Abstract
Background
Hypertension and the use of antihypertensive medications have been intensively investigated in relation to colorectal cancer (CRC). Prior epidemiologic studies have not been able to examine this topic with adequate confounding control and follow-up time, or disentangle the effects of antihypertensive agents and hypertension.
Methods
Eligible participants in the Nurses’ Health Study and Health Professionals Follow-up Study were followed for up to 28 years, with repeat assessments of exposures. Cox proportional hazards models were used to estimate hazard ratios and 95% confidence intervals.
Results
In fully adjusted analyses based on both new-user and prevalent-user designs, there was no association between the use of beta-blockers, calcium-channel blockers, thiazide diuretics, angiotensin-converting enzyme inhibitors, furosemide, other antihypertensive drugs and CRC risk and mortality reached the statistically significant threshold after Bonferroni correction. The results remained similar in sensitivity analyses among participants with hypertension. Before Bonferroni correction, suggestive associations between beta-blocker use and CRC risk and between furosemide use and CRC-specific mortality were observed specifically in analyses using a new-user design. Hypertension was not associated with CRC risk in analyses based on both new-user and prevalent-user designs.
Conclusions
Hypertension and long-term use of major classes of antihypertensive medications are unlikely to be associated with CRC risk and mortality.
Subject terms: Colon cancer, Cancer epidemiology
Introduction
Over the past decades, hypertension [1] and major classes of antihypertensive agents, such as angiotensin-converting enzyme inhibitors (ACEIs) [2–10], angiotensin receptor blockers (ARBs) [3, 9–13], beta-blockers [8, 10, 14–19], calcium-channel blockers (CCBs) [7, 8, 20, 21] and diuretics [7], have been investigated in relation to colorectal cancer (CRC) risk [1–3, 6–9, 12–15, 21] and mortality [3, 9, 10, 14, 16–20]. Current epidemiologic evidence remains inconclusive with inconsistent findings reported [1–21], though mechanistic studies provided initial evidence on a range of plausible biological links [22–25].
The interpretation of most of the previous effect estimates may have been limited by their study design (e.g., the concern of recall bias in retrospective studies), imprecise/incomplete assessment of exposures (e.g., missing repeat assessments of antihypertensive medication use and hypertension throughout follow-up, and therefore unable to explore duration-dependent associations), inadequate control for confounding (missing information on crucial covariates, missing updated covariate information, or without taking into account confounding by indication, etc.), and unable to disentangle the mixed effects of antihypertensive medication use and hypertension. In addition, adenomas (CRC precursors) generally require at least 10 years to evolve into CRC [26, 27]. The crucial concept of “delayed effect”, initially suggested by aspirin and CRC risk analyses [27–30], has been acknowledged by the US Preventive Services Task Force Recommendation Statement based on high-quality epidemiologic and interventional evidence [31], suggesting that the apparent effect of medication use (e.g., aspirin) on the risk of CRC in average-risk population would require approximately 10 or more years of usage. To our knowledge, few prior investigations have been able to assess this topic in the context of long-term exposure (i.e., >10 years), presenting another important disadvantage.
To address these limitations, we investigated prospectively the associations between long-term use of major classes of antihypertensive medications, hypertension, and CRC risk and mortality in the Nurses’ Health Study (NHS) [32] and Health Professionals Follow-up Study (HPFS) [33]. We hypothesised that long-term use of antihypertensive medications and hypertension may be associated with colorectal cancer incidence.
Methods
Study population
The NHS and HPFS are ongoing large prospective cohort studies of US healthcare professionals [32, 33]. The NHS was initiated in 1976 [32], when 121,700 female registered nurses aged 30–55 years were enrolled. The HPFS, which began in 1986 [33], enrolled 51,529 male health professionals between 40 and 75 years of age. Demographics were collected at cohort recruitment. Via self-administered questionnaires, participants biennially or quadrennially reported data on anthropometrics, lifestyles, dietary factors, menstrual and reproductive history (women only), family history, and medical history thereafter throughout follow-up. In both cohorts, follow-up rates have consistently exceeded 90%. Participants who were alive and free of cancer at the time when information on exposures were first assessed were included in analyses of both CRC risk and CRC-specific mortality.
Ascertainment of antihypertensive medication use and hypertension
Information on antihypertensive medication use by classes (beta-blockers, CCBs, thiazide diuretics, ACEIs and other antihypertensive drugs) was first assessed in the NHS in 1988, and with biennial updates throughout follow-up since 1994 (except ACEI use which was updated biennially from 1996); furosemide use was first assessed in 1994 and with biennial updates thereafter. In the HPFS, beta-blockers, CCBs, thiazide diuretics, and furosemide were first assessed in 1986, and with subsequent updates every 2 years throughout follow-up; ACEI use was ascertained biennially from 2004 and thereafter. In both cohorts, participants were queried about if they used antihypertensive medications regularly in the past 2 years. The total duration of antihypertensive medication use was calculated by summing previous years of regular use, first based on new-user design to mitigate confounding by indication (primary analyses) [34], then based on prevalent-user design (sensitivity analyses). We were not able to consider individual evaluation of ARBs because ARBs were first queried relatively late (NHS: 2008; HPFS: 2006). ARBs were therefore combined with “other antihypertensive drugs”. Throughout follow-up in both the NHS and HPFS, participants biennially updated information on physician-diagnosed hypertension, the validity of which have previously been reported (among random sample of participants in the NHS reporting physician-diagnosed hypertension, all self-reports were confirmed by medical records) [35].
Ascertainment of covariates
Age, sex, race, smoking, physical activity, body mass index (BMI), alcohol intake, regular use of aspirin, family history of CRC, history of diabetes mellitus, multivitamin use, CRC screening and diet were selected a priori as potential covariates [36, 37] (Supplementary Methods).
Ascertainment of outcomes
Physician-diagnosed incident CRC events were reported by participants via biennial questionnaires. Cohort investigators obtained written informed consent from participants or next-of-kin to retrieve medical records and pathology reports for the purpose of ascertaining CRC diagnoses or referred to state cancer registries when medical records were unavailable.
Through the National Death Index, postal authorities or next-of-kin, death events were confirmed with an identifying rate exceeding 96% [38]. Cohort investigators reviewed death certificates and medical records after obtaining permission from the next-of-kin of deceased participants. Causes of death were assigned according to the International Classification of Diseases, 8th Revision (ICD-8).
Statistical analysis
In analyses of CRC risk, person-years of follow-up were calculated from the return of the baseline questionnaire until the date of diagnosis of any cancer (except nonmelanoma skin cancers), death recorded, loss to follow-up, or predefined follow-up completion (NHS: June 1, 2014; HPFS: January 31, 2014), whichever was earliest. In analyses of CRC-specific mortality, follow-up accrued from the baseline questionnaire return date until the date of death recorded, loss to follow-up or follow-up completion, whichever occurred first.
The primary exposures include the total duration of antihypertensive medication use and the total duration of hypertension. No statistically significant heterogeneity by cohort was detected for each exposure (P > 0.37 for heterogeneity by cohort). Pooled analyses were conducted in the combined dataset. The proportional hazard assumption was verified using interactions between the exposures of interest and the log-transformed time scale, and we detected no violation. We used Cox proportional hazard models to estimate age- and multivariable-adjusted hazard ratios (HRs) and 95% confidence intervals (95% CIs) for incident CRC and CRC-specific mortality, across categories of exposures. To further explore potential duration-dependent associations between the exposures and outcomes, we estimated HRs and 95% CIs per 5-year increment in total duration of antihypertensive medication use and history of hypertension separately, where the P value for trend was calculated by modelling duration as a continuous variable.
We used a new-user design in primary analyses. We conducted additional analyses based on an active-comparator new-user design to further ensure minimised confounding by indication. We performed joint association analyses of hypertension and any antihypertensive medication use with CRC risk by using the collapsed categories to disentangle their effects. In sensitivity analyses, we repeated antihypertensive medication analyses among participants with hypertension only, and conducted further analyses according to anatomic subsites. Finally, we repeated analyses based on prevalent-user design.
Multivariable analyses were stratified by age (months), sex (women, men), and questionnaire cycle (each 2-year interval), and adjusted for the race (White, Black, others), pack-years of smoking (pack-years, continuous), physical activity (MET-hours/week, continuous), BMI (kg/m2, continuous), alcohol consumption (grams/day, continuous), Alternate Healthy Eating Index (continuous), regular use of aspirin (tablets/week, continuous), family history of CRC (yes, no), history of diabetes mellitus (yes, no), multivitamin use (yes, no), screening colonoscopy or sigmoidoscopy (yes, no), total calorie intake (kcal/day, continuous), red or processed meat intake (servings/day, continuous), fibre intake (g/day, continuous), folate intake (µg/day, continuous), calcium intake (mg/day, continuous), and vitamin D intake (IU/day, continuous). Specifically, in analyses of antihypertensive medications, models were additionally adjusted for history of hypertension (yes, no), plus mutually adjusted for other classes of antihypertensive medications than the one under study.
All data analyses were conducted using SAS statistical software (version 9.4 for UNIX; SAS Institute Inc., Cary, NC). All tests were two-sided. To account for multiple testing in analyses of antihypertensive medication use, Bonferroni correction was used to adjust the statistical significance level to α of 0.008 (0.05/6).
Results
Population characteristics
During up to 28 years of follow-up, we identified 2343 incident CRC cases among 110,431 eligible participants in analyses based on new-user design, and we identified 3328 incident CRC cases among 143,309 eligible participants in analyses based on prevalent-user design. Compared to participants never used antihypertensive medications, those reporting longer duration of antihypertensive medication use tended to be older, had a higher BMI, and engaged in less physical activity. They were more likely to be past-smokers but less likely to be current smokers, and reported fewer pack-years of smoking. They were much more likely to have diagnoses of type 2 diabetes and hypertension, received more frequent colonoscopy or sigmoidoscopy screening, and used more aspirin and multivitamins. They also tended to have a higher intake of folate, vitamin D, and calcium, and worse overall diet quality (Table 1).
Table 1.
Participant characteristics according to total duration of antihypertensive medication usea (new-user design).
| Characteristics | Total duration of antihypertensive medication use (years) | |||
|---|---|---|---|---|
| 0 | 1–5 | 6–10 | >11 | |
| Age, years | 62.8 (9.7) | 68.6 (9.0) | 71.3 (8.2) | 74.5 (7.4) |
| White, % | 96.9 | 96.5 | 97.1 | 97.1 |
| Body mass indexb, kg/m² | 24.7 (3.8) | 26.0 (4.3) | 26.6 (4.4) | 27.2 (4.5) |
| Physical activityc, METs-hours/week | 22.8 (21.9) | 20.9 (19.5) | 20.4 (18.1) | 20.2 (16.2) |
| Past smoking, % | 42 | 46 | 46 | 46 |
| Current smoking, % | 10 | 8 | 6 | 4 |
| Pack-years of smokingd | 12.1 (18.9) | 12.9 (19.4) | 12.0 (18.7) | 10.9 (17.2) |
| Type 2 diabetes, % | 3 | 8 | 10 | 15 |
| Hypertension, % | 15 | 76 | 88 | 93 |
| Family history of colorectal cancer, % | 16 | 15 | 15 | 14 |
| Screening colonoscopy or sigmoidoscopy, % | 36 | 44 | 52 | 66 |
| Polyps, % | 9 | 9 | 10 | 11 |
| Aspirin use, tablets/week | 2.5 (3.2) | 3.0 (3.4) | 3.2 (3.4) | 3.3 (3.0) |
| Multivitamin use, % | 47 | 55 | 61 | 65 |
| Total folate intake, µg/day | 489 (221) | 503 (213) | 522 (202) | 557 (188) |
| Total vitamin D, IU/day | 396 (230) | 403 (220) | 409 (208) | 435 (195) |
| Total calcium intake, mg/day | 986 (370) | 1002 (368) | 1027 (362) | 1060 (346) |
| AHEI | 47.6 (9.4) | 47.2 (9.1) | 47.1 (9.0) | 47.6 (8.6) |
| Alcohol, g/day | 7.4 (10.7) | 7.4 (10.9) | 7.3 (10.5) | 7.3 (10.0) |
| Red or processed meat, servings/week | 6.0 (3.6) | 6.0 (3.5) | 5.9 (3.2) | 5.9 (3.0) |
| Total fibre, g/day | 19.6 (5.8) | 19.4 (5.5) | 19.3 (5.2) | 19.7 (4.9) |
| Total calorie intake, kcal/day | 1815 (502) | 1811 (495) | 1811 (489) | 1816 (460) |
METs metabolic equivalent tasks, AHEI Alternate Healthy Eating Index.
aUpdated information throughout follow-up (NHS: 1990–2014; HPFS: 1988–2014) was used to calculate the mean (SD) for continuous variables and percentage for categorical variables. All variables are age-standardised except age.
bBody mass index was calculated as weight in kilograms divided by the square of height in metres.
cWeekly energy expenditure in MET-hours/week from recreational and leisure time physical activity.
dCumulative among smokers.
Antihypertensive medication use and CRC risk
In fully adjusted analyses using new-user design, long-term use of beta-blockers (HR, 95% CI vs. never users was 1.38, 0.90–2.11 for ≥11 years; 1.13, 1.01–1.27 for per 5-year increment; P = 0.03 for trend) was associated with increased CRC risk; however, it did not reach the statistically significant threshold after Bonferroni correction. Long-term use of CCBs (0.92, 0.57–1.46 for ≥11 years; 1.01, 0.90–1.12 for per 5-year increment; P = 0.91 for trend), thiazide diuretics (0.84, 0.47–1.51 for ≥11 years; 1.03, 0.91–1.17 for per 5-year increment; P = 0.65 for trend), ACEIs (0.57, 0.23–1.39 for ≥11 years; 0.91, 0.76–1.08 for per 5-year increment; P = 0.26 for trend), furosemide (1.00, 0.65–1.53 for ≥6 years; 1.12, 0.90–1.38 for per 5-year increment; P = 0.31 for trend), and other antihypertensive drugs (1.12, 0.59–2.12 for ≥11 years; 1.08, 0.94–1.23 for per 5-year increment; P = 0.29 for trend) was not associated with risk of CRC (Table 2).
Table 2.
Colorectal cancer incidence according to the total duration of antihypertensive medication use (new-user design).
| Total duration of antihypertensive medication use (years) | Per 5-year increment HR (95% CI) | P trende | ||||
|---|---|---|---|---|---|---|
| 0 | 1–5 | 6–10* | ≥11 | |||
| Beta-blockersa | ||||||
| No. of cases | 1942 | 244 | 120 | 37 | ||
| Age-adjustedb | 1 [Ref] | 1.04 (0.91–1.20) | 1.15 (0.95–1.39) | 1.18 (0.85–1.65) | 1.08 (0.99–1.17) | 0.10 |
| MVc | 1 | 1.09 (0.94–1.25) | 1.23 (1.01–1.50) | 1.32 (0.94–1.85) | 1.12 (1.03–1.23) | 0.01 |
| MV + mutual adjustmentd | 1 | 1.03 (0.85–1.25) | 1.24 (0.96–1.60) | 1.38 (0.90–2.11) | 1.13 (1.01–1.27) | 0.03 |
| CCBsa | ||||||
| No. of cases | 2049 | 198 | 77 | 19 | ||
| Age-adjustedb | 1 | 1.08 (0.93–1.25) | 0.95 (0.75–1.19) | 0.89 (0.56–1.42) | 1.00 (0.90–1.11) | 0.96 |
| MVc | 1 | 1.11 (0.95–1.29) | 1.00 (0.79–1.26) | 0.94 (0.59–1.50) | 1.02 (0.92–1.14) | 0.66 |
| MV + mutual adjustmentd | 1 | 1.08 (0.93–1.27) | 0.96 (0.76–1.23) | 0.92 (0.57–1.46) | 1.01 (0.90–1.12) | 0.91 |
| Thiazide diureticsa | ||||||
| No. of cases | 2058 | 200 | 73 | 12 | ||
| Age-adjustedb | 1 | 1.08 (0.93–1.26) | 1.05 (0.82–1.33) | 0.85 (0.48–1.51) | 1.03 (0.92–1.16) | 0.59 |
| MVc | 1 | 1.10 (0.94–1.29) | 1.09 (0.85–1.39) | 0.89 (0.50–1.58) | 1.05 (0.93–1.19) | 0.41 |
| MV + mutual adjustmentd | 1 | 1.08 (0.92–1.27) | 1.05 (0.82–1.35) | 0.84 (0.47–1.51) | 1.03 (0.91–1.17) | 0.65 |
| ACEIsa | ||||||
| No. of cases | 1279 | 127 | 40 | 5 | ||
| Age-adjustedb | 1 | 0.97 (0.80–1.18) | 0.90 (0.65–1.25) | 0.56 (0.23–1.35) | 0.91 (0.78–1.07) | 0.28 |
| MVc | 1 | 1.00 (0.81–1.22) | 0.92 (0.66–1.29) | 0.59 (0.24–1.45) | 0.92 (0.78–1.09) | 0.34 |
| MV + mutual adjustmentd | 1 | 0.97 (0.79–1.20) | 0.88 (0.63–1.24) | 0.57 (0.23–1.39) | 0.91 (0.76–1.08) | 0.26 |
| Furosemidea | ||||||
| No. of cases | 1948 | 101 | 23 | – | ||
| Age-adjustedb | 1 | 1.37 (1.11–1.68) | 1.17 (0.77–1.77) | – | 1.26 (1.03–1.54) | 0.02 |
| MVc | 1 | 1.27 (1.03–1.56) | 1.04 (0.68–1.59) | – | 1.16 (0.94–1.42) | 0.16 |
| MV + mutual adjustmentd | 1 | 1.23 (1.00–1.52) | 1.00 (0.65–1.53) | – | 1.12 (0.90–1.38) | 0.31 |
| Other antihypertensive drugsa | ||||||
| No. of cases | 1972 | 284 | 77 | 10 | ||
| Age-adjustedb | 1 | 1.04 (0.91–1.18) | 1.12 (0.88–1.41) | 1.06 (0.56–2.00) | 1.05 (0.93–1.19) | 0.40 |
| MVc | 1 | 1.06 (0.92–1.22) | 1.16 (0.91–1.48) | 1.11 (0.59–2.10) | 1.08 (0.95–1.23) | 0.26 |
| MV + mutual adjustmentd | 1 | 1.06 (0.91–1.22) | 1.15 (0.89–1.47) | 1.12 (0.59–2.12) | 1.08 (0.94–1.23) | 0.29 |
CCB calcium-channel blocker, ACEI angiotensin-converting enzyme inhibitor, MV multivariate, HR hazard ratio, CI confidence interval, AHEI Alternate Healthy Eating Index.
aIndividual classes of antihypertensive drugs have different timelines in analyses based on new-user design: beta-blockers (NHS: 1990–2014; HPFS: 1988–2014), CCBs (NHS: 1990–2014; HPFS: 1988–2014), thiazide diuretics (NHS: 1990–2014; HPFS: 1988–2014), ACEIs (NHS: 1990–2014; HPFS: 2006–2014), furosemide (NHS: 1996–2014; HPFS: 1988–2014), other antihypertensive drugs (NHS: 1990–2014; HPFS: 1988–2014).
bStratified by age, sex and follow-up cycle.
cStratified by age, sex and follow-up cycle; adjusted for race, pack-years of smoking, physical activity, BMI, alcohol consumption, AHEI, regular use of aspirin, family history of colorectal cancer, history of diabetes mellitus, multivitamin use, screening colonoscopy or sigmoidoscopy, total calorie intake, red or processed meat intake, fibre intake, folate intake, calcium intake, vitamin D intake and history of hypertension.
dAdjusted for all above-mentioned covariates, plus mutually adjusted for other classes of antihypertensive medications than the one under study.
eP value for trend was the P value for variables modelled as continuous.
*In all, 6–10 for beta-blockers, CCBs, thiazide diuretics, ACEIs and other antihypertensive drugs; ≥6 for furosemide.
In analyses based on prevalent-user design, we detected no association between major individual classes of antihypertensive medication use and CRC risk as well (Table 3). The results remained similar in sensitivity analyses restricted to participants with hypertension, and were largely similar for all subsites (Supplementary Tables 1–4). In further analyses using active-comparator new-user design, we detected no association between major classes of antihypertensive medication and CRC risk as well (Table 4).
Table 3.
Colorectal cancer incidence according to the total duration of antihypertensive medication use (prevalent-user design).
| Total duration of antihypertensive medication use (years) | Per 5-year increment HR (95% CI) | P trende | ||||
|---|---|---|---|---|---|---|
| 0 | 1–5 | 6–10 | ≥11 | |||
| Beta-blockersa | ||||||
| No. of cases | 2493 | 418 | 267 | 150 | ||
| Age-adjustedb | 1 [Ref] | 1.04 (0.93–1.15) | 0.97 (0.85–1.11) | 1.14 (0.96–1.35) | 1.03 (0.98–1.08) | 0.24 |
| MVc | 1 | 1.04 (0.93–1.16) | 0.99 (0.87–1.13) | 1.18 (0.99–1.41) | 1.04 (0.99–1.09) | 0.11 |
| MV + mutual adjustmentd | 1 | 0.99 (0.84–1.16) | 0.99 (0.83–1.18) | 1.08 (0.86–1.36) | 1.02 (0.95–1.08) | 0.63 |
| CCBsa | ||||||
| No. of cases | 2739 | 271 | 222 | 96 | ||
| Age-adjustedb | 1 | 1.04 (0.92–1.18) | 1.09 (0.95–1.26) | 1.03 (0.83–1.27) | 1.03 (0.97–1.09) | 0.39 |
| MVc | 1 | 1.05 (0.92–1.19) | 1.12 (0.97–1.30) | 1.06 (0.86–1.31) | 1.04 (0.98–1.10) | 0.22 |
| MV + mutual adjustmentd | 1 | 1.04 (0.91–1.19) | 1.11 (0.96–1.28) | 1.03 (0.83–1.28) | 1.03 (0.97–1.09) | 0.41 |
| Thiazide diureticsa | ||||||
| No. of cases | 2511 | 390 | 333 | 94 | ||
| Age-adjustedb | 1 | 1.08 (0.97–1.21) | 1.13 (1.01–1.27) | 1.05 (0.85–1.30) | 1.05 (1.00–1.11) | 0.053 |
| MVc | 1 | 1.08 (0.96–1.21) | 1.13 (1.00–1.28) | 1.07 (0.87–1.33) | 1.05 (1.00–1.11) | 0.07 |
| MV + mutual adjustmentd | 1 | 1.06 (0.94–1.20) | 1.11 (0.98–1.27) | 1.04 (0.83–1.29) | 1.04 (0.98–1.11) | 0.16 |
| ACEIsa | ||||||
| No. of cases | 1817 | 222 | 104 | 24 | ||
| Age-adjustedb | 1 | 1.02 (0.87–1.19) | 0.89 (0.71–1.11) | 0.79 (0.52–1.18) | 0.95 (0.87–1.04) | 0.26 |
| MVc | 1 | 1.03 (0.88–1.21) | 0.88 (0.70–1.12) | 0.78 (0.52–1.18) | 0.95 (0.86–1.04) | 0.26 |
| MV + mutual adjustmentd | 1 | 1.02 (0.87–1.20) | 0.86 (0.68–1.09) | 0.75 (0.49–1.13) | 0.93 (0.84–1.03) | 0.14 |
| Furosemidea | ||||||
| No. of cases | 2624 | 175 | 58 | 9 | ||
| Age-adjustedb | 1 | 1.20 (1.03–1.41) | 1.25 (0.96–1.63) | 0.98 (0.50–1.89) | 1.16 (1.03–1.31) | 0.02 |
| MVc | 1 | 1.11 (0.95–1.30) | 1.13 (0.86–1.48) | 0.86 (0.44–1.67) | 1.08 (0.95–1.23) | 0.22 |
| MV + mutual adjustmentd | 1 | 1.08 (0.92–1.27) | 1.08 (0.83–1.42) | 0.81 (0.42–1.58) | 1.05 (0.93–1.20) | 0.44 |
| Other antihypertensive drugsa | ||||||
| No. of cases | 2651 | 469 | 176 | 32 | ||
| Age-adjustedb | 1 | 1.03 (0.93–1.14) | 1.11 (0.95–1.30) | 1.05 (0.74–1.50) | 1.05 (0.97–1.13) | 0.23 |
| MVc | 1 | 1.02 (0.92–1.14) | 1.10 (0.93–1.29) | 1.03 (0.72–1.47) | 1.04 (0.96–1.13) | 0.35 |
| MV + mutual adjustmentd | 1 | 1.01 (0.91–1.13) | 1.07 (0.91–1.27) | 1.00 (0.70–1.45) | 1.03 (0.94–1.12) | 0.55 |
CCB calcium-channel blocker, ACEI angiotensin-converting enzyme inhibitor, MV multivariate, HR hazard ratio, CI confidence interval, AHEI Alternate Healthy Eating Index.
aIndividual classes of antihypertensive drugs have different timelines in analyses based on prevalent-user design: beta-blockers (NHS: 1988–2014; HPFS: 1986–2014), CCBs (NHS: 1988–2014; HPFS: 1986–2014), thiazide diuretics (NHS: 1988–2014; HPFS: 1986–2014), ACEIs (NHS: 1988–2014; HPFS: 2004–2014), furosemide (NHS: 1994–2014; HPFS: 1986–2014), other antihypertensive drugs (NHS: 1988–2014; HPFS: 1986–2014).
bStratified by age, sex and follow-up cycle.
cStratified by age, sex and follow-up cycle; adjusted for race, pack-years of smoking, physical activity, BMI, alcohol consumption, AHEI, regular use of aspirin, family history of colorectal cancer, history of diabetes mellitus, multivitamin use, screening colonoscopy or sigmoidoscopy, total calorie intake, red or processed meat intake, fibre intake, folate intake, calcium intake, vitamin D intake and history of hypertension.
dAdjusted for all above-mentioned covariates, plus mutually adjusted for other classes of antihypertensive medications than the one under study.
eP value for trend was the P value for variables modelled as continuous.
Table 4.
Colorectal cancer incidence according to the total duration of antihypertensive medication use (active-comparator, new-user design).
| Reference population | |||||||
|---|---|---|---|---|---|---|---|
| Versus normotensive | Versus hypertensive, untreated | Versus hypertensive, treated with any other antihypertensive medications | |||||
| No. of cases | Age-adjustedb | MV-adjustedc | Age-adjustedb | MV-adjustedc | Age-adjustedb | MV-adjustedc | |
| Beta-blockersa | |||||||
| Duration of use (years) | |||||||
| 1–5 | 185 | 1.03 (0.88–1.21) | 1.04 (0.88–1.22) | 1.07 (0.88–1.31) | 1.17 (0.95–1.42) | 0.99 (0.83–1.18) | 1.00 (0.84–1.20) |
| 6–10 | 101 | 1.17 (0.95–1.45) | 1.21 (0.98–1.50) | 1.22 (0.96–1.56) | 1.36 (1.07–1.74) | 1.12 (0.90–1.40) | 1.17 (0.94–1.46) |
| ≥11 | 32 | 1.19 (0.83–1.71) | 1.29 (0.90–1.85) | 1.24 (0.85–1.81) | 1.44 (0.99–2.12) | 1.14 (0.79–1.65) | 1.24 (0.86–1.79) |
| CCBsa | |||||||
| Duration of use (years) | |||||||
| 1–5 | 167 | 1.10 (0.94–1.30) | 1.11 (0.94–1.31) | 1.15 (0.94–1.41) | 1.25 (1.01–1.53) | 1.04 (0.87–1.24) | 1.04 (0.87–1.25) |
| 6–10 | 66 | 0.95 (0.74–1.23) | 0.98 (0.76–1.27) | 1.00 (0.75–1.32) | 1.10 (0.83–1.46) | 0.90 (0.69–1.17) | 0.92 (0.71–1.20) |
| ≥11 | 18 | 0.95 (0.59–1.53) | 0.98 (0.61–1.58) | 0.99 (0.60–1.61) | 1.10 (0.67–1.80) | 0.89 (0.55–1.44) | 0.92 (0.57–1.48) |
| Thiazide diureticsa | |||||||
| Duration of use (years) | |||||||
| 1–5 | 179 | 1.08 (0.92–1.28) | 1.09 (0.93–1.29) | 1.13 (0.92–1.38) | 1.23 (1.00–1.51) | 1.03 (0.86–1.23) | 1.04 (0.87–1.24) |
| 6–10 | 71 | 1.08 (0.84–1.39) | 1.11 (0.87–1.43) | 1.13 (0.86–1.49) | 1.25 (0.95–1.65) | 1.03 (0.80–1.33) | 1.06 (0.82–1.36) |
| ≥11 | 12 | 0.89 (0.50–1.59) | 0.92 (0.52–1.65) | 0.93 (0.52–1.68) | 1.04 (0.57–1.87) | 0.85 (0.47–1.52) | 0.88 (0.49–1.57) |
| ACEIsa | |||||||
| Duration of use (years) | |||||||
| 1–5 | 117 | 0.96 (0.79–1.18) | 0.97 (0.79–1.19) | 0.99 (0.77–1.28) | 1.07 (0.83–1.38) | 0.89 (0.72–1.10) | 0.90 (0.73–1.12) |
| 6–10 | 40 | 0.96 (0.69–1.33) | 0.96 (0.69–1.33) | 0.99 (0.69–1.42) | 1.06 (0.74–1.52) | 0.89 (0.64–1.24) | 0.89 (0.64–1.24) |
| ≥11 | 5 | 0.58 (0.24–1.41) | 0.61 (0.25–1.48) | 0.60 (0.24–1.47) | 0.67 (0.27–1.65) | 0.54 (0.22–1.31) | 0.56 (0.23–1.38) |
| Furosemidea | |||||||
| Duration of use (years) | |||||||
| 1– 5 | 80 | 1.43 (1.13–1.81) | 1.34 (1.05–1.70) | 1.39 (1.06–1.81) | 1.41 (1.08–1.85) | 1.38 (1.09–1.75) | 1.27 (1.00–1.61) |
| ≥6 | 19 | 1.25 (0.79–1.98) | 1.13 (0.71–1.80) | 1.21 (0.75–1.95) | 1.19 (0.74–1.93) | 1.21 (0.76–1.92) | 1.07 (0.67–1.71) |
CCB calcium-channel blocker, ACEI angiotensin-converting enzyme inhibitor, MV multivariate, AHEI Alternate Healthy Eating Index.
aIndividual classes of antihypertensive drugs have different timelines in analyses based on new-user design: beta-blockers (NHS: 1990–2014; HPFS: 1988–2014), CCBs (NHS: 1990–2014; HPFS: 1988–2014), thiazide diuretics (NHS: 1990–2014; HPFS: 1988–2014), ACEIs (NHS: 1990–2014; HPFS: 2006–2014), furosemide (NHS: 1996–2014; HPFS: 1988–2014), other antihypertensive drugs (NHS: 1990–2014; HPFS: 1988–2014).
bStratified by age, sex and follow-up cycle.
cStratified by age, sex and follow-up cycle; adjusted for race, pack-years of smoking, physical activity, BMI, alcohol consumption, AHEI, regular use of aspirin, family history of colorectal cancer, history of diabetes mellitus, multivitamin use, screening colonoscopy or sigmoidoscopy, total calorie intake, red or processed meat intake, fibre intake, folate intake, calcium intake, vitamin D intake.
Moreover, compared to normotensive individuals, no association was observed for either those with hypertension and ever used any antihypertensive medication (new-user design: 1.07, 0.96–1.18; prevalent-user design: 1.06, 0.98–1.15) or those with hypertension but never used any antihypertensive medication (new-user design: 0.89, 0.77–1.03; prevalent-user design: 0.92, 0.81–1.05) (Supplementary Tables 5 and 6).
Hypertension and CRC risk
Hypertension was not associated with CRC risk (1.00, 0.87–1.14 for ≥11 years; 1.01, 0.97–1.05 for per 5-year increment; P = 0.63 for trend), and the results remained similar in the analysis according to subsites (Table 5 and Supplementary Table 7).
Table 5.
Colorectal cancer incidence according to the total duration of hypertension.
| Total duration of hypertension (years) | Per 5-year increment HR (95% CI) | P trendc | ||||
|---|---|---|---|---|---|---|
| 0 | 1–5 | 6–10 | ≥11 | |||
| No. of cases | 2692 | 292 | 245 | 272 | ||
| Age-adjusteda | 1 [Ref] | 0.93 (0.82–1.05) | 0.96 (0.84–1.10) | 1.00 (0.88–1.14) | 1.01 (0.97–1.05) | 0.61 |
| MVb | 1 | 0.94 (0.83–1.06) | 0.97 (0.85–1.11) | 1.00 (0.87–1.14) | 1.01 (0.97–1.05) | 0.63 |
MV multivariate, HR hazard ratio, CI confidence interval, AHEI Alternate Healthy Eating Index.
aStratified by age, sex and follow-up cycle.
bStratified by age, sex and follow-up cycle; adjusted for race, pack-years of smoking, physical activity, BMI, alcohol consumption, AHEI, regular use of aspirin, family history of colorectal cancer, history of diabetes mellitus, multivitamin use, screening colonoscopy or sigmoidoscopy, total calorie intake, red or processed meat intake, fibre intake, folate intake, calcium intake and vitamin D intake.
cP value for trend was the P value for variables modelled as continuous.
Joint association of hypertension and antihypertensive medication use with CRC risk
In joint association analysis of hypertension and antihypertensive medication use based on both new-user design and prevalent-user design, compared with individuals who never had hypertension and never used antihypertensive medications, no association was observed for all collapsed categories. There was no interaction between hypertension and antihypertensive medication use on CRC incidence (new-user design: P = 0.69 for interaction; prevalent-user design: P = 0.39 for interaction) (Table 6).
Table 6.
Joint associations of hypertension and antihypertensive medication use with colorectal cancer incidence (new-user design and prevalent-user design).
| Joint associations according to hypertension and antihypertensive medication usec | |||||
|---|---|---|---|---|---|
| Never hypertension and never-use antihypertensive medication | Never hypertension and ever-use antihypertensive medication | Ever hypertension and never-use antihypertensive medication | Ever hypertension and ever-use antihypertensive medication | P for interactiond | |
| New-user design | |||||
| No. of cases | 1273 | 158 | 218 | 694 | |
| Age-adjusteda | 1 [Ref] | 1.09 (0.92–1.29) | 0.97 (0.84–1.12) | 1.07 (0.97–1.19) | 0.89 |
| MVb | 1 | 1.15 (0.97–1.36) | 0.90 (0.78–1.05) | 1.09 (0.98–1.21) | 0.69 |
| Prevalent-user design | |||||
| No. of cases | 1422 | 257 | 273 | 1376 | |
| Age-adjusteda | 1 | 1.03 (0.90–1.18) | 0.99 (0.87–1.13) | 1.09 (1.01–1.18) | 0.49 |
| MVb | 1 | 1.06 (0.93–1.22) | 0.93 (0.81–1.06) | 1.07 (0.99–1.17) | 0.39 |
MV multivariate, AHEI Alternate Healthy Eating Index.
aStratified by age, sex and follow-up cycle.
bStratified by age, sex and follow-up cycle; adjusted for race, pack-years of smoking, physical activity, BMI, alcohol consumption, AHEI, regular use of aspirin, family history of colorectal cancer, history of diabetes mellitus, screening colonoscopy or sigmoidoscopy, multivitamin use, total calorie intake, red or processed meat intake, fibre intake, folate intake, calcium intake and vitamin D intake.
cHistory of hypertension was dichotomised into “never hypertension” category versus “ever hypertension” category; status of antihypertensive medication use was dichotomised into “never use antihypertensive medication” category versus “ever use antihypertensive medication” category, resulting in 4 categories (never hypertension and never-use antihypertensive medication, never hypertension and ever use antihypertensive medication, ever hypertension and never-use antihypertensive medication, and ever hypertension and ever use antihypertensive medication).
dP value for interaction was calculated using the multiplication of each low/high category of the status of hypertension and antihypertensive medication use.
Antihypertensive medication use and CRC mortality
In analyses using new-user design, we detected no association between long-term use of beta-blockers (0.57, 0.28–1.16 for ≥11 years; 0.91, 0.78–1.07 for per 5-year increment; P = 0.25 for trend), CCBs (0.49, 0.20–1.19 for ≥11 years; 0.95, 0.80–1.13 for per 5-year increment; P = 0.57 for trend), thiazide diuretics (0.96, 0.42–2.20 for ≥11 years; 0.99, 0.81–1.21 for per 5-year increment; P = 0.94 for trend), ACEIs (0.62, 0.36–1.09 for ≥6 years; 0.81, 0.61–1.08 for per 5-year increment; P = 0.15 for trend) and CRC-specific mortality. Longer duration of furosemide use was associated with higher CRC-specific mortality (1.64, 1.00–2.68 for ≥6 years; 1.31, 1.01–1.70 for per 5-year increment; P = 0.046 for trend); however, after Bonferroni correction, it did not reach the statistically significant threshold (Supplementary Table 8). Essentially similar results were observed in analyses using a prevalent-user design (Supplementary Table 9).
Discussion
In this large prospective cohort study of US healthcare professionals, we comprehensively investigated the associations between hypertension, long-term use of major classes of antihypertensive medications, and CRC risk and mortality.
Hypertension remains the leading cause of premature mortality, the prevalence of which is increasing globally [39]. In the US, hypertension affects 30% of the population, and 59.4% of those diagnosed with high blood pressure used hypertensive medications [40]. On the other hand, CRC accounted for 10.0% of new cancer cases and 9.4% of cancer-related deaths worldwide [41], the disease burden of which is also substantial. Elucidating whether exposure to hypertension and long-term use of antihypertensive agents would have a secondary influence on the incidence and mortality of CRC has clinical importance and major public health implications [42, 43].
ACEIs and ARBs are among the major classes of antihypertensive agents most frequently investigated in relation to CRC risk and mortality [3, 9]. The hypothesis has been driven by evidence from mechanistic studies supporting the role of renin–angiotensin system (RAS) signalling pathway dysregulation in cancer inflammation and angiogenesis, and cancer cell proliferation, migration, metastasis, and apoptosis [22]. A recent meta-analyses reported a statistically significant association of RAS inhibitor use with modestly lower CRC risk (pooled relative risk, 95% CI was 0.86, 0.78–0.93) and lower CRC-specific mortality (0.80, 0.66–0.98). The protective association was observed for both ACEIs and ARBs, and was pronounced in both case–control and cohort studies [9]. These results are largely consistent with a subsequent well-designed large retrospective cohort investigation based on prescription and dispensing data from electronic healthcare database that observed ACEI/ARB use was associated with a lower risk of post-colonoscopy CRC in a duration–response manner [44], but inconsistent with two other large prospective cohort studies afterwards that observed no association [45, 46].
The hypothesis linking beta-blocker use with CRC is based on mechanistic studies supporting the role of weakening norepinephrine signalling in inhibiting CRC cell proliferation and migration [23, 24]. Previously the only case–control study that investigated beta-blocker use in relation to the risk of CRC showed no association [15]. With one exception [16], most of the prior investigations (all retrospective) do not support any association between use of beta-blockers and CRC-specific mortality among CRC patients as well [14, 17, 19].
Previous epidemiologic investigations generally offer no evidence supporting the association between CCB use and risk [7, 8] and mortality [8, 20] of CRC, except for a matched cohort study that reported a lower CRC risk among CCB users [21]. However, the hypothesis of which has been initially driven by mechanistic studies suggesting the complex bidirectional effects of CCBs on cancer cell apoptosis through regulating intracellular calcium ion balance [25]. The associations of thiazide diuretic, furosemide and CRC risk and mortality have scarcely been explored. The only existing investigation, a case–control study, reported diuretics as a composite was not associated with risk of CRC [7]. The associations between hypertension and the risk of CRC have been intensively examined. A meta-analysis, summarising 16 cohort studies and 9 case–control studies, supported a suspected link between hypertension and risk of CRC (an increased pooled relative risk of 1.15-fold, 95% CI 1.08–1.23) [1].
It is noteworthy that very few of the prior studies has been able to examine prospectively this topic with repeat assessments of exposures, rigorous confounding control, and adequate follow-up time, or disentangle the effects of antihypertensive medications and hypertension, limiting interpretation of their effect estimates. We were able to address the above-mentioned limitations in our study. In our analyses based on both new-user design and prevalent-user design, we detected no association between hypertension, long-term use of major classes of antihypertensive agents (beta-blockers, CCBs, thiazide diuretics and ACEIs) and CRC risk and mortality that reached the statistically significant threshold after Bonferroni correction. Interestingly, before Bonferroni correction, suggestive associations between beta-blocker use and CRC risk and between furosemide use and CRC-specific mortality were observed specifically in analyses using a new-user design.
In our study, the potential inverse associations between long-term ACEI use and risk of CRC, though all not statistically significant in analyses based on both new-user design and prevalent-user design, cannot be completely ruled out. On the other hand, we were not able to consider individual evaluation of ARBs because ARBs were queried relatively late in our cohorts. Further large prospective investigations on these RAS inhibitors are warranted.
We observed a suggestively higher risk of CRC-specific mortality associated with participants reported the longer duration of furosemide use, which did not reach the statistically significant threshold after Bonferroni correction and has not been reported before. Interestingly, similar result was not observed for overall CRC incidence. These results are hypothesis-generating and warrants re-evaluation in other investigations, though several initial clues suggested the potential role of sodium–potassium-chloride cotransporter and gene–calcium interactions [47, 48].
We observed no association between hypertension and colorectal cancer risk, which does not corroborate the positive evidence reported by prior meta-analysis [1]. However, it should be noted that the positive association observed in prior meta-analysis was modest, and was restricted to men only [1]. Further, the reported association seemed to be stronger in case–control studies, and in investigations of relatively low general quality, including those with unsatisfying confounding control (missing information on crucial covariates and missing repeat assessments of covariates are common) [1], partially explaining the inconsistent findings.
This study has several major strengths. First, it represents one of the largest prospective cohort studies so far to investigate this topic. Second, the long-term follow-up, high follow-up response rates attained, and repeat assessments of exposures enabled us to comprehensively assess the effects of hypertension and antihypertensive medications at a wide range of durations in the context of long-term exposure and “10-year delayed effect” for the first time, which has not been satisfactorily evaluated by prior investigations. Third, we were able to examine major classes of antihypertensive medications separately and disentangle the effects of antihypertensive agents and hypertension. Fourth, we were able to consider new-user design plus active-comparator design to minimise confounding by indication [34]. Fifth, time-varying information based on repeated measures of a wide spectrum of covariates throughout follow-up offers high-quality resources to rigorously control for potential confounding, further ensuring the accuracy of our effect estimates. Sixth, a standardised review of medical records and pathology reports by cohort investigators afforded us an opportunity to explore potential heterogeneity across anatomic subsites. Lastly, the nature of our cohort participants (all healthcare professionals) presented important advantages in minimising socioeconomic confounding, enhancing internal validity, and ensuring data quality.
The limitations of our study should be acknowledged. First, we were not able to conduct further analysis by investigating dose-dependent associations because information on these domains of exposure history was not assessed in our cohorts. Second, information on medication use and hypertension were self-reported. Nonetheless, we have confidence in the quality of medication data documented, considering a wide spectrum of lifestyle habits and medical history measurements (including hypertension [35]) in our cohorts have been demonstrated to be highly valid. Third, the possibility of residual confounding cannot be completely ruled out due to the observational study design. However, this possibility has been minimised given extensive control for all known or plausible confounding only had minimal influence on our effect estimates. On the other hand, large, randomised CRC primary prevention trials may not be feasible given the high prevalence of antihypertensive medication use in clinical settings, the long-term follow-up required in the context of “10-year delayed effect”, and the availability and efficacy of CRC screening practices (ethical concerns). Lastly, the homogeneity of cohort participants may limit the generalisability.
Conclusions
Hypertension and long-term use of major classes of antihypertensive medications are unlikely to be associated with CRC risk and mortality.
Supplementary information
Acknowledgements
The authors thank all participants and staff of the Nurses’ Health Study and Health Professionals Follow-up Study for their contributions to this research. The authors thank the Channing Division of Network Medicine, Brigham and Women’s Hospital and Harvard Medical School for leading the Nurses’ Health Study. The authors would like to acknowledge the contribution to this study from central cancer registries supported through the Centers for Disease Control and Prevention’s National Program of Cancer Registries (NPCR) and/or the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program. Central registries may also be supported by state agencies, universities, and cancer centers. Participating central cancer registries include the following: Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, Florida, Georgia, Hawaii, Idaho, Indiana, Iowa, Kentucky, Louisiana, Massachusetts, Maine, Maryland, Michigan, Mississippi, Montana, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Puerto Rico, Rhode Island, Seattle SEER Registry, South Carolina, Tennessee, Texas, Utah, Virginia, West Virginia, Wyoming. The authors assume full responsibility for the analyses and interpretation of these data.
Author contributions
Concept and design: YZ, WCW and ELG. Acquisition, analysis or interpretation of the data: YZ. Drafting of the manuscript: YZ. Critical revision of the manuscript for important intellectual content: ELG, WCW, JAM, ATC and MS. Statistical analysis: YZ. Obtained funding: YZ, MS, ATC and ELG. Administrative, technical or material support: ELG and WCW. Supervision: ELG. The authors assume full responsibility for the analyses and interpretation of these data.
Funding
Nurses’ Health Study and Health Professionals Follow-up Study was supported by grants UM1 CA186107, U01 CA176726, P01 CA87969, and U01 CA167552 from the National Institutes of Health (NIH). This work was additionally supported by the American Cancer Society Mentored Research Scholar Grant MRSG-17–220–01-NEC and the NIH grant R00 CA215314 (Dr. Song); NIH grants R01 CA137178, R35 CA253185 (Dr. Chan); and a grant from the World Cancer Research Fund (Dr. Giovannucci). Dr. Zhang is supported by Irene M. & Fredrick J. Stare Nutrition Education Fund Doctoral Scholarship and Mayer Fund Doctoral Scholarship. Dr. Chan is a Stuart and Suzanne Steele MGH Research Scholar. The funding sources played no role in the study design, data collection, data analysis and interpretation of the results, or the decisions made in the preparation and submission of the article.
Data availability
Data described in the manuscript may be made available upon application to and approval by the Channing Division of Network Medicine at Brigham and Women’s Hospital, and Harvard T.H. Chan School of Public Health. Further information, including the procedures to obtain and access data from the Nurses’ Health Study and Health Professionals Follow-up Study, is described at https://www.nurseshealthstudy.org/researchers (email: nhsaccess@channing.harvard.edu) and https://sites.sph.harvard.edu/hpfs/for-collaborators/.
Competing interests
Dr. Meyerhardt declares institutional research funding from Boston Biomedical and consulting for Ignyta, Taiho Pharmaceutical, and Cota, outside the submitted work. Dr. Chan declares research funding from Bayer and consulting for Bayer, Pfizer Inc. and Boehringer Ingelheim, outside the submitted work. The remaining authors declare no competing interests.
Ethics approval and consent to participate
The Nurses’ Health Study and Health Professionals Follow-up Study were approved by the Institutional Review Boards of the Brigham and Women’s Hospital and the Harvard T.H. Chan School of Public Health (Boston, MA), and those of participating registries as required. Written informed consent was required to retrieve medical records.
Consent to publish
Not applicable.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1038/s41416-022-01975-4.
References
- 1.Xuan K, Zhao T, Sun C, Patel AS, Liu H, Chen X, et al. The association between hypertension and colorectal cancer: a meta-analysis of observational studies. Eur J Cancer Prev. 2021;30:84–96. doi: 10.1097/CEJ.0000000000000578. [DOI] [PubMed] [Google Scholar]
- 2.Makar GA, Holmes JH, Yang YX. Angiotensin-converting enzyme inhibitor therapy and colorectal cancer risk. J Natl Cancer Inst. 2014;106:djt374. doi: 10.1093/jnci/djt374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dai YN, Wang JH, Zhu JZ, Lin JQ, Yu CH, Li YM. Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers therapy and colorectal cancer: a systematic review and meta-analysis. Cancer Causes Control. 2015;26:1245–55. doi: 10.1007/s10552-015-0617-1. [DOI] [PubMed] [Google Scholar]
- 4.Hicks BM, Filion KB, Yin H, Sakr L, Udell JA, Azoulay L. Angiotensin converting enzyme inhibitors and risk of lung cancer: population based cohort study. BMJ. 2018;363:k4209. doi: 10.1136/bmj.k4209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cronin-Fenton D. Angiotensin converting enzyme inhibitors and lung cancer. BMJ. 2018;363:k4337. doi: 10.1136/bmj.k4337. [DOI] [PubMed] [Google Scholar]
- 6.Lever AF, Hole DJ, Gillis CR, McCallum IR, McInnes GT, MacKinnon PL, et al. Do inhibitors of angiotensin-I-converting enzyme protect against risk of cancer? Lancet. 1998;352:179–84. doi: 10.1016/S0140-6736(98)03228-0. [DOI] [PubMed] [Google Scholar]
- 7.Boudreau DM, Koehler E, Rulyak SJ, Haneuse S, Harrison R, Mandelson MT. Cardiovascular medication use and risk for colorectal cancer. Cancer Epidemiol Biomark Prev. 2008;17:3076–80. doi: 10.1158/1055-9965.EPI-08-0095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Michels KB, Rosner BA, Walker AM, Stampfer MJ, Manson JE, Colditz GA, et al. Calcium channel blockers, cancer incidence, and cancer mortality in a cohort of U.S. women: the nurses’ health study. Cancer. 1998;83:2003–7. doi: 10.1002/(SICI)1097-0142(19981101)83:9<2003::AID-CNCR17>3.0.CO;2-3. [DOI] [PubMed] [Google Scholar]
- 9.Chen X, Yi CH, Ya KG. Renin-angiotensin system inhibitor use and colorectal cancer risk and mortality: a dose-response meta analysis. J Renin Angiotensin Aldosterone Syst. 2020;21:1470320319895646. doi: 10.1177/1470320319895646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Engineer DR, Burney BO, Hayes TG, Garcia JM. Exposure to ACEI/ARB and beta-blockers is associated with improved survival and decreased tumor progression and hospitalizations in patients with advanced colon cancer. Transl Oncol. 2013;6:539–45. doi: 10.1593/tlo.13346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bangalore S, Kumar S, Kjeldsen SE, Makani H, Grossman E, Wetterslev J, et al. Antihypertensive drugs and risk of cancer: network meta-analyses and trial sequential analyses of 324,168 participants from randomised trials. Lancet Oncol. 2011;12:65–82. doi: 10.1016/S1470-2045(10)70260-6. [DOI] [PubMed] [Google Scholar]
- 12.Bhaskaran K, Douglas I, Evans S, van Staa T, Smeeth L. Angiotensin receptor blockers and risk of cancer: cohort study among people receiving antihypertensive drugs in UK General Practice Research Database. BMJ. 2012;344:e2697. doi: 10.1136/bmj.e2697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sipahi I, Debanne SM, Rowland DY, Simon DI, Fang JC. Angiotensin-receptor blockade and risk of cancer: meta-analysis of randomised controlled trials. Lancet Oncol. 2010;11:627–36. doi: 10.1016/S1470-2045(10)70106-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jansen L, Weberpals J, Kuiper JG, Vissers PAJ, Wolkewitz M, Hoffmeister M, et al. Pre- and post-diagnostic beta-blocker use and prognosis after colorectal cancer: results from a population-based study. Int J Cancer. 2017;141:62–71. doi: 10.1002/ijc.30717. [DOI] [PubMed] [Google Scholar]
- 15.Jansen L, Below J, Chang-Claude J, Brenner H, Hoffmeister M. Beta blocker use and colorectal cancer risk: population-based case-control study. Cancer. 2012;118:3911–9. doi: 10.1002/cncr.26727. [DOI] [PubMed] [Google Scholar]
- 16.Ahl R, Matthiessen P, Sjolin G, Cao Y, Wallin G, Ljungqvist O, et al. Effects of beta-blocker therapy on mortality after elective colon cancer surgery: a Swedish nationwide cohort study. BMJ Open. 2020;10:e036164. doi: 10.1136/bmjopen-2019-036164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Musselman RP, Bennett S, Li W, Mamdani M, Gomes T, van Walraven C, et al. Association between perioperative beta blocker use and cancer survival following surgical resection. Eur J Surg Oncol. 2018;44:1164–9. doi: 10.1016/j.ejso.2018.05.012. [DOI] [PubMed] [Google Scholar]
- 18.Ahl R, Matthiessen P, Cao Y, Sjolin G, Ljungqvist O, Mohseni S. The relationship between severe complications, beta-blocker therapy and long-term survival following emergency surgery for colon cancer. World J Surg. 2019;43:2527–35.. doi: 10.1007/s00268-019-05058-z. [DOI] [PubMed] [Google Scholar]
- 19.Hicks BM, Murray LJ, Powe DG, Hughes CM, Cardwell CR. beta-Blocker usage and colorectal cancer mortality: a nested case-control study in the UK Clinical Practice Research Datalink cohort. Ann Oncol. 2013;24:3100–6. doi: 10.1093/annonc/mdt381. [DOI] [PubMed] [Google Scholar]
- 20.Cui Y, Wen W, Zheng T, Li H, Gao YT, Cai H, et al. Use of Antihypertensive medications and survival rates for breast, colorectal, lung, or stomach cancer. Am J Epidemiol. 2019;188:1512–28.. doi: 10.1093/aje/kwz106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Grimaldi-Bensouda L, Klungel O, Kurz X, de Groot MC, Maciel Afonso AS, de Bruin ML, et al. Calcium channel blockers and cancer: a risk analysis using the UK Clinical Practice Research Datalink (CPRD) BMJ Open. 2016;6:e009147. doi: 10.1136/bmjopen-2015-009147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.George AJ, Thomas WG, Hannan RD. The renin-angiotensin system and cancer: old dog, new tricks. Nat Rev Cancer. 2010;10:745–59. doi: 10.1038/nrc2945. [DOI] [PubMed] [Google Scholar]
- 23.Mravec B, Horvathova L, Hunakova L. Neurobiology of cancer: the role of beta-adrenergic receptor signaling in various tumor environments. Int J Mol Sci. 2020;21:7958. [DOI] [PMC free article] [PubMed]
- 24.Chin CC, Li JM, Lee KF, Huang YC, Wang KC, Lai HC, et al. Selective beta2-AR blockage suppresses colorectal cancer growth through regulation of EGFR-Akt/ERK1/2 signaling, G1-phase arrest, and apoptosis. J Cell Physiol. 2016;231:459–72. doi: 10.1002/jcp.25092. [DOI] [PubMed] [Google Scholar]
- 25.Mason RP. Calcium channel blockers, apoptosis and cancer: is there a biologic relationship? J Am Coll Cardiol. 1999;34:1857–66. doi: 10.1016/S0735-1097(99)00447-7. [DOI] [PubMed] [Google Scholar]
- 26.Janne PA, Mayer RJ. Chemoprevention of colorectal cancer. N Engl J Med. 2000;342:1960–8. doi: 10.1056/NEJM200006293422606. [DOI] [PubMed] [Google Scholar]
- 27.Giovannucci E. Aspirin and delayed chemoprevention of colorectal cancer. Clin Chem. 2018;64:1668–9. doi: 10.1373/clinchem.2018.290809. [DOI] [PubMed] [Google Scholar]
- 28.Cuzick J, Otto F, Baron JA, Brown PH, Burn J, Greenwald P, et al. Aspirin and non-steroidal anti-inflammatory drugs for cancer prevention: an international consensus statement. Lancet Oncol. 2009;10:501–7. doi: 10.1016/S1470-2045(09)70035-X. [DOI] [PubMed] [Google Scholar]
- 29.Zhang Y, Chan AT, Meyerhardt JA, Giovannucci EL. Timing of aspirin use in colorectal cancer chemoprevention: a prospective cohort study. J Natl Cancer Inst. 2021;113:841–51.. doi: 10.1093/jnci/djab009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Zhang Y, Wu K, Chan AT, Meyerhardt JA, Giovannucci EL. Long-term statin use, total cholesterol level, and risk of colorectal cancer: a prospective cohort study. Am J Gastroenterol. 2022;117:158–66.. doi: 10.14309/ajg.0000000000001543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Bibbins-Domingo K. Force USPST. Aspirin use for the primary prevention of cardiovascular disease and colorectal cancer: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med. 2016;164:836–45. doi: 10.7326/M16-0577. [DOI] [PubMed] [Google Scholar]
- 32.Belanger CF, Hennekens CH, Rosner B, Speizer FE. The nurses’ health study. Am J Nurs. 1978;78:1039–40. [PubMed] [Google Scholar]
- 33.Harvard School of Public Health. Health professionals follow-up study. https://sites.sph.harvard.edu/hpfs/ (Accessed September 10, 2022).
- 34.Yoshida K, Solomon DH, Kim SC. Active-comparator design and new-user design in observational studies. Nat Rev Rheumatol. 2015;11:437–41. doi: 10.1038/nrrheum.2015.30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Colditz GA, Martin P, Stampfer MJ, Willett WC, Sampson L, Rosner B, et al. Validation of questionnaire information on risk factors and disease outcomes in a prospective cohort study of women. Am J Epidemiol. 1986;123:894–900. doi: 10.1093/oxfordjournals.aje.a114319. [DOI] [PubMed] [Google Scholar]
- 36.Lee DH, Keum N, Giovannucci EL. Colorectal cancer epidemiology in the Nurses’ Health Study. Am J Public Health. 2016;106:1599–607. doi: 10.2105/AJPH.2016.303320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Keum N, Giovannucci E. Global burden of colorectal cancer: emerging trends, risk factors and prevention strategies. Nat Rev Gastroenterol Hepatol. 2019;16:713–32.. doi: 10.1038/s41575-019-0189-8. [DOI] [PubMed] [Google Scholar]
- 38.Stampfer MJ, Willett WC, Speizer FE, Dysert DC, Lipnick R, Rosner B, et al. Test of the national death index. Am J Epidemiol. 1984;119:837–9. doi: 10.1093/oxfordjournals.aje.a113804. [DOI] [PubMed] [Google Scholar]
- 39.Mills KT, Stefanescu A, He J. The global epidemiology of hypertension. Nat Rev Nephrol. 2020;16:223–37.. doi: 10.1038/s41581-019-0244-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Samanic CM, Barbour KE, Liu Y, Fang J, Lu H, Schieb L, et al. Prevalence of self-reported hypertension and antihypertensive medication use among adults—United States, 2017. MMWR Morb Mortal Wkly Rep. 2020;69:393–8. doi: 10.15585/mmwr.mm6914a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209–49. [DOI] [PubMed]
- 42.Song M, Giovannucci EL. Antihypertension and colorectal cancer prevention: getting two birds with one stone. J Natl Cancer Inst. 2014;106:djt438. doi: 10.1093/jnci/djt438. [DOI] [PubMed] [Google Scholar]
- 43.Nissen SE. Angiotensin-receptor blockers and cancer: urgent regulatory review needed. Lancet Oncol. 2010;11:605–6. doi: 10.1016/S1470-2045(10)70142-X. [DOI] [PubMed] [Google Scholar]
- 44.Cheung KS, Chan EW, Seto WK, Wong ICK, Leung WK. ACE (angiotensin-converting enzyme) inhibitors/angiotensin receptor blockers are associated with lower colorectal cancer risk: a territory-wide study with propensity score analysis. Hypertension. 2020;76:968–75.. doi: 10.1161/HYPERTENSIONAHA.120.15317. [DOI] [PubMed] [Google Scholar]
- 45.Brasky TM, Flores KF, Larson JC, Newton AM, Shadyab AH, Watanabe JH, et al. Associations of angiotensin-converting enzyme inhibitor or angiotensin receptor blocker use with colorectal cancer risk in the Women’s Health Initiative. Cancer Epidemiol Biomarkers Prev. 2021;30:1029–32. [DOI] [PMC free article] [PubMed]
- 46.Htoo PT, Sturmer T, Jonsson-Funk M, Pate V, Simpson RJ, Jr., Lund JL. Renin-angiotensin-aldosterone system-based antihypertensive agents and the risk of colorectal cancer among medicare beneficiaries. Epidemiology. 2019;30:867–75.. doi: 10.1097/EDE.0000000000001065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Zhu X, Liang J, Shrubsole MJ, Ness RM, Cai Q, Long J, et al. Calcium intake and ion transporter genetic polymorphisms interact in human colorectal neoplasia risk in a 2-phase study. J Nutr. 2014;144:1734–41. doi: 10.3945/jn.114.196709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Zhao J, Zhu X, Shrubsole MJ, Ness RM, Hibler EA, Cai Q, et al. Interactions between calcium intake and polymorphisms in genes essential for calcium reabsorption and risk of colorectal neoplasia in a two-phase study. Mol Carcinog. 2017;56:2258–66.. doi: 10.1002/mc.22678. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data described in the manuscript may be made available upon application to and approval by the Channing Division of Network Medicine at Brigham and Women’s Hospital, and Harvard T.H. Chan School of Public Health. Further information, including the procedures to obtain and access data from the Nurses’ Health Study and Health Professionals Follow-up Study, is described at https://www.nurseshealthstudy.org/researchers (email: nhsaccess@channing.harvard.edu) and https://sites.sph.harvard.edu/hpfs/for-collaborators/.