heart rate and prostate cancer mortality: results of a ...thetic activity, such as blood pressure...

Vol. 4, 61 1-616, Seplember 1995 Cancer Epidemiology. Biomarkers & Prevention 611

3 The abbreviations used are: NGF, nerve growth factor; ECG, electrocardiogram;

RR, relative risk; CI, confidence interval; BPH, benign prostatic hyperplasia.

Heart Rate and Prostate Cancer Mortality: Results of a

Prospective Analysis1

Peter H. Gann,2 Martha L. Daviglus, Alan R. Dyer, andJeremiah Stamler

Department of Preventive Medicine IP. H. G., M. L. D., A. R. D., J. S.] and

Robert H. Lurie Cancer Center [P. H. G.], Northwestem University Medical

School, Chicago, Illinois 6061 1

Abstract

Recent studies suggest that local levels of sympatheticnervous activity influence the growth of prostatic tissue.In several epidemiological studies, resting heart rate, anindicator of overall sympathetic activity, was positivelyassociated with all noncardiovascular and cancer deathamong men. However, no previous analyses have focusedon the specific relationship of heart rate to prostatecancer mortality.

We studied 22,380 men enrolled in the ChicagoHeart Association cohort from 1967-1973, who had heartrate (HR) determined by electrocardiogram. Mean lengthof follow-up (for mortality) was 19.2 years. We computedage-adjusted rates for prostate cancer death by variable

of interest and fitted proportional hazards models toestimate relative risks (RRs) adjusted for potentialconfounders. In a model controlling for age, body massindex, blood pressure, serum cholesterol, smoking,postload plasma glucose, and years of education, the RRfor a 10 beat/mm higher HR was 1.26 (95% confidenceinterval = 1.04-1.51). Age-adjusted RRs across higherquintiles for HR were 1.00, 1.55, 1.85, 2.18, and 2.69(P trend = 0.006). Survival curves indicated that theelevated risk was not confined to the early years offollow-up.

Because little is known about factors that determinerisk of prostate cancer death, these results could proveimportant even if due to an unmeasured etiological factorother than heart rate itself. The results are consistentwith the hypothesis that local neurotrophic factorsassociated with sympathetic activity influence theprogression of prostate cancer.

Introduction

The causes of prostate cancer remain largely unknown, al-though the disease is now the second leading cause of cancer

Received 1/19/95; revised 3/21/95; accepted 4/3/95.

I This research was supported by grants from the American Heart Association,

American Heart Association of Metropolitan Chicago, Illinois Heart Association,

National Heart, Lung and Blood Institute of the NIH (5ROI-HL21O1O), and the

Chicago Health Research Foundation. Part of the material in this paper waspresented at the Annual Meeting of the American Society for Preventive Oncol-

ogy. Bethesda, MD, March, 1994.

2 To whom requests for reprints should be addressed, at Department of Preventive

Medicine, Northwestern University Medical School, 680 North Lake Shore Drive,

Suite 1102, Chicago. IL 60611.

death among men in many Western countries (1). Studies ofmigrants and international comparisons of occult prostate can-cer at autopsy indicate that environmental or lifestyle factors

play a large role and that high-risk populations differ fromlow-risk ones in the frequency with which small, insignificant

tumors undergo progression (2). The general hypothesis thatsympathetic nervous activity is involved in controlling tissuegrowth is an old one. Previous research has established thatsympathetic activity contributes to skeletal muscle hypertrophy

and to compensatory hypertrophy after removal of a pairedorgan (3,4). Recent evidence suggests that sympathetic inner-

vation or, more precisely, the balance of local sympathetic andparasympathetic activity, can influence the rate of prostatic

growth (5). This phenomenon could be due to NGF3 and similarneurotrophic peptides that have demonstrable growth-stimulat-ing properties in non-neuronal tissues and the tissue concen-trations of which correlate with the level of sympathetic inner-vation and activity (6). Thus, because androgens are already

implicated in prostate growth, the prostate could be similar to

a more extensively studied secretory organ, the salivary gland,in which growth is modulated by an interaction of androgenic

and neural factors (7).Epidemiological analyses relating prostate cancer to phys-

iological variables that serve as surrogate markers of sympa-thetic activity, such as blood pressure and heart rate, are rare.However, data from several prospective cohort studies haveindicated an association between heart rate or blood pressure

and total noncardiovascular or cancer mortality (8-13). Restingheart rate, and to a lesser extent blood pressure, are regulated bythe balance of activity of sympathetic and cholinergic controlcenters in the brain stem (14). Output of adrenergic tone fromthese centers through sympathetic nerves to innervated targettissues varies considerably from person to person (15).

In this paper we evaluate the associations of heart rate andblood pressure with prostate cancer mortality in a prospective

cohort with over 19 years of follow-up. We also review epide-miological and other types of evidence bearing on the hypoth-esis that sympathetic activity plays a causal role in stimulatingabnormal prostate growth.

Materials and Methods

Study Population. Between 1967 and 1973, the ChicagoHeart Association Detection Project in Industry examined39,572 people (including 22,380 men ranging in age from 15 to90 years) who were employed by 81 firms in the Chicago area.At each employment site, one of two teams of nurses and

technicians collected baseline data. A self-administered ques-tionnaire was used to collect demographic data, smoking his-tory, and information on previous medical diagnoses and treat-

on April 24, 2020. © 1995 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

http://cebp.aacrjournals.org/

612 Heart Rate and Prostate Cancer Mortality

ment. In addition to resting heart rate and blood pressure, othervariables measured at baseline included height and weight, total

serum cholesterol, an ECG, and (for those without self-reporteddiabetes) plasma glucose level 1 h after a 50 g p.o. glucose load.

Data on these variables were available for over 99% of maleparticipants. Vital status has been ascertained through 1989,

yielding a mean follow-up of 19.2 years.

Heart Rate and Blood Pressure Measurement. During theinitial phase of the examinations, resting heart rate was deter-mined to the nearest beat/mm by measuring the interval be-

tween R-waves for three consecutive QRS complexes on theECG. During the final phase of the examinations, a heart ratemeter attached to skin electrodes performed the same measure-ment electronically and provided a digital reading. Heart ratewas not systematically measured during a brief middle phase

and, thus, is missing for 2,492 (1 1%) of the men. Technicians

took a single casual supine blood pressure reading using a

standard sphygmomanometer and the fifth phase criterion for

diastolic pressure.

Mortality Ascertainment. Deaths among cohort members be-fore 1979 were ascertained annually by direct mail contact or

matching of cohort records against Social Security Adminis-

tration files. For deaths after 1979, study records were matchedagainst National Death Index records. As of 1979, vital statuswas unknown for only 3% of the cohort. For these men, we

censored their observation period at the year in which they werelast known to be alive. Death certificates were obtained fromstate agencies and were coded by a physician using the 8th

Revision of the International Classification of Diseases,adapted for use in the United States. All coding decisions were

cross-checked by additional study team members. Ninety-onedeaths associated with prostate cancer were ascertained. In 73cases, which form the basis for these analyses, prostate cancerwas the primary or underlying cause of death.

Data Analysis. After reviewing the univariate distributions toassess normality and missing data, we computed crude andage-standardized RR for eight variables: heart rate, systolic and

diastolic blood pressure, body mass index, postload glucose,total serum cholesterol, number of cigarettes currently smokedper day, and years of education. We then used proportionalhazards modeling with SPSS-PC to estimate the same RRs andtheir 95% CIs. We computed correlation coefficients to esti-mate associations between heart rate and other covariates. Be-

cause some variables were not normally distributed, we usedrank order (Spearman) rather than product moment (Pearson)

correlation coefficients. We fitted a “full” proportional hazardsmodel containing a continuous variable for heart rate plus allseven other covariates; we then compared these results to those

obtained in a simpler model with just heart rate and age ascovariates. We divided the population into quintiles by heartrate and computed new proportional hazards models with in-

dicator variables for quintile (lowest quintile as referent). Weevaluated the trend across heart rate categories by using an

ordinal variable for heart rate quintile in the models. To assesswhether these results were distorted by modeling assumptions,we reviewed the age-standardized rates and RRs for each heartrate quintile and fit models with dummy variables for age

categories. We also used Kaplan-Meier plots to check forproportional hazards and to investigate changes in prostate

cancer-free survival over time in subgroups with highly con-trasting baseline heart rate (highest versus lowest tertile). Inaddition, we fitted separate proportional hazards models for the

years 1-10 and 1 1-19 segments of follow-up time. To explorefor effect modification, we compared model coefficients

Table I Selected baseline characteristics of men in the Chicago Heart

Association Detection Project in Industry cohort, l967-l973, total a 22,380

Variable No. missing Mean SD

Age (yrs) (1 40.2 12.5

Education (yrs) 34 13.5 3.2

Heart rate (beats/mm) 2,492 75.9 12.3

Body mass index (kg/m2) 9 26.6 3.7

Serum cholesterol, total (mg/dl) 63 200.9 38.2

Postload plasma glucose (mg/dl) 661 134.0 45.2

Systolic blood pressure (mm Hg) 12 138.8 18.7

Diastolic blood pressure (mm Hg) 12 81.3 11.7

Race 0 n

White 20,184 90.2

Black 1,487 6.6

Other 709 3.2

History of diabetes 87 578 2.6

Insulin or oral agents 263 1.2

No drug treatment 313 1.4

Unknown treatment 2 0

History of hypertension 80 2,864 12.8

Current treatment 887 4.0

No treatment 1,965 8.8

Treatment status unknown 12 0.1

Smoking status 13

Never . 5,935 26.5

Past 6,692 29.9

Current 9.740 43.5

for heart rate from separate models applied to dichotomizedcovariate subgroups.

Results

Table 1 shows selected characteristics of the men at baseline.They were predominantly white (90%) and had a widely spread

distribution for resting heart rate (SD - 12.3). Data on exercisefrequency were not obtained. A substantial percentage (43.5%)were currently smoking; only 4% were taking antihypertensivemedication. A percentage (2.6%) of subjects reported a historyof diabetes, and 1 .2% reported use of insulin or oral antidiabeticdrugs.

Table 2 presents crude and age-adjusted RRs of prostate

cancer death for several baseline variables. Heart rate had theclearest age-adjusted association, with risk higher by 26% foreach 10 beat/mm increment in heart rate. Diastolic and systolicblood pressure had weak positive associations with prostatecancer mortality after age adjustment. Among the small sub-group of Black men, only one prostate cancer death occurred.Because RR according to race could not be estimated preciselyand because including adjustment for race did not affect theassociation of heart rate with prostate cancer mortality, weomitted a separate variable for race from further analyses. To

determine which covariate associations might be confoundingthe age-adjusted risk estimates, we calculated correlations be-

tween heart rate and other baseline variables (Table 3). Bloodpressure (both systolic and diastolic) and the number of ciga-rettes smoked per day had substantial positive correlations withheart rate. P values for the correlation coefficients are notinterpretable because even a small correlation produces a low

P value because of the large number of men.Table 4 presents RR estimates from a proportional hazards

model that included heart rate as a continuous variable and hadsimultaneous adjustment for seven other covariates. This model



Cancer Epidemiology, Biomarkers & Prevention 613

Table 2 Crude and age-adjusted associati ons between bas dine variab les and prostate ca ncer mortality

VariableRR

.increment

Crude Age adjusted

RR 95% Cl RR 95% CI

Heart rate (beats/mm) 10 1.19 0.99-1.41 1.26 1.05-1.49

Body mass index (kg/m2) 5 1.23 0.92-1.64 1.03 0.74-1.42

Postload plasma glucose (mg/dI) 10 1.07 1.04-1.09 0.99 0.94-1.03

Serum cholesterol, total (mg/dI) 10 1.08 1.02-1.13 1.01 0.94-1.07

No. of cigarettes/day S 0.94 0.86-1.04 1.05 0.95-1.14

Systolic BP5 (mm Hg) 10 1.32 1.21-1.45 1.06 0.95-1.17

Diastolic BP (mm Hg) 10 1.48 1.25-1.74 1.11 0.92-1.34

Education (years) I 0.85 0.78-0.92 0.96 0.90-1.03

“ These numbers are convenient increments for expressing relative risk and do not represent equivalent portions of the distribution for each variable. Thus, the RRs shown

cannot he directly compared across variables. RR increments can be compared to the standard deviations shown in Table I . RR, relative risk estimated by proportional

hazards regression.

I, BP, blood pressure.

Table 3 Spearman correlation coefficients for heart rate with other

baseline variables

Variable Correlation coefficient No. of men

Age (yrs) 0.033 19,888

Body mass index (kg/m2) 0.033 19,879

Postload plasma glucose (mg/dl) 0.1 1 1 19,117

Serum cholesterol, total (mg/dl) 0.093 19,828

Number of cigarettes/day 0.207 19,873

Diastolic BP” (mm Hg) 0.185 19,879

Systolic BP (mm Hg) 0.262 19,879

Education (yrs) -0.122 19,855

‘: BP, blood pressure.

Table 4 Multivariable proportional hazards regression model for prostate

cancer death: Chicago Heart Association Detection Project in Industry cohort

(n 72 deaths)

Variable . RR RR#{176} 95% CIincrement

Age (yrs) S 2.23 1.94-2.55

Heart rate (heats/mm) 10 1.26 1.04-1.51

Body mass index (kg/m2) S 1.02 0.73-1.42

Serum cholesterol, total (mgjdl) 10 1.00 0.94-1.06

Systolic BP” (mm Hg) 10 1.03 0.92-1.15

No. of cigarettes/day 5 1.02 0.93-1.12

Postload plasma glucose (mg/dl) 10 0.97 0.92-1.02

Education (yrs) I 0.98 0.91-1.02

“ Relative risk, multiple of risk for each difference of x units for a variable. The

numbers of units used to calculate RR are shown in the column headed ‘�RR

increment.”

F, BP, blood pressure.

included systolic but not diastolic blood pressure because thesemeasures were highly collinear. Similar results were obtainedwhen diastolic rather than systolic pressure was included. The

RR for a 10 beat/mm-higher heart rate was 1.26, identical to theRR adjusted for age alone. We found no evidence for effectmodification of the heart rate-prostate canc#{231}rassociation based

on comparison of RRs across various subcohort strata, includ-ing those who were defined as older or younger than the median

age at baseline. ..

Fig. 1 shows age-adjusted RRs and 95% CIs for each

quintile of heart rate, with the lowest quintile as the referentgroup. These estimates are derived from a model; age standard-ization analysis produced nearly identical point estimates. The

data are consistent with a linear increase in risk across heart rate

quintiles (P = 0.006) and an approximate 2.7-fold highermortality risk in the highest quintile compared to the lowest(median heart rate, 92.0 and 54.5 beats/mm, respectively).

Curves for prostate cancer-specific survival of the men inthe lowest and highest tertiles of baseline heart rate are shownin Fig. 2. The curves clearly separate after about 6 years offollow-up and remain proportionally separated throughout theremainder of follow-up. This supports applicability of the pro-portional hazards model used in the analyses and also indicatesthat excess cancer deaths in the high heart rate group were notconfined to the early follow-up period. Separate proportionalhazards models for deaths occurring in years 1-10 and 1 1-19 offollow-up gave similar coefficients for heart rate.

Discussion

In a cohort of over 22,000 men followed for an average of 19years, we found evidence for an association between resting

heart rate and mortality due to prostate cancer. This associationwas unchanged by adjustment for body mass, serum choles-terol, blood pressure, smoking, postload plasma glucose, race,and educational level. Positive associations between bloodpressure and prostate cancer mortality were apparent in crude

analyses, but were primarily attributable to the associationbetween blood pressure and age. Our data indicate that each 10

beat/mm increment in heart rate is associated with a 26% higherrisk of prostate cancer death, and that men in the highestquintile for heart rate have more than a doubling in risk com-

pared to men in the lowest quintile. This association was notattributable to deaths occurring close to baseline and, thus, toincreased heart rate among those ill with cancer but rather itwas apparent throughout the follow-up period.

Several previous epidemiological studies provide resultsthat bear at least indirectly on the hypothesis that sympatheticactivity affects prostate cancer risk. No studies have used directmeasures of sympathetic activity, such as urinary catechol

output. Members of our group first reported a possible linkbetween blood pressure and cancer mortality 19 years ago (16),a finding replicated in some (9) but not other (10) analyses. Asubsequent analysis by our group found evidence for an asso-

ciation between baseline heart rate and overall and site-specificcancer death in two cohorts, including the 5-year follow-up ofthe Chicago Heart Association cohort (17). Prostate cancerdeaths were too few to be analyzed separately. An analysisbased on a 30-year follow-up of the original Framingham

cohort revealed a strong trend for the increasing noncardiovas-cular death rate by category of baseline heart rate, especially formen (11). A recent report on the British Regional Heart Study



p-trend=0.OO6

5.66

4.70

4.08

-�

5)>

5)

3.45

0.840.69

1.28

1.01

Q1 Q2 Q3 04 Q5

54.5 69.5 75.5 82.5 92.0

6.00

5.00

4.00

3.00

2.00

1.00

0.00 -

Median(beats/mm)

Fig. 1. Age-adjusted RR and 95% CIs for prostate cancer death by quintile (Q) of baseline heart rate.

1.002

1.000

.998

.996

.994

P92

highest tertile( � 81 beats/mi.i)

Lowest tertile

.( � 70 beats/mitt)

o 2000 4000 6000

Length of Follow-up (Days)

Fig. 2. Prostate cancer mortality survival curves for men in the highest and lowest tertiles for resting heart rate.

8000


C,)

.�

a8ac)

( all male) cohort concluded that resting heart rate was anindependent predictor of cancer mortality (8). Furthermore,analysis by cancer site showed the strongest association forthose with cancers outside the lung and digestive systems.

In a prospective analysis of cancer incidence amongJapanese men in Hawaii, Severson et a!. (18) reported anassociation between resting heart rate and colon but not prostatecancer (18). Two recent prospective studies on risk factors forBPH obtained results consistent with a relationship betweensympathetic activity and prostate growth. In one, an association

was seen between abdominal obesity and both incidence of

BPH surgery and symptom severity (19). In the other, risk ofBPH surgery was greater among men with higher blood pres-

sure (20). Neither of these studies, however, was able to sep-arate the possible effect of sympathetic tone on dynamic blad-der outflow obstruction from an effect on gland size itself.

Increased heart rate in itself is not a plausible causal factorfor prostate cancer death. Although we assert the possibilitythat these results are explained by the strong relationship of

heart rate to increased sympathetic tone, they might have arisen



.-�

ly.. ___________

“��‘A . .

Hyperinsulinemia9,

Prostate ProstateStromal Cell Epithelial Cell

Cancer Epidemiology, Biomarkers & Prevention 615

Fig. 3. Schematic diagram illustrating the

hypothetical relationship of sympathetic ac-

tivity and associated traits with neurotrophic

mechanisms affecting prostate growth.

NGF-, nerve growth factor-like peptide;

EGF, epidermal growth factor; IGF-1, insu-

lin-like growth factor; trk, tyrosine kinase.

if factors associated with sympathetic activity, rather than thisactivity itself, were directly involved in the causal pathway.

Saturated fat and red meat intake, e.g. , have been linked toprostate cancer risk in some studies and could be positivelyrelated to resting heart rate (21). We were unable to directlyadjust for this potential source of confounding due to lack ofdietary data for this cohort, but we found that adjustment fortotal serum cholesterol, a factor associated with saturated fat

intake, did not affect our point estimates. We also found nomaterial difference due to adjustment for measurements ofglucose intolerance or body mass index, both of which havebeen linked with increased sympathetic activity (22). We had

no data on waist or hip size and, thus, could not adjust for

central obesity, which has been specifically associated withincreased sympathetic tone, as well as changes in hormonalmetabolism that could affect prostate growth (23, 24).

A positive association of heart rate and prostate cancer

mortality could also be explained by confounding due to phys-ical activity. A large cohort study of physical activity recentlyreported a deficit of prostate cancer deaths among men in the

highest activity categories (25). However, in the British Re-gional Heart Study, the association between cancer deaths and

heart rate remained after adjustment for physical activity (8).The relationship of physical activity to heart rate is complexbecause it involves both variation in sympathetic activity andvariation in cardiovascular fitness, including improved strokeefficiency. Our data are not particularly consistent with theview that fitness is the important feature because we observed

a smooth, linear relationship with mortality throughout therange of heart rates rather than merely a deficit among men withthe lowest heart rates.

We also considered the possibility that men with elevatedheart rates are more likely to have cardiovascular disease,which could lead to increased medical surveillance and a higherlikelihood of having prostate cancer diagnosed or recorded onthe death certificate. The association between heart rate andprostate mortality remained, however, after adjustment for 5ev-

era] major cardiovascular risk factors. Adjustment for major orminor ECG abnormalitites also had no effect on this associa-

tion. Furthermore, by including in the analysis only prostatecancers responsible for death we minimized problems due toincidental recording of prostate cancer. We note that studies ofprostate cancer mortality as opposed to incidence have the

advantage of restricting analysis to aggressive cases and ofinsulating the study from excessive bias due to variations inutilization of medical care.

At least four lines of evidence from laboratory studiessupport the biological plausibility of a hypothesis linking sym-

pathetic nervous activity to the control of prostate growth: (a)sectioning of the pelvic ganglion in rats produces a markeddecrease in prostate gland size and loss of secretory activity (5).

Selective sectioning of the sympathetic nerves on one side inyoung rats leads to inhibition of cell division and growth on the

affected side (26); (b) transfection of fibroblasts with a geneexpressing a constitutively activated al-adrenergic receptorinduces mitogenesis and neoplastic transformation (27). In theprostate, most of the autonomic innervation consists of abun-dant cel-adrenergic nerve endings in stromal tissue;

(c) many investigators currently believe that paracrine

growth factors secreted by stromal cells play a dominant role incontrol of epithelial cell growth in the prostate (28). Djakiewand coworkers recently reported that in the human prostate,NGF-like proteins are localized to the stroma whereas NGFreceptors are found primarily on epithelial cells (29). Mediataken from cultures of stromal cells stimulates growth of pros-tate cancer cells in vitro, but this stimulation is blocked sub-

stantially by the addition of anti-NGF antibodies (30); and(d) the prostate gland was the second type of non-neuronal

tissue in which NGF activity was found, the first being themouse submaxillary salivary gland (31). It is now establishedthat NGF production in the salivary gland is regulated bya-adrenergic stimulation in the presence of sufficient androgen(7). It is not yet known whether production of the NGF-likeprotein by prostate stroma is also androgen dependent. How-




ever, recent reports indicate that adrenergic neurotransmitters

increase NGF synthesis in non-neuronal tissues, and that al-adrenergic agonists cause large increases in the levels of epi-dermal growth factor, a potent mitogen, in rat prostatic secre-tions (32, 33). In summary, it is plausible that men with higherlevels of central sympathetic tone (as reflected in higher heartrate) have higher levels of adrenergic activity and, thus, neu-rotrophins in the prostate, leading to androgen-mediated stim-ulation of prostate cancer growth and higher mortality. Keyrelationships in this hypothesis are summarized in the schemain Fig. 3.

The results of this study need to be evaluated by attemptsat replication in other datasets, particularly because the hypoth-esis we have outlined has not been explored previously in

detail. Confirmation of an association between heart rate andprostate cancer mortality could open new avenues for research,even if sympathetic activity is not directly causal. We believe,however, that these results emphasize the need to devote moreattention to the role of neurotrophic factors in prostate carci-nogenesis. The large number of men now receiving a-adren-

ergic-blocking drugs for benign prostatic hyperplasia couldprovide a population for future studies of this question. A recent

analysis of cancer incidence among chronic schizophrenics inDenmark found a 44% reduction in risk of prostate cancer that

was largely attributable to patients who had received high doses

of chlorpromazine, an antipsychotic medication with pro-nounced anti-adrenergic side effects (34). The hypothesis mightalso provide some insight into the reasons for higher prostate

cancer mortality among African-American men, who also haveelevated rates of hypertension, perhaps as a result of higherbasal and reactive sympathetic activity (35).

Acknowledgments

The authors thank Daniel Garside for expert assistance with statistical program-

ming and the many individuals whose work made it possible to establishand maintain the Chicago Heart Association cohort. See W. Levine ci al.,

Am. J. Epidemiol.. /31: 254-262, 1990, for a complete list of acknowledged

contributors.

References

I. Boring, C. C., Squires, T. S., Tong, M. S., and Montgomery, S. Cancer

Statistics. 1994. CA Cancer J. Clin., 44: 7-26, 1994.

2. Nomura, A. M. Y., and Kolonel, L. N. Prostate cancer: a current perspective.

Am. J. Epidemiol., 13: 200-227, 1991.

3. Kim, Y. S.. and Sainz, R. D. �-Adrenergic agonists and hypertrophy of skeletalmuscles. Life Sci., 50: 397-407, 1992.

4. Gerendai. I., Marchetti, B., Maugeri, S., Prato, A., Scapagnini, U., and Halasz,B. Norepinephrinergic involvement in the neuronal mechanism regulating com-

pensatory ovarian hypertrophy. Acta Eur. Fertil., 8: 327-331, 1977.

S. Wang, J., McKenna, K. E., McVary, K. T., and Lee, C. Requirement of

innervation for maintenance of structural and functional integrity in the rat

prostate. Biol. Reprod.. 44: 1171-1176, 1991.

6. Levi, A., and Alema, S. The mechanism of action of nerve growth factor.

Annu. Rev. Pharmacol. Toxicol., 31: 205-228, 1991.

7. Walker, P. The mouse submaxillary gland: a model for the study of hormon-

ally dependent growth factors. J. Endocrinol. Invest., 5: 183-196, 1982.

8. Wannamethee, G., Shaper, A. G., Macfarlane, P. W. Heart rate, physical

activity, and mortality from cancer and other noncardiovascular diseases. Am. J.

Epidemiol., 137: 735-748, 1993.

9. Svardsudd, K., and Tibblin, G. Mortality and morbidity during 13.5 years’

followup in relation to blood pressure. Acta Med. Scand., 205: 483-492, 1979.

10. Raynor, W. J.. Jr., Shekelle, R. B., Rossof, A. I-I., Maliza, C., and Paul, 0.

High blood pressure and 17-year cancer mortality in the Western Electric HealthStudy. Am. J. Epidemiol., 113: 371-377. 1981.

11. Kannel, W. B., Kannel, C., Paffenberger, R. S., Jr., and Cupples, L. A. Heart

rate and cardiovascular mortality: the Framingham Study. Am. Heart J. 113:

1489-1494, 1987.

12. Garcia-Palmieri, M., Sorlie, P. D., Costas, R., Jr., and Havlik, R. J. An

apparent inverse relationship between serum cholesterol and cancer mortality in

Puerto Rico. Am. J. Epidemiol., 114: 29-40, 1981.

13. Dyer, A. R., Persky, V., Stamler, J., Paul, 0., Shekelle, R. B., Berkson, D. M.,

Lepper, M., Schoenberger, J. A., and Lindberg, H. A. Heart rate as a prognostic

factor for coronary heart disease and mortality: findings in three Chicago epide-

miologic studies. Am. J. Epidemiol., 112: 736-749, 1980.

14. Julius, J., Esler, M. D., and Randall, 0. S. Role of the autonomic nervous

system in mild human hypertension. Clin. Sci. Mol. Med., 48: 243s-252s, 1975.

15. Young, J. B., Troisi, R. J., Weiss, S. T., Parker, 0. R., Sparrow, D., and

Landsberg, L. Relationship of catecholamine excretion to body size, obesity, and

nutrient intake in middle-aged and elderly men. Am. J. Clin. Nutr., 56: 827-834,

1992.

16. Dyer, A. R., Stamler, J., Berkson, D. M., Lindberg, H. A., and Stevens, E.

High blood-pressure: a risk factor for cancer mortality? Lancet 1: 1051-1056,1975.

17. Persky, V., Dyer, A. R., Leonas, J., Stamler, J., Berkson, D. M., Lindberg, H.

A., Paul, 0., Shekelle, R. B., Lepper, M. H.. and Schoenberger, J. A. Heart rate:

a risk factor for cancer? Am. J. Epidemiol., /14: 477-487, 1981.

18. Severson, R. K., Nomura, A. M. Y., Grove, J. S., and Stemmermann, G. N.

A prospective analysis of physical activity and cancer. Am. J. Epidemiol., 130:

522-529, 1989.

19. Giovannucci, E., Rimm, E. B., Chute, C. C., ci al. Obesity and benign

prostatic hyperplasia: a prospective study. Am. J. Epidemiol., 140: 989-1002,

1994.

20. Gann, P. H., Hennekens, C. H., Longcope, C., Verhoek-Oftedahl, W.,

Grodstein, F., and Stampfer, M. J. A prospective study of plasma hormonelevels, non-hormonal factors, and development of benign prostatie hyperplasia.

Prostate, 26: 40-49, 1995.

21. Young, J. B., and Walgren, M. C. Differential effects of dietary fats on

sympathetic nervous system activity in the rat. Metab. Clin. Exp., 43: 51-60,

1994.

22. Landsberg.L. Hyperinsulinemia: possible role in obesity-induced hyperten-

sian. Hypertension, 19: 1-61-1-66, 1991.

23. Troisi, R. J., Weiss, S. T., Parker, D. R., Sparrow, D., Young, J. B., and

Landsberg, L. Relation of obesity and diet to sympathetic nervous system activity.

Hypertension, 17: 669-677, 1991.

24. Pasquali, R., Casimirri, F., and Cantobelli, S. Effect of obesity and body fat

distribution on sex hormones and insulin in men. Metabolism, 40: 101-104, 1991.

25. Lee, I. M., Paffenbarger, R. S., Jr., and Hsieh, CC. Physical activity and risk

of prostatic cancer among college alumni. Am. J. Epidemiol., 135: 169-179,

1992.

26. McVary, K. T., Razzaq, A., Lee, C., Venegas, M. F., Rademaker, A., and

McKenna, K. E. Growth of the rat prostate gland is facilitated by the autonomic

nervous system. Biol. Reprod., 51: 99-107, 1994.

27. AlIen, L. F., Lefkowitz, R. J., Caron, M. G., and Cotecchia, S. G-protein-

coupled receptor genes as protooncogenes: constitutively activating mutation of

the a 1/3-adrenergic receptor enhances mitogenesis and tumorigenicity. Proc.

NatI. Acad. Sci. USA, 88: 11354-11358, 1991.

28. Camps, J. L., Chang, S., and Hsu, T. C. Fibroblast-mediated acceleration ofhuman epithelial tumor growth in rico. Proc. NatI. Acad. Sei. USA, 87: 75-79,

1990.

29. Graham, C. W., Lynch, J. H., and Djakiew, D. Distribution of nerve growth

factor-like protein and nerve growth factor receptor in human benign prostatichyperplasia and prostatic adenocarcinoma. J. Urol., 147: 1444-1447, 1992.

30. Djakiew, D. Role of nerve growth factor-like protein in the paracrine regu-lation of prostate growth. J. Androl., 13: 476-487, 1992.

31. Harper, G. P., Barde, Y. A., Burnstock, G., el al. Guinea pig prostate is a rich

source of nerve growth factor. Nature (Lond.), 279: 160-162, 1979.

32. Furukawa, S., and Furukawa, Y. Nerve growth factor synthesis and its

regulatory mechanisms: an approach to therapeutic induction of nerve growthfactor synthesis. Cerebrovase. Brain Metab. Rev., 2: 328-344, 1990.

33. Jacobs, S. C., and Story, M. T. Exocrine secretion of epidermal growth factorby the rat prostate: effect of adrenergie agents, cholinergic agents, and vasoactive

intestinal peptide. Prostate, 13: 79-87, 1988.

34. Mortensen, P. B. Neuroleptic medication and reduced risk of prostate cancerin schizophrenic patients. Ada Psychiatr. Scand., 85: 390-393, 1992.

35. Falkner, B. Differences in blacks and whites with essential hypertension:biochemistry and endocrine. Hypertension, 15: 681-686, 1990.



1995;4:611-616. Cancer Epidemiol Biomarkers Prev P H Gann, M L Daviglus, A R Dyer, et al. prospective analysis.Heart rate and prostate cancer mortality: results of a

Updated version

http://cebp.aacrjournals.org/content/4/6/611

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cebp.aacrjournals.org/content/4/6/611To request permission to re-use all or part of this article, use this link



http://cebp.aacrjournals.org/cgi/alerts

mailto:[email protected]



heart rate and prostate cancer mortality: results of a ...thetic activity, such as blood pressure...

Documents