Regular, long-duration multivitamin use and risk of overall and aggressive prostate cancer in the Health Professionals Follow-up Study

Abstract

Introduction:

Multivitamin supplement is commonly used among older adults in the United States. Evidence on multivitamin use and prostate cancer risk is limited, and some suggested potential risk for clinically important subtypes of cancer.

Methods:

A total of 48,137 men from the Health Professionals Follow-up Study were followed from 1986 to 2017. Multivitamin use and frequency were self-reported at baseline and updated biennially. Clinical features of prostate cancer included advanced, lethal, and high-grade outcomes. Cox proportional hazards models were used to estimate multivariable hazard ratios (HRs) and 95% confidence intervals (CIs) of multivitamin use and incidence of prostate cancer.

Results:

The median follow-up for men without prostate cancer diagnosis or death was 30.7 years. We documented 7,108 incident prostate cancer cases including 564 advanced and 1,065 lethal. Overall, we observed a null association between multivitamin use and risk of prostate cancer. Compared to never-users, men who used multivitamin 10 or more tablets per week had similar risk of advanced (HR:1.14, 95% CI: 0.77–1.70, P_trend=0.46) or lethal (HR:1.07, 95% CI: 0.80–1.44, P_trend=0.59) prostate cancer, and multivitamin use of 15 years or more was not associated with risk of advanced (HR:1.10, 95% CI: 0.80–1.50, P_trend=0.46) or lethal (HR:1.04, 95% CI: 0.83–1.31, P_trend=0.13) prostate cancer.

Conclusions:

Regular or long-duration multivitamin use among older generally well-nourished men was not associated with either increased or lower risk of overall or aggressive prostate cancer.

Introduction

Multivitamins are the most widely used dietary supplements among older men in the US^1,2. According to the National Health and Nutrition Examination Survey (NHANES) data in 2011–2014, 34–49% of older adults regularly took multivitamin/multimineral supplements³. The most common reasons reported by consumers of multivitamins were to maintain or improve overall health, prevent health problems, and promote bone or heart health^3,4. However, current guidelines^5,6 concluded that the present evidence is insufficient to recommend either for or against the use of multivitamins to prevent chronic diseases.

Evidence on the association between multivitamin use and prostate cancer is limited and inconsistent. The large randomized clinical trial (RCT) Physicians’ Health Study II (PHS II)⁷ and observational studies^8–10 reported null associations between multivitamin use and overall prostate cancer risk. However, some evidence suggests regular or long-term multivitamin use was associated with increased risk of advanced prostate cancer, characterized by disease progression and poorer survival^11–13.

The potential harmful effect of regular long-term use of multivitamin on advanced prostate cancer warrants further examination. RCTs would be the ideal design to examine the association, but they usually do not have long-term multivitamin use of more than 10 years and have small number of aggressive prostate cancer cases^7,14. Here, leveraging data from a large prospective cohort with 30 years of follow up with a large number of prostate cancer cases and detailed repeated assessments of multivitamin use, we aimed to investigate the association between multivitamin use frequency and duration in relation to subtype specific prostate cancer.

Methods

Study population

The Health Professionals Follow-up Study (HPFS) is an ongoing prospective cohort study of 51,529 US male health professionals who enrolled in 1986 at age of 40 to 75. Participants were mailed questionnaires at baseline and biennially to collect updated information on demographic, lifestyle factors, medical history, medication, and disease outcome. Dietary information was collected using a validated semi-quantitative food frequency questionnaire (FFQ) at baseline and every 4 years thereafter^15,16. Follow up rates exceeded 90% for each questionnaire. The study protocol was approved by the institutional review boards of the Brigham and Women’s Hospital and Harvard T.H. Chan School of Public Health, and those of participating registries as required.

Assessment of multivitamin

At baseline, participants were asked ‘have you regularly taken multivitamins?’ and identify themselves as ‘never-user’, ‘past only’ or ‘current user’. Current users further provided number of tablets per week (≤2, 3–5, 6–9, ≥10 per week), duration of use (0–1, 2–4, 5–9, ≥10 years), and the usual brand and type of multivitamin. Subsequent biennial questionnaires collected updated information on current multivitamin use and frequency at each follow-up cycle.

Men who reported current multivitamin use but missing frequency (<7.0% in each questionnaire cycle) were assigned the most common frequency level (6–9 times per week). Never-users were defined as participants who never reported multivitamin use on current or any previous questionnaires. The duration of multivitamin use was calculated based on reported duration at baseline and updated with subsequent response to current use status. Men who reported multivitamin current use but missing duration at baseline (9.3%) were assigned the most common duration category (≥10 years). Participants also provided other dietary supplements use every two years since 1986¹⁷. The validity of supplement use by these health professionals was confirmed with validation studies based on dietary records¹⁵.

Ascertainment of prostate cancer cases and death

Diagnoses of prostate cancer were initially self-reported on biennial questionnaires by participants and then confirmed by review of medical records and pathology reports. Clinical and pathologic stage, Gleason score, and prostate cancer progression after diagnosis were collected by medical records and questionnaires sent to prostate cancer survivors and their attending physicians. Deaths in the cohort were ascertained through reports by family members and searches of National Death Index. Underlying cause of death were determined by study physicians based on all available data and were blinded to any exposure information. The mortality ascertainment in the cohort is more than 98%¹⁸.

For overall prostate cancer endpoint, stage T1a prostate cancer were excluded since these cases are incidentally diagnosed and prone to detection bias. We classified clinical subtypes of prostate cancer as advanced (stage T3b, T4, N1, or M1), lethal (distant metastases or prostate cancer specific death), and high-grade (Gleason 8–10) using information from prostatectomy or biopsy pathology reports.

Population for analysis

At baseline, we excluded participants who had a history of cancer (other than nonmelanoma skin cancer), who did not have multivitamin use information, and those who left more than 70 items blank on FFQ or reported implausible energy intake (<800 or >4200 kcal/day), leaving 48,137 men for current analysis.

Statistical analyses

Cox proportional hazards models were applied to estimate the hazard ratios (HR’s) and 95% confidence intervals (CI’s) for the association between multivitamin supplement use and risk of prostate cancer. Person-time was calculated from the return of the baseline questionnaire until date of diagnosis, death, or end of the follow-up (January 2017), whichever came first. For example, for the analysis of high-grade prostate cancer, people who developed other subtypes of prostate cancer at diagnosis were censored at the time they had a diagnosis because they were not longer at risk of developing high-grade prostate cancer at diagnosis by definition. Age in months and calendar year at start of follow-up for each 2-year questionnaire cycle were used as stratification variables in the model. Our basic model included race, family history of prostate cancer, PSA testing, history of diabetes, aspirin use, body mass index (BMI) at age 21, current BMI, height, smoking, vigorous physical activity, and intake of total calories, red meat, tomato-based foods and fish. To control potential confounding from other supplements¹⁹, we additionally adjusted for zinc, selenium, vitamin E and A supplement use and number of overall supplement use. Linear trends across categories were evaluated using the median of each category as a continuous variable.

We conducted stratified analysis by risk factors of advanced prostate cancer²⁰, including family history, BMI, smoking status, and vigorous physical activity. We also evaluated the joint effect of multivitamin and other dietary supplements, including zinc, selenium, vitamin E and A. To evaluate potential latency and minimize possible reverse causation, we used multivitamin reported at different latencies (i.e., 4–6 and 8–10 years) before prostate cancer diagnosis. For example, in the 8–10 years lag analysis, we examined the association of multivitamin use in 1990 with risk of prostate cancer between 1998 and 2000. To address residual confounding from PSA screening, we further examined the association specifically in post-PSA era (1994–2016). We also conducted a complete case analysis by restricting to men with complete multivitamin use, frequency, and duration information as a sensitivity analysis.

All analyses were performed using SAS version 9.4 (SAS Institute, Inc; Cary, NC) and results with a two-sided p-value <0.05 were considered statistically significant.

Results

The median follow-up for men without prostate cancer diagnosis or death was 30.7 years. We documented a total of 7,108 incident prostate cancer cases including 564 advanced, 932 high-grade and 1,065 lethal. At baseline, multivitamin users of ≥10 tablets per week were more likely to be physical active, use aspirin and other supplements (Table 1). Differences in PSA screening patterns were minor.

Table 1.

Selected characteristics of 48,137 participants in the Health Professionals Follow-up Study by frequency of multivitamin use in 1986^*.

Characteristic	Level of multivitamin dosage, tablets/week
	Past use	Never use	1–2 tablets/week	3–5 tablets/week	6–9 tablets/week	≥10 tablets/week
	n=9,148	n=18,166	n=1,648	n=3,644	n=12,812	n=2,719
Age in years, mean (SD)	52.8 (9.6)	54.4 (9.7)	52.5 (9.5)	53.4 (9.5)	56.0 (9.8)	54.9 (9.4)
White, %	96	96	95	95	96	96
Family history of prostate cancer, %	11.9	11.9	11.3	12.1	12.1	12.2
History of prostatitis or prostatic infection in 1992, %	13.9	11.6	13.9	13.9	13.3	13.3
PSA test in 1994, %	36.5	35.7	36.4	36.9	39.2	37.8
PSA test in 2000, %	68.6	68.1	68.0	69.3	70.0	69.9
Ever had negative biopsy in 1994, %	9.8	9.4	8.4	9.3	10.0	9.7
History of diabetes, %	3.3	3.1	2.9	2.0	3.6	2.9
Current aspirin use, %	28.9	23.8	31.6	32.9	35.7	32.1
BMI at age 21 in kg/m2, mean (SD)	22.9 (3.0)	23.1 (3.0)	23.0 (2.9)	22.9 (3.0)	23.0 (2.9)	23.1 (2.9)
Current BMI in in kg/m2, mean (SD)	25.6 (3.4)	25.8 (3.4)	25.5 (3.2)	25.4 (3.4)	25.2 (3.2)	25.2 (3.3)
Height in inches, mean (SD)	70.1 (2.9)	70.1 (2.9)	69.9 (2.8)	70.1 (2.9)	70.2 (2.8)	70.1 (2.9)
Smoking status
Never smoker, %	43.5	45.5	46.4	45.0	43.5	46.3
Past smoker quit >10 years, %	31.4	27.7	27.9	27.5	31.0	30.7
Past smoker quit <=10 years, %	12.5	12.5	12.0	13.4	12.5	10.7
Current smoker, %	8.7	10.3	10.7	10.3	9.4	7.9
Smoking unknown, %	3.9	4.1	3.1	3.8	3.6	4.4
Vigorous activity in MET-h/week, mean (SD) †	5.8 (18.0)	5.8 (20.1)	5.4 (17.8)	5.8 (15.2)	6.9 (21.9)	7.8 (22.2)
Energy in kcal/d, mean (SD)	1976 (616)	1986 (612)	2043 (627)	1995 (621)	1979 (601)	1963 (617)
Red meat in servings/week, mean (SD)	6.8 (5.1)	7.3 (5.2)	7.2 (5.1)	6.8 (4.7)	6.4 (4.9)	5.6 (4.8)
Tomato-based foods in servings/week, mean (SD)	4.6 (3.5)	4.7 (3.6)	4.7 (3.6)	4.6 (3.4)	4.7 (3.6)	4.9 (3.8)
Fish in servings/week, mean (SD)	2.3 (2.0)	2.3 (2.2)	2.3 (2.0)	2.3 (1.9)	2.5 (2.2)	2.9 (2.6)
Current zinc supplement use, %	3.7	1.5	8.6	12.0	15.0	38.4
Current selenium supplement use, %	1.5	0.7	3.3	5.1	8.2	28.2
Current vitamin A supplement use, %	2.3	0.9	5.3	8.5	10.7	34.8
Current vitamin E supplement use, %	6.9	3.2	13.3	22.7	29.0	57.0

^*All values (except age) are standardized to the age distribution of the study population.

^†METs=Metabolic Equivalents per week.

Overall, we observed no association between multivitamin use and risk of prostate cancer overall or its clinical subtypes. Compared to never-users, men who took multivitamins ≥10 tablets per week were not at increased risk for advanced (HR [95% CI], 1.14 [0.77–1.70], P_trend = 0.46), lethal (1.07 [0.80–1.44], P_trend = 0.59), or high-grade (1.20 [0.88–1.63], P_trend = 0.83) prostate cancer (Table 2). In terms of duration, multivitamin use of 15 years or more was not associated with risk for advanced (1.10 [0.80–1.50], P_trend=0.46), lethal (1.04 [0.83–1.31], P_trend =0.13), or high-grade (0.98 [0.77–1.24], P_trend = 0.37) prostate cancer compared to never-users (Table 3).

Table 2.

Multivariable hazard ratios (and 95% confidence intervals, CI) of prostate cancer in relation to multivitamin use frequency in the Health Professionals Follow-up Study, 1986–2016.

Risk of prostate cancer	Hazard Ratio (95% CI)
	Never use	Past use	1–2 tablets/week	3–5 tablets/week	6–9 tablets/week	≥10 tablets/week	P-trend †
All prostate cancer, n	1306	1416	198	480	2939	479
Model 1	1.0	1.07 (0.99, 1.16)	1.06 (0.91, 1.23)	1.05 (0.94, 1.17)	1.09 (1.02, 1.17)	1.09 (0.98, 1.22)	0.03
Model 2	1.0	1.05 (0.97, 1.14)	1.03 (0.88, 1.21)	1.02 (0.91, 1.14)	1.05 (0.97, 1.14)	1.05 (0.93, 1.18)	0.30
Advanced prostate cancer, n	137	114	21	36	213	43
Model 1	1.0	1.04 (0.81, 1.35)	1.26 (0.79, 2.02)	0.93 (0.64, 1.35)	1.07 (0.85, 1.34)	1.32 (0.93, 1.88)	0.23
Model 2	1.0	1.00 (0.77, 1.30)	1.21 (0.74, 1.99)	0.88 (0.58, 1.32)	1.01 (0.77, 1.33)	1.14 (0.77, 1.70)	0.46
Lethal prostate cancer, n	253	202	33	74	424	79
Model 1	1.0	0.98 (0.81, 1.19)	1.06 (0.73, 1.54)	1.03 (0.79, 1.34)	1.10 (0.93, 1.30)	1.25 (0.96, 1.63)	0.11
Model 2	1.0	0.94 (0.78, 1.14)	0.98 (0.67, 1.45)	0.95 (0.72, 1.27)	1.01 (0.82, 1.23)	1.07 (0.80, 1.44)	0.59
High-grade, n	180	204	26	61	388	73
Model 1	1.0	1.09 (0.89, 1.34)	1.03 (0.67, 1.56)	0.99 (0.73, 1.33)	1.01 (0.84, 1.22)	1.20 (0.91, 1.59)	0.55
Model 2	1.0	1.05 (0.85, 1.29)	1.01 (0.65, 1.55)	0.98 (0.71, 1.34)	1.00 (0.80, 1.24)	1.20 (0.88, 1.63)	0.83

All cases: T1a excluded; Advanced cases: stage T3b, T4, N1, or M1; Lethal cases: distant metastases or prostate cancer specific death; High-grade cases: Gleason score 8–10.

Model1: adjusted for current age, time period, race(European American, African American, Asian American, other), family history of prostate cancer (yes, no), lagged PSA testing history (yes, no, lagged by one period to avoid counting diagnostic PSA tests), lagged PSA testing in >50% of possible time periods (yes, no, lagged by one period to avoid counting diagnostic PSA tests), history of diabetes (yes, no), current aspirin use (yes, no), body mass index at age 21 (<20, 20–22.5, 22.5–25, ≥25 kg/m²), current body mass index (<21, 21–25, 25–30, ≥30 kg/m²), height (<68, 68–70, 70–72, ≥72 inches), smoking (never, former/quit > 10 years ago, former/quit ≤ 10 years ago, current), vigorous physical activity (quintiles of metabolic equivalent off task (MET)-h/week), total calories (quintiles of kcal/day), red meat (quintiles of servings/week), tomato-based foods (quintiles of servings/week) and fish (quintiles of servings/week).

Model2: additionally adjusted for supplemental zinc use (past, never, low-dose, high-dose), selenium use (past, never, low-dose, high-dose), vitamin E use (past, never, low-dose, high-dose), vitamin A use (past, never, low-dose, high-dose) and number of overall supplement use (to capture overall pattern of tendency to use supplements; 0, 1, 2, 3, ≥4 per day). All covariates, except race, height, body mass index at age 21 and family history of prostate cancer, were updated with each questionnaire available.

^†:using median value of each category (0, 1, 4, 7.5, 12). Past use as missing.

Table 3.

Multivariable hazard ratios (and 95% confidence intervals, CI) of prostate cancer in relation to duration of multivitamin use in the Health Professionals Follow-up Study, 1986–2016.

Risk of prostate cancer	Hazard Ratio (95% CI)
	Never use	Past use of any duration	1–4 years	5–9 years	10–14 years	≥15 years	P-trend †
All prostate cancer, n	1306	1416	722	766	868	1740
Model 1	1.0	1.07 (0.99, 1.16)	1.06 (0.97, 1.17)	1.18 (1.08, 1.30)	1.02 (0.93, 1.11)	1.10 (1.02, 1.19)	0.14
Model 2	1.0	1.05 (0.97, 1.14)	1.04 (0.94, 1.15)	1.14 (1.03, 1.26)	1.00 (0.91, 1.10)	1.03 (0.94, 1.13)	0.80
Advanced prostate cancer, n	137	114	56	52	80	125
Model 1	1.0	1.05 (0.82, 1.36)	1.07 (0.78, 1.48)	1.10 (0.79, 1.53)	0.98 (0.74, 1.29)	1.23 (0.94, 1.61)	0.18
Model 2	1.0	1.01 (0.78, 1.31)	1.03 (0.73, 1.47)	1.04 (0.72, 1.50)	0.94 (0.68, 1.29)	1.10 (0.80, 1.50)	0.46
Lethal prostate cancer, n	253	202	96	102	178	234
Model 1	1.0	0.98 (0.81, 1.19)	0.90 (0.71, 1.15)	1.11 (0.87, 1.41)	1.16 (0.96, 1.42)	1.18 (0.97, 1.43)	0.02
Model 2	1.0	0.94 (0.78, 1.14)	0.84 (0.65, 1.10)	1.03 (0.79, 1.33)	1.08 (0.86, 1.35)	1.04 (0.83, 1.31)	0.13
High grade, n	180	204	88	102	120	238
Model 1	1.0	1.09 (0.89, 1.34)	0.99 (0.77, 1.29)	1.15 (0.89, 1.48)	1.00 (0.79, 1.27)	1.02 (0.83, 1.26)	0.76
Model 2	1.0	1.05 (0.85, 1.29)	0.97 (0.73, 1.29)	1.12 (0.85, 1.48)	1.03 (0.79, 1.34)	0.98 (0.77, 1.24)	0.37

All cases: T1a excluded; Advanced cases: stage T3b, T4, N1, or M1; Lethal cases: distant metastases or prostate cancer specific death; High-grade cases: Gleason score 8–10.

Model1: adjusted for current age, time period, race(European American, African American, Asian American, other), family history of prostate cancer (yes, no), lagged PSA testing history (yes, no, lagged by one period to avoid counting diagnostic PSA tests), lagged PSA testing in >50% of possible time periods (yes, no, lagged by one period to avoid counting diagnostic PSA tests), history of diabetes (yes, no), current aspirin use (yes, no), body mass index at age 21 (<20, 20–22.5, 22.5–25, ≥25 kg/m²), current body mass index (<21, 21–25, 25–30, ≥30 kg/m²), height (<68, 68–70, 70–72, ≥72 inches), smoking (never, former/quit > 10 years ago, former/quit ≤ 10 years ago, current), vigorous physical activity (quintiles of metabolic equivalent off task (MET)-h/week), total calories (quintiles of kcal/day), red meat (quintiles of servings/week), tomato-based foods (quintiles of servings/week) and fish (quintiles of servings/week).

Model2: additionally adjusted for supplemental zinc use (past, never, low-dose, high-dose), selenium use (past, never, low-dose, high-dose), vitamin E use (past, never, low-dose, high-dose), vitamin A use (past, never, low-dose, high-dose) and number of overall supplement use (to capture overall pattern of tendency to use supplements; 0, 1, 2, 3, ≥4 per day). All covariates, except race, height, body mass index at age 21 and family history of prostate cancer, were updated with each questionnaire available.

^†:using median of each category (0, 2.5, 7, 12, 17). Past use of any duration as missing.

The null association between multivitamin use and risk of prostate cancer remained when stratified by family history, BMI, smoking status and vigorous physical activity (eTable 1). For high-grade prostate cancer, multivitamin use of ≥10 tablets/week was associated with an increased risk among men with family history of prostate cancer (2.47 [1.15–5.31]) compared to never-use. However, the HRs fluctuated across categories of duration and no overall trend was noted (P_trend=0.75). No association was observed among men without family history (0.99 [0.70–1.42], P_trend=0.81). Among men who were current smoker or quit ≤ 10 years, those who took multivitamin ≥10 tablets/week were at lower risk of high-grade prostate cancer compared to never-users (0.66 [0.24–1.87], P_trend=0.04).

In joint analysis of multivitamin and individual supplements, we observed increased risk for advanced, lethal and high-grade prostate cancer comparing zinc vs. no zinc supplement use across levels of multivitamin use frequency (eTable 2). To investigate whether zinc supplement or multivitamin was driving the positive joint association, we conducted a stratified analysis by zinc supplement use and found multivitamin use was not associated with higher risk of aggressive prostate cancer among zinc supplement non-users and users (eTable 3). The association pattern remained null after applying a 4–6- or 8–10-year lag (eTable 4) and when restricted to post-PSA era (eTable 5). Restriction to men with complete multivitamin information did not change the results (eTable 6 and 7).

Discussion

We found regular or long-duration multivitamin use was not associated with risk of overall and aggressive prostate cancer. While RCT such as the Selenium and Vitamin E Cancer Prevention Trial (SELECT) showed high-dose vitamin E significantly increased risk of prostate cancer²¹, how multivitamins, which usually contain low-dose of vitamins and minerals, are associated with prostate cancer risk was unclear ^7–14. In PHS II, men taking a daily multivitamin had no reduced or increased risk of prostate cancer (0.98 [0.88–1.09])⁷. However, the majority of cases were earlier stage, low-grade cancer, resulting from prevalent PSA screening. Similarly, one cohort⁸ and two case-control studies^9,10 also found null associations between multivitamin use and risk of prostate cancer. Another RCT, SU.VI.MAX¹⁴, showed that supplementation with a mixture of low doses antioxidant vitamin and minerals (vitamin C, vitamin E, beta-carotene, selenium and zinc) daily was associated with reduced risk of prostate cancer (0.52 [0.29–0.92]) among men with normal PSA at baseline (<3 ug/L). The comparability of the low doses antioxidant mixtures with over-the-counter multivitamins remains in question, as the latter usually contain more than 5 vitamins and minerals with dosage approximate to the recommended dietary allowance (RDA)²².

Two cohort studies raised concern over potential harmful effects of multivitamin use on risk of advanced prostate cancer. The Cancer Prevention Study II¹² showed that men who took multivitamins ≥15 times/month had marginally increased risk of prostate cancer compared to non-users (1.07 [0.99–1.15]). The observed positive association could be explained as multivitamins promoted the growth of preexisting tumors; however, it could also be caused by reverse causation: men who died during the initial follow-up years could have increased their multivitamin use because they were experiencing symptoms of advanced but undiagnosed prostate cancer. The other cohort study, the NIH-AARP Diet and Health study¹¹, showed increased risk of advanced (RR:1.32, 95%CI: 1.04–1.67) and fatal (RR:1.98, 95%CI: 1.07–3.66) prostate cancers for men with multivitamin use of ≥7 times/week compared with never-users. However, information on PSA screening and digital rectal examination was not available for all participants and was not included in the multivariable model. In current analysis, we observed null association between regular or long-term multivitamin use and risk of advanced and lethal prostate cancer. The results remain unchanged when we applied various latency periods and restricted to the post-PSA era.

Both our and previous study¹¹ observed a potential higher risk of advanced prostate cancer associated with multivitamin use among men with family history of prostate cancer, although the trend tests were all statistically insignificant. It could be that the presence of one or several components of the multivitamin interact with certain genetic factors that promoted the growth of advanced prostate cancer, and such interaction effect does not differ by nutrient dosage. Previously, data from PHS II suggested a non-significant reduction in cancer risk with multivitamin use among current smokers (0.72 [0.50–1.03])⁷. In current analysis, we found a possible lower risk of high-grade prostate cancer associated with multivitamin use among current smokers or who quit less than 10 years. However, the number of cases for each frequency groups was very small (n=5, 6, 46, 7 for 1–2, 3–5, 6–9 and ≥10 tablets/week, respectively) and confidence intervals were wide, these results should be interpreted with causation. The increased risk of aggressive prostate cancer observed in joint analysis of zinc supplement and multivitamin use was most likely due to potential harmful effect of supplemental zinc on advanced prostate cancer, as in stratified analysis we did not observe increased risk of prostate cancer with higher frequency of multivitamin use among zinc supplement non-users and users. Chronic high intake of supplemental zinc can cause abnormalities of genitourinary functions²³. Although we previously showed that excessive zinc supplement was associated with increased risk of aggressive prostate cancer²⁴, our current analysis suggested multivitamin use did not increase the risk further in men taking zinc supplement. Future analysis from PHS II⁷ and the ongoing Cocoa Supplement and multivitamin Outcomes Study (COSMOS)²⁵ may provide additional insights into these areas.

Our study has several strengths, including the prospective design, repeated assessment of multivitamin use and frequency every two years, long follow-up, large number of events, comprehensive information on lifestyle factors and PSA screening. However, our study is not without limitations. First, there are potential measurement errors in the self-reported questionnaires. However, given the prospective study design, any mismeasurement in the exposures would likely be random with respect to endpoints, resulting in conservative associations. Second, we cannot entirely rule out the possibility of unmeasured or residual confounding. However, the homogeneity of the study population and comprehensive data on the risk factors may help minimize potential confounding. Third, the commercial forms of multivitamins have changed over time and add to inevitable measurement error. In HPFS, participants provided detailed brands and types of multivitamins, but it was difficult to come up with consistent definitions of multivitamin use because of too many types and inconsistent formulations throughout time. Finally, the results of our study may not be generalizable to different population or ethnic groups, especially with different nutrient status. Both our cohort and the PHS II⁷ consists of male physicians who were expected to be already well-nourished. Additional studies evaluating how baseline nutritional status may modify the effect of multivitamin on cancer are needed.

In summary, we found no evidence to suggest regular, or long-duration multivitamin use was associated with risk of overall or aggressive prostate cancer in a well-nourished adult male population.

Abbreviations:

BMI

body mass index

CI

confidence interval

FFQ

food frequency questionnaire

HPFS

Health Professionals Follow-up Study

HR

Hazard ratio

NHANES

National Health and Nutrition Examination Survey

NIH

National Institutes of Health

PHS II

Physicians’ Health Study II

PSA

prostate-specific antigen

RCT

randomized clinical trial

RDA

recommended dietary allowance

RR

Relative risk

Data availability:

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.