] | 0.93 (0.71,1.22) | 0.87 (0.65,1.15) | 0.93 (0.69,1.23) | 0.87 (0.64,1.19) | 0.43 | 0.95 (0.85,1.08) | 0.45 |
Model 1 adjusted for age and race*field center;
Model 2 adjusted for the age, race*field center, obesity, waist-hip ratio, diabetes status, alcohol drinking status, smoking status and family history of prostate cancer;
Model 3 adjusted for the variables in model 2 and the propensity score of socio-economic status and access to and uptake of medical care;
Model 4 adjusted for the variables in model 3 but without BMI or diabetes status;
P value for trend was calculated by modeling the median of each quintile as a continuous term;
The standard deviations (SD) for APRI, FIB-4, and NFS are 0.19, 0.68, and 1.41, respectively.
Table 4.
Hazard Ratios and 95% CIs for the Association between Fibrosis Scores and Total Prostate Cancer Incidence Among Black Men, Atherosclerosis Risk in Communities Study, 1990–1992
| Quintile of Fibrosis Scores | P value for Trende | Per SD increase in fibrosis scoref | ||||||
|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | HR (95%CI) | P value | ||
| APRI | ||||||||
| Events, No. | 33 | 50 | 36 | 48 | 48 | 215 | ||
| Person-years | 2761 | 2985 | 3402 | 3537 | 4578 | 17262 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 1.41 (0.91,2.18) | 0.89 (0.55,1.42) | 1.12 (0.72,1.75) | 0.87 (0.56,1.35) | 0.16 | 0.94 (0.81,1.09) | 0.41 |
| Model 2 HR (95%CI)b | 1 [Reference] | 1.36 (0.87,2.12) | 0.89 (0.55,1.44) | 1.09 (0.70,1.71) | 0.79 (0.50,1.23) | 0.06 | 0.91 (0.78,1.07) | 0.26 |
| Model 3 HR (95%CI)c | 1 [Reference] | 1.26 (0.74,2.15) | 0.69 (0.37,1.28) | 1.02 (0.59,1.76) | 0.91 (0.54,1.55) | 0.57 | 0.89 (0.73,1.09) | 0.26 |
| FIB-4 | ||||||||
| Events, No. | 39 | 40 | 40 | 49 | 47 | 215 | ||
| Person-years | 2633 | 3345 | 3135 | 3834 | 4316 | 17262 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 0.78 (0.50,1.21) | 0.77 (0.49,1.20) | 0.75 (0.49,1.15) | 0.62 (0.40,0.96) | 0.05 | 0.92 (0.79,1.07) | 0.28 |
| Model 2 HR (95%CI)b | 1 [Reference] | 0.79 (0.51,1.23) | 0.75 (0.48,1.17) | 0.74 (0.48,1.14) | 0.58 (0.37,0.91) | 0.03 | 0.90 (0.77,1.06) | 0.21 |
| Model 3 HR (95%CI)c | 1 [Reference] | 0.81 (0.51,1.30) | 0.64 (0.39,1.04) | 0.67 (0.42,1.07) | 0.47 (0.29,0.77) | 0.004 | 0.77 (0.63,0.93) | 0.01 |
| NFS | ||||||||
| Events, No. | 35 | 35 | 47 | 50 | 48 | 215 | ||
| Person-years | 2568 | 2900 | 3491 | 3794 | 4509 | 17262 | ||
| Model 1 HR (95%CI)a | 1 [Reference] | 0.85 (0.53,1.36) | 0.91 (0.59,1.42) | 0.86 (0.56,1.34) | 0.67 (0.43,1.05) | 0.09 | 0.82 (0.70,0.97) | 0.02 |
| Model 2 HR (95%CI)b | 1 [Reference] | 0.87 (0.54,1.39) | 0.90 (0.58,1.41) | 0.83 (0.53,1.30) | 0.64 (0.39,1.05) | 0.08 | 0.79 (0.66,0.96) | 0.01 |
| Model 3 HR (95%CI)c | 1 [Reference] | 0.91 (0.55,1.52) | 0.84 (0.52,1.36) | 0.80 (0.49,1.30) | 0.56 (0.33,0.97) | 0.03 | 0.72 (0.59,0.88) | 0.00 |
| Model 4 HR (95%CI)d | 1 [Reference] | 0.92 (0.55,1.52) | 0.88 (0.55,1.42) | 0.84 (0.52,1.36) | 0.60 (0.36,0.99) | 0.03 | 0.75 (0.62,0.90) | 0.00 |
Model 1 adjusted for age and race*field center;
Model 2 adjusted for the age, race*field center, obesity, waist-hip ratio, diabetes status, alcohol drinking status, smoking status and family history of prostate cancer;
Model 3 adjusted for the variables in model 2 and the propensity score of socio-economic status and access to and uptake of medical care;
Model 4 adjusted for the variables in model 3 but without BMI or diabetes status;
P value for trend was calculated by modeling the median of each quintile as a continuous term;
The standard deviations (SD) for APRI, FIB-4, and NFS are 0.19, 0.68, and 1.41, respectively.
Quintiles of the liver fibrosis scores were not clearly associated with lethal or fatal prostate cancer (all P>0.05), and most of the HRs for per standard deviation increase in fibrosis scores were close to 1 overall and in both Black and in White men (fully-adjusted model; Table 5).
Table 5.
Hazard Ratios and P for Trend for Per Standard Deviation Increase in Liver Fibrosis Scores of Lethal and Fatal Prostate Cancer, Atherosclerosis Risk in Communities Study, 1990–1992
| Totala | White | Black | ||||
|---|---|---|---|---|---|---|
| HR per SD increase in fibrosis score (95%CI)c | P value for Trendd | HR per SD increase in fibrosis score (95%CI)c | P value for Trendd | HR per SD increase in fibrosis score (95%CI)c | P value for Trendd | |
| Lethal Prostate Cancer b | ||||||
| Events, No. | 83 | 55 | 28 | |||
| Person-years | 92532 | 75263 | 17270 | |||
| APR | 1.02 (0.84,1.24) | 0.80 | 0.91 (0.61,1.35) | 0.09 | 1.08 (0.83,1.41) | 0.13 |
| FIB-4 | 1.03 (0.86,1.24) | 0.67 | 1.00 (0.76,1.32) | 0.77 | 1.06 (0.75,1.51) | 0.84 |
| NFS | 1.04 (0.76,1.43) | 0.61 | 1.22 (0.83,1.80) | 0.44 | 0.74 (0.42,1.29) | 0.79 |
| Fatal Prostate Cancer | ||||||
| Events, No. | 77 | 51 | 26 | |||
| Person-years | 105682 | 85453 | 20228 | |||
| APR | 0.89 (0.64,1.25) | 0.79 | 0.83 (0.50,1.37) | 0.17 | 0.89 (0.55,1.44) | 0.36 |
| FIB-4 | 0.95 (0.71,1.26) | 0.70 | 0.98 (0.67,1.43) | 0.62 | 0.90 (0.58,1.40) | 0.95 |
| NFS | 1.06 (0.76,1.49) | 0.50 | 1.31 (0.87,1.97) | 0.24 | 0.71 (0.39,1.29) | 0.64 |
All results were adjusted for the variables in model 3;
The associations for incidence among total, White, and Black are shown in Tables 2, 3, and 4, respectively;
The standard deviations (SD) for APRI, FIB-4, and NFS are 0.19, 0.68, and 1.41, respectively;
P value for trend was calculated by modeling the median of each quintile as a continuous term.
In the cross-categorized fibrosis scores analysis, compared with men with zero abnormal scores, men with any abnormal score had a lower risk of prostate cancer if they were Black (HR=0.46, 95%CI: 0.24–0.89), but not if they were White (HR=1.04, 95%CI: 0.69–1.58, P- interaction=0.04).
Discussion
In middle-aged and older community dwelling men without a diagnosis of chronic liver disease, we observed an inverse association between higher liver fibrosis scores and prostate cancer incidence in Black men but not in White men. Liver fibrosis scores did not appear to be associated with lethal or fatal prostate cancer in either Black or White men.
We were motivated to conduct this study following our prior work suggesting that men with higher liver fibrosis scores had lower serum PSA.(4) We had two hypotheses about how liver fibrosis might affect the apparent risk of total and fatal prostate cancer: 1) biological influences on the development and progression of prostate cancer and 2) related to the influence of liver disease on PSA levels (decreasing the accuracy). If among those with liver fibrosis, reduced PSA concentration delays prostate cancer detection, then we would expect a lower prostate cancer risk in men with liver fibrosis than in men without liver fibrosis. We would also expect a higher risk of lethal prostate cancer and prostate cancer mortality in men with liver fibrosis, given that a delay in detection leading to later treatment could increase the likelihood of lethal disease progression. Regarding biological influences, if liver fibrosis results in a lower probability of prostate cancer development, we would expect that the risk of total, lethal, and fatal prostate cancer would be lower in men with liver fibrosis than those without. Overall, our findings did not support one of these hypotheses over the other. In Black men, we observed higher liver fibrosis scores were associated with a lower risk of prostate cancer, but not with lethal or fatal prostate cancer. Thus, in Black men, the patterns together suggest that liver fibrosis may influence the observed risk of prostate cancer through the combination of biology and detection bias, if other non-causal explanations can be ruled out. Conversely, in White men, associations appeared to be null for liver fibrosis scores with total prostate cancer, and with lethal or fatal disease. In White men, both detection bias and biology could be acting, but their balance differs from that in Black men. Alternatively, we cannot rule out the roles of chance and other sources of bias.
Several aspects of this study warrant discussion. First, while the number of prostate cancer cases is relatively large, the number of prostate cancer deaths was small (77 cases in 5,284 men), resulting in limited statistical power to detect modest to moderate associations. Second, the non-invasive indicators might result in misclassification of liver fibrosis compared with imaging and the gold standard liver biopsy. However, cross-categorizing the men with all three scores supported the main findings. We expected that men “truly” without liver fibrosis would tend to score low on all three scores and thus increase the sensitivity. Third, NFS includes factors known to be associated with prostate cancer (abnormal fasting glucose, diabetes) or with lethal/fatal disease (BMI). Results for NFS, with or without further adjustment for BMI and diabetes status, showed similar results for APRI, suggesting the BMI and diabetes do not fully underlie the findings. Fourth, while we excluded men with a known clinical diagnosis of liver disease, men with and without unrecognized severe liver fibrosis might be different on many demographic, health, and lifestyle characteristics. We accounted for these by multivariable adjustment, including propensity scores for SES and associated access to care variables . Fifth, PSA-based prostate cancer screening and prostate biopsy histories are not available in ARIC. Whether any differences by race in the patterns of the recommendation and uptake of PSA screening contributed to the differences in findings between Black and White men cannot be directly determined. Finally, we were not able to conclusively distinguish between detection bias versus biology as explanations for the observed pattern. If detection bias, which stems from lower PSA in those with liver fibrosis, were the sole explanation for the inverse association between liver fibrosis scores and prostate cancer risk, we would have expected a positive association for lethal/fatal prostate cancer as result of delayed detection and treatment. While we did not observe a positive association with lethal/fatal prostate cancer, including in Black men, the number of events was small. Even longer follow up, and time-varying measures of both PSA and liver fibrosis scores, would be needed to fully capture the long-term impact of delayed detection on prostate cancer death. Thus, additional studies to determine whether liver impairment affects prostate cancer detection in Black men, is of particular importance because Black men have more than twice the risk of dying from prostate cancer than non-Hispanic White men.(29)
Despite possible limitations, ARIC has several strengths for addressing the present research question. First, to our knowledge, this study is the first and largest to examine the association between mild to moderate liver disease and prostate cancer risk. Second, ARIC is a prospective study with clinical and biomarker measures relevant to liver impairment and cancer follow-up. Liver enzymes were measured in the cohort without indication/suspicion of liver fibrosis, and when included in the scores calculations, provide a non-invasive assessment of the likelihood and severity of liver fibrosis. We were able to exclude participants with a diagnosis of liver disease, who may not be eligible for prostate cancer screening, and have a higher risk of death from liver disease, reducing the competing risk of death. Additionally, the participants were from four distinct US geographic sites and included a large proportion of Black participants, making our findings generalizable to similar groups in the US.(21)
In conclusion, in men without a clinical diagnosis of liver disease, having a higher liver fibrosis score was associated with a lower risk of prostate cancer, but only in Black men. Whether the findings for Black men reflect the influence of subclinical liver disease on the development versus the detectability of prostate cancer, including by PSA, or whether there is another explanation requires further study, as does the difference in findings by race.
Prevention Relevance Statement.
Investigating the link between liver fibrosis and prostate cancer risk and mortality, our study reveals the potential influence of liver health on prostate cancer detection using prostate-specific antigen (PSA) test, urging further research to optimize prevention and intervention strategies.
Acknowledgments
The authors thank the staff and participants of the ARIC study for their important contributions. Cancer data were provided by the Maryland Cancer Registry, Center for Cancer Prevention and Control, Maryland Department of Health, with funding from the State of Maryland and the Maryland Cigarette Restitution Fund. The collection and availability of cancer registry data is also supported by the Cooperative Agreement NU58DP006333, funded by the Centers for Disease Control and Prevention. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.
Funding:
The Atherosclerosis Risk in Communities study has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under Contract nos. (75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005). Studies on cancer in ARIC are also supported by the National Cancer Institute (U01 CA164975). E.A. Platz and J. Lu were supported by the Maryland Cigarette Restitution Fund at Johns Hopkins. C.E. Joshu was supported by the Prostate Cancer Foundation.
Footnotes
Conflict of interest statement: The authors declare no potential conflicts of interest.
References
- 1.Vicentini FC, Botelho LA, Hisano M, Ebaid GX, Lucon M, Lucon AM, et al. Are total prostate-specific antigen serum levels in cirrhotic men different from those in normal men? Urology 2009;73(5):1032–5 doi 10.1016/j.urology.2009.01.015. [DOI] [PubMed] [Google Scholar]
- 2.Akdogan M, Hassoun BS, Gurakar A, El-Sahwi K, Jazzar A, Wright H, et al. Prostate-specific antigen levels among cirrhotic patients. Int J Biol Markers 2002;17(3):161–4 doi 10.1177/172460080201700303. [DOI] [PubMed] [Google Scholar]
- 3.Oesterling JE, Jacobsen SJ, Chute CG, Guess HA, Girman CJ, Panser LA, et al. Serum prostate-specific antigen in a community-based population of healthy men. Establishment of age-specific reference ranges. JAMA 1993;270(7):860–4. [PubMed] [Google Scholar]
- 4.Wang A, Lazo M, Carter HB, Groopman JD, Nelson WG, Platz EA. Association between Liver Fibrosis and Serum PSA among US Men: National Health and Nutrition Examination Survey (NHANES), 2001–2010. Cancer Epidemiol Biomarkers Prev 2019;28(8):1331–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fenton JJ, Weyrich MS, Durbin S, Liu Y, Bang H, Melnikow J. U.S. Preventive Services Task Force Evidence Syntheses, formerly Systematic Evidence Reviews. Prostate-Specific Antigen-Based Screening for Prostate Cancer: A Systematic Evidence Review for the US Preventive Services Task Force, U.S. Preventive Services Task Force Evidence Syntheses, formerly Systematic Evidence Reviews. Rockville (MD): Agency for Healthcare Research and Quality (US); 2018. [PubMed] [Google Scholar]
- 6.Mokdad AA, Lopez AD, Shahraz S, Lozano R, Mokdad AH, Stanaway J, et al. Liver cirrhosis mortality in 187 countries between 1980 and 2010: a systematic analysis. BMC Med 2014;12:145doi 10.1186/s12916-014-0145-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Denniston MM, Jiles RB, Drobeniuc J, Klevens RM, Ward JW, McQuillan GM, et al. Chronic hepatitis C virus infection in the United States, National Health and Nutrition Examination Survey 2003 to 2010. Ann Intern Med 2014;160(5):293–300 doi 10.7326/M13-1133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Heidenreich A, Abrahamsson PA, Artibani W, Catto J, Montorsi F, Van Poppel H, et al. Early detection of prostate cancer: European Association of Urology recommendation. Eur Urol 2013;64(3):347–54 doi 10.1016/j.eururo.2013.06.051. [DOI] [PubMed] [Google Scholar]
- 9.Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG, et al. The influence of finasteride on the development of prostate cancer. The New England journal of medicine 2003;349(3):215–24 doi 10.1056/NEJMoa030660. [DOI] [PubMed] [Google Scholar]
- 10.Nickel JC, Gilling P, Tammela TL, Morrill B, Wilson TH, Rittmaster RS. Comparison of dutasteride and finasteride for treating benign prostatic hyperplasia: the Enlarged Prostate International Comparator Study (EPICS). BJU International 2011;108(3):388–94 doi 10.1111/j.1464-410X.2011.10195.x. [DOI] [PubMed] [Google Scholar]
- 11.Young CY, Andrews PE, Tindall DJ. Expression and androgenic regulation of human prostate-specific kallikreins. J Androl 1995;16(2):97–9. [PubMed] [Google Scholar]
- 12.Banez LL, Loftis RM, Freedland SJ, Presti JC Jr., Aronson WJ, Amling CL, et al. The influence of hepatic function on prostate cancer outcomes after radical prostatectomy. Prostate Cancer Prostatic Dis 2010;13(2):173–7 doi 10.1038/pcan.2010.3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Al-Azab R, Toi A, Lockwood G, Kulkarni GS, Fleshner N. Prostate volume is strongest predictor of cancer diagnosis at transrectal ultrasound-guided prostate biopsy with prostate-specific antigen values between 2.0 and 9.0 ng/mL. Urology 2007;69(1):103–7 doi 10.1016/j.urology.2006.09.041. [DOI] [PubMed] [Google Scholar]
- 14.Marchesini G, Brizi M, Morselli-Labate AM, Bianchi G, Bugianesi E, McCullough AJ, et al. Association of nonalcoholic fatty liver disease with insulin resistance. Am J Med 1999;107(5):450–5 doi 10.1016/s0002-9343(99)00271-5. [DOI] [PubMed] [Google Scholar]
- 15.Albanes D, Weinstein SJ, Wright ME, Mannisto S, Limburg PJ, Snyder K, et al. Serum insulin, glucose, indices of insulin resistance, and risk of prostate cancer. J Natl Cancer Inst 2009;101(18):1272–9 doi 10.1093/jnci/djp260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yun SJ, Min BD, Kang HW, Shin KS, Kim TH, Kim WT, et al. Elevated insulin and insulin resistance are associated with the advanced pathological stage of prostate cancer in Korean population. J Korean Med Sci 2012;27(9):1079–84 doi 10.3346/jkms.2012.27.9.1079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Krystyna A, Safi T, Briggs WM, Schwalb MD. Correlation of hepatitis C and prostate cancer, inverse correlation of basal cell hyperplasia or prostatitis and epidemic syphilis of unknown duration. Int Braz J Urol 2011;37(2):223–9; discussion 30 doi 10.1590/s1677-55382011000200009. [DOI] [PubMed] [Google Scholar]
- 18.Perez-Cornago A, Fensom GK, Andrews C, Watts EL, Allen NE, Martin RM, et al. Examination of potential novel biochemical factors in relation to prostate cancer incidence and mortality in UK Biobank. Br J Cancer 2020;123(12):1808–17 doi 10.1038/s41416-020-01081-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Choi WM, Lee JH, Yoon JH, Kwak C, Lee YJ, Cho YY, et al. Nonalcoholic fatty liver disease is a negative risk factor for prostate cancer recurrence. Endocr Relat Cancer 2014;21(2):343–53 doi 10.1530/ERC-14-0036. [DOI] [PubMed] [Google Scholar]
- 20.Wright JD, Folsom AR, Coresh J, Sharrett AR, Couper D, Wagenknecht LE, et al. The ARIC (Atherosclerosis Risk In Communities) Study: JACC Focus Seminar 3/8. J Am Coll Cardiol 2021;77(23):2939–59 doi 10.1016/j.jacc.2021.04.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Joshu CE, Barber JR, Coresh J, Couper DJ, Mosley TH, Vitolins MZ, et al. Enhancing the Infrastructure of the Atherosclerosis Risk in Communities (ARIC) Study for Cancer Epidemiology Research: ARIC Cancer. Cancer Epidemiol Biomarkers Prev 2018;27(3):295–305 doi 10.1158/1055-9965.EPI-17-0696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Udompap P, Mannalithara A, Heo NY, Kim D, Kim WR. Increasing prevalence of cirrhosis among U.S. adults aware or unaware of their chronic hepatitis C virus infection. J Hepatol 2016;64(5):1027–32 doi 10.1016/j.jhep.2016.01.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.McPherson S, Stewart SF, Henderson E, Burt AD, Day CP. Simple non-invasive fibrosis scoring systems can reliably exclude advanced fibrosis in patients with non-alcoholic fatty liver disease. Gut 2010;59(9):1265–9 doi 10.1136/gut.2010.216077. [DOI] [PubMed] [Google Scholar]
- 24.Vallet-Pichard A, Mallet V, Nalpas B, Verkarre V, Nalpas A, Dhalluin-Venier V, et al. FIB-4: an inexpensive and accurate marker of fibrosis in HCV infection. comparison with liver biopsy and fibrotest. Hepatology 2007;46(1):32–6 doi 10.1002/hep.21669. [DOI] [PubMed] [Google Scholar]
- 25.Mondul AM, Joshu CE, Barber JR, Prizment AE, Bhavsar NA, Selvin E, et al. Longer-term Lipid-lowering Drug Use and Risk of Incident and Fatal Prostate Cancer in Black and White Men in the ARIC Study. Cancer Prev Res (Phila) 2018;11(12):779–88 doi 10.1158/1940-6207.CAPR-17-0396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Flores YN, Yee HF Jr., Leng M, Escarce JJ, Bastani R, Salmeron J, et al. Risk factors for chronic liver disease in Blacks, Mexican Americans, and Whites in the United States: results from NHANES IV, 1999–2004. Am J Gastroenterol 2008;103(9):2231–8 doi 10.1111/j.1572-0241.2008.02022.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Shoham DA, Vupputuri S, Diez Roux AV, Kaufman JS, Coresh J, Kshirsagar AV, et al. Kidney disease in life-course socioeconomic context: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Kidney Dis 2007;49(2):217–26 doi 10.1053/j.ajkd.2006.11.031. [DOI] [PubMed] [Google Scholar]
- 28.Foraker RE, Patel MD, Whitsel EA, Suchindran CM, Heiss G, Rose KM. Neighborhood socioeconomic disparities and 1-year case fatality after incident myocardial infarction: the Atherosclerosis Risk in Communities (ARIC) Community Surveillance (1992–2002). Am Heart J 2013;165(1):102–7 doi 10.1016/j.ahj.2012.10.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin 2022;72(1):7–33 doi 10.3322/caac.21708. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data analyzed in this study are available from the Atherosclerosis Risk in Communities (ARIC) study. Detailed information about data access and sharing is available online (https://sites.cscc.unc.edu/aric/sites/default/files/public/listings/ARIC%20data%20sharing.pdf).