Cancer pattern among hypertensive patients in North Karelia ... - Nature

Journal of Human Hypertension (2005) 19, 373–379 & 2005 Nature Publishing Group All rights reserved 0950-9240/05 $30.00 www.nature.com/jhh

ORIGINAL ARTICLE

Cancer pattern among hypertensive patients in North Karelia, Finland AM Lindgren1, AM Nissinen1,2,3, JO Tuomilehto2,4,5 and E Pukkala6 1

Department of Public Health and General Practice, University of Kuopio, Finland; 2Department of Epidemiology and Health Promotion, National Public Health Institute, Helsinki, Finland; 3Department of Neurology, Kuopio University Hospital, Finland; 4Department of Public Health, University of Helsinki, Helsinki, Finland; 5South Ostrobotnia Central Hospital, Seina¨joki, Finland; 6Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, Helsinki, Finland

The objective of this study was to determine the sitespecific cancer incidence of hypertensive patients and examine the effect of blood pressure-related variables on the risk of cancers with elevated incidence among the hypertensive patients. A record linkage study of Hypertension Register of the North Karelia Project and the Finnish Cancer Registry was conducted. The mean follow-up time was 16 years. A total of 20 529 hypertensive patients were studied. Main outcome measures were standardised incidence ratios and hazard ratios. The overall cancer incidence was close to that of the general population for both men and women. The incidence rate for the kidney cancer was significantly increased in hypertensive patients (standardised incidence ratio 1.34, 95% confidence interval (CI) 1.11–1.60), as well as incidence rates for cancers of pancreas (1.26, 1.02–1.54), and endometrium (1.22, 1.01–1.44) in hypertensive women. The incidence of lung cancer was

significantly decreased (0.86, 0.77–0.95). The incidence of liver cancer was elevated with borderline significance (1.36, 0.99–1.82). In Cox regression models, the use of antihypertensive drugs at baseline was a significant predictor of kidney (hazard ratio for use of antihypertensive drugs 1.89, 95% CI 0.96–3.75) and pancreatic cancer (1.78, 0.99–3.22) in women but not in men. The incidence of endometrial cancer or liver cancer was not related to blood pressure levels or the use of antihypertensive drugs. In women, obesity was a significant predictor of cancers of the endometrium, kidney and liver. In conclusion, increased occurrence of some cancer types among hypertensive patients seem to be partly explained by obesity and the use of antihypertensive drugs. Journal of Human Hypertension (2005) 19, 373–379. doi:10.1038/sj.jhh.1001834 Published online 10 February 2005

Keywords: blood pressure; antihypertensive drugs; kidney cancer; endometrial cancer; pancreatic cancer; liver cancer

Introduction The association between hypertension and an increased risk of cancer for all sites has been discussed since Dyer and colleagues first suggested this in 1975.1 Several cohort studies,2–9 although not all,10–13 have repeated this finding. The cancer sites with most evidence of an increased risk are the kidney14,15 and the endometrium,16,17 but association has also been suggested for other sites.1,4,7,8,13,18 The discussion about antihypertensive (AH) drugs and cancer has been controversial. Mainly diure-

Correspondence: Dr A Lindgren, Department of Public Health and General Practice, University of Kuopio, PB 1627, 70211 Kuopio, Finland. E-mail: [email protected] This work has not been published elsewhere in whole or in part. There are no potential conflicts of interest. The research was funded with a grant from the Finnish Cultural Foundation. Received 2 September 2004; revised 16 December 2004; accepted 17 December 2004; published online 10 February 2005

tics,19,20 but also calcium channel blockers21 have been blamed for increasing the cancer risk, while ACE inhibitors have been suggested even to protect against cancer.22 Nevertheless, the evidence that these drugs directly influence the cancer risk is inconclusive. In this paper, we report the cancer incidence among a large cohort of hypertensive patients registered in the community-based hypertension register of North Karelia project. For cancer sites with an increased or decreased incidence among the hypertensives, we examined how blood pressure levels and the use of antihypertensive drugs predict the cancer incidence.

Materials and methods The study cohort consists of 20 886 hypertensive patients included in the community-based hypertension register of the North Karelia project. The details of the register are described elsewhere.23 This

Cancer pattern among hypertensive patients AM Lindgren et al 374

province-wide register was established in 1972, and new cases were registered up until 1988. The register was run within the primary health-care system in North Karelia as part of the North Karelia Project.24 All subjects either who were on antihypertensive drug treatment or who had elevated blood pressure in three subsequent measurements were registered. The blood pressure criteria were: 150 and/or 90 mmHg (o29 years), 160 and/or 95 mmHg (30–64 years) and 170 and/or 95 mmHg (465 years). Blood pressure measurements were performed by local nurses and physicians who had received relevant training. Blood pressure was measured in a local clinic from the right arm in a sitting position after a 5–10 min rest. Diastolic blood pressure (DBP) was recorded as the fifth phase of Korotkoff sounds. The local physician, following the common protocol, completed the hypertension register record forms after medical examination of the patient. All residents of Finland since 1 January 1967 have a unique personal identifier that is used in all main registers in Finland. The cohort was linked with the Population Register of Finland to check the validity of the personal identifiers and to obtain the dates of death, emigrations, and the number of children for the female subjects. In all, 36 persons (0.1%) were excluded from the cohort because they could not be identified. The record linkage with the cancer registry has been described earlier.25,26 Briefly, all incident cases of cancer were identified through the populationbased country-wide Finnish Cancer Registry. This was carried out automatically using the personal identifier as key. The follow-up started at the date when the patient was included into the hypertension register, and ended at death, emigration or on 31 December 1996, whichever occurred first. In all, 321 patients with diagnosed cancer before the beginning of the follow-up were excluded. After these patients were excluded, 12 621 women and 7908 men remained in the cohort. The numbers of observed cases and person-years at risk were calculated by 5-year age groups and by 5-year calendar periods. The expected numbers of cases for total cancer and for specific cancer types were calculated by multiplying the number of person-years in each age group by the corresponding cancer incidence rate in Eastern Finland during the period of observation. To calculate the standardised incidence ratio (SIR), the observed number of cases was divided by the expected number. The 95% confidence intervals (CI) for the SIR were based on the assumption that the number of observed cases followed a Poisson distribution. The cancer sites with SIRs significantly different from unity were selected for further analysis. The Cox regression model was used to analyse the independent effects of selected variables within the cohort on the incidence of these cancers. In this analysis, the end point was the person’s first Journal of Human Hypertension

diagnosis of cancer of interest. Those who died without cancer during the follow-up were censored at death. The best-fitting multivariate model for each cancer of interest was developed using a backward stepwise method. Only variables that significantly improved the model were included in the final model. The significance level 0.05 was used for exclusion of variables. The following variables were initially included in the models: age, year of registration, DBP and systolic blood pressure (SBP) as continuous variables, smoking (yes/no), body mass index (BMI) quartiles (the cutoff points for men were 24, 27 and 29 kg/m2, and for women 25, 28 and 32 kg/m2), use of antihypertensive drugs at baseline, functional diagnosis of hypertension and number of children for women. The functional diagnosis of hypertension was based on WHO recommendations.27 Stage I denoted elevated blood pressure without evidence of organic changes in the cardiovascular system. Stage II denoted high blood pressure with left ventricular hypertrophy but without evidence of other organ damage, and stage III denoted high blood pressure with other organ damage attributable to hypertension. Antihypertensive drug treatment indicated the use of any blood pressure lowering drug at least five times a week during the preceding 3 weeks. Smoking was coded as number of cigarettes smoked per day, but since the number of female smokers was so small, the distinction yes/no was used in most analyses for women.

Results The mean age of hypertensive women at the beginning of the follow-up was 58 years (s.d. 12) and of men 51 years (s.d. 13). Of the patients, 69% had antihypertensive drug treatment at the time of registration, and 31% received nonpharmacological treatment alone or were untreated. The mean length of follow-up for a person was 16 years (s.d. 7). During the follow-up period, 2691 cases of cancer were diagnosed; the expected number was 2700 (Table 1). The incidence of kidney cancer was significantly higher in hypertensive patients than in the general population. This was more evident in women (SIR 1.43, 95% CI 1.12– 1.80) than in men (SIR 1.20, 95% CI 0.86–1.63). Also the incidence of liver cancer was elevated, although this did not reach the statistical significance. In women, the incidence of endometrial cancer (corpus uteri) was significantly increased. The sites with markedly different SIRs in men and women were lung cancer and pancreatic cancer. The incidence of lung cancer was significantly reduced in women (SIR 0.55, 95% CI 0.36–0.78), and only slightly reduced in men (SIR 0.91 95% CI 0.81–1.02). In pancreatic cancer, the incidence was higher than expected in hypertensive women, hazard ratios (HR)


1.26 (95% CI 1.02–1.54), whereas in hypertensive men the incidence of pancreatic cancer was 0.73 (95% CI 0.49–1.04). For other sites, the incidence rates were rather similar for both hypertensive men and women (not shown). The sites that showed increased or decreased incidence among hypertensive patients were selected for further analysis. For kidney cancer in women, the best-fitting Cox regression model included age, registration year, use of antihypertensive drugs and BMI (Table 2). The HR for use of

Table 1 Observed (Obs) and expected (Exp) numbers of cancer cases by site, with SIRs and 95% CI, diagnosed during 1972–1996 among patients included in the hypertension register of the North Karelia project Primary site

Obs

Exp

SIR

95% CI

All sites Oesophagus Stomach Colon Rectum Liver Gall-bladder Pancreas Lung, bronchus Breast Cervix uteri Corpus uteri Ovary Prostate Kidney Bladder Skin melanoma Other skina Nervous system NHL Leukaemia Multiple myeloma

2691 59 201 148 99 44 54 123 305 307 29 117 70 175 113 113 52 111 65 55 61 48

2700.5 47.3 197.7 153.5 117.0 32.2 52.1 114.7 354.5 328.7 34.4 96.0 86.3 171.5 84.2 103.2 53.4 100.7 75.5 58.8 66.2 41.6

1.00 1.25 1.02 0.96 0.85 1.36 1.04 1.07 0.86 0.94 0.84 1.22 0.81 1.02 1.34 1.10 0.97 1.10 0.86 0.93 0.92 1.15

0.96–1.03 0.95–1.61 0.88–1.16 0.82–1.12 0.69–1.03 0.99–1.82 0.78–1.35 0.89–1.27 0.77–0.95 0.84–1.04 0.57–1.21 1.01–1.44 0.63–1.02 0.87–1.17 1.11–1.60 0.90–1.30 0.73–1.27 0.91–1.31 0.66–1.09 0.70–1.21 0.71–1.18 0.85–1.52

antihypertensive drugs in this model was 1.89 (95% CI 0.96–3.75). To minimise the possible bias that results from kidney cancer causing the hypertension, the analysis was repeated after excluding subjects with kidney cancer diagnosed within 5 years after the beginning of the follow-up. The HR for the use of antihypertensive drugs in this model was 1.79 (95% CI 0.91–3.87). Figure 1 shows the Kaplan–Meier curves for kidney cancer in women using antihypertensive drugs and not using antihypertensive drugs at baseline. There was no association between blood pressure levels and kidney cancer incidence either in women using antihypertensive drugs or in women not using the antihypertensive drugs at baseline. In the Cox regression model for men, only age remained as a significant factor predicting kidney cancer incidence. The increased incidence of liver cancer among our hypertensive population was not associated to blood pressure levels or use of antihypertensive drugs. Only age in men, and age and BMI in women, were significant predictors of liver cancer incidence. For endometrial cancer (Table 3), the final model included age, registration year and BMI (significant increase in the highest quartile). The possible interaction between BMI and blood pressure was also tested; neither DBP nor SBP was associated with endometrial cancer risk in any BMI category (P for interaction between BMI and DBP is 0.413 and that between BMI and SBP is 0.792). For pancreatic cancer in women, smoking had a strong independent effect. However, the use of

0.012

0.010

a

Excludes basal cell carcinoma. NHL ¼ Non-Hodgkin’s lymphoma.

0.008

hazard ratios (HR), 95% CI and number of cancer cases in each category (n), as diagnosed during 1972–1996 among patients included in the hypertension register of the North Karelia project Variable

n

HR

95% CI

Age (per year) Year of registration in the hypertension register Antihypertensive drugs at baseline No Yes Smoking at baseline No Yes Body mass index at baseline 25.3 kg/m2 or under 25.3–28.2 kg/m2 28.2–31.6 kg/m2 31.6 kg/m2 or more

66 66

1.02 1.08

1.00–1.05 1.01–1.16

10 56

1 1.89

0.96–3.75

60 6

1 1.97

0.83–4.65

13 14 18 21

1 1.06 1.35 1.63

0.50–2.25 0.66–2.78 0.81–3.27

0.006 Rate

Table 2 The model best fit for kidney cancer in women, with

0.004

0.002

0.000 − 0.002 −10

0

10

20

30

Time (years) Figure 1 Kaplan–Meier curves for kidney cancer incidence in women using antihypertensive drugs at baseline (upper curve) and not using antihypertensive drugs at baseline (lower curve), controlling for age, year of registration to hypertension register, smoking and BMI, among patients included in the hypertension register of the North Karelia project. Journal of Human Hypertension


Table 3 The model best fit for endometrial cancer in women, with HR, 95% CI and number of cancer cases in each category (n), diagnosed during 1972–1996 among patients included in the hypertension register of the North Karelia project Variable

n

Age (per year) 110 Year of registration in the hypertension 110 register Body mass index at baseline 17 25.3 kg/m2 or under 20 25.3–28.2 kg/m2 2 32 28.2–31.6 kg/m 41 31.6 kg/m2 or more

HR

95% CI

1.02 1.21

1.00–1.04 1.12–1.31

1 1.15 1.78 2.57

0.60–2.21 0.98–3.23 1.45–4.54

Table 4 The model best fit for pancreas cancer in women, with HR, 95% CI and number of cancer cases in each category (n), diagnosed during 1972–1996 among patients included in the hypertension register of North Karelia project Variable

n

HR

95% CI

Age (per year) Antihypertensive drugs at baseline No Yes Smoking at baseline No Yes

91

1.09

1.06–1.12

13 78

1 1.78

0.99–3.22

82 9

1 3.27

1.62–6.60

cancer,1,8 liver cancer13 and cancers of gastrointestinal tract.1,5,18 As in our study, a recent study among Swedish hypertensive patients found theory overall risk of cancer to be the same as that in the general population.28 In this study, the incidence of kidney cancer was also increased in both sexes (RR 1.4), but this was statistically nonsignificant (95% CI 0.85– 2.1). They found no significant increase or decrease in the incidence of cancers of the endometrium, pancreas, lung, liver or any other site. It is evident that hypertensive patients can differ from the general population in aspects other than blood pressure levels alone. One important factor is obesity, which is known to be associated with hypertension.29 Also in our cohort, BMI was higher in hypertensive patients than in the Finnish general population.30 In the cross-sectional survey in 1972 in Eastern Finland, the mean BMI among men aged 30–59 years was 26.0 kg/m2 and among women 26.8 kg/m2, whereas in our cohort it was 27.6 and 29.6 kg/m2, respectively. On the other hand, the prevalence of smoking was lower among the hypertensive patients in our cohort than in the general population, as discussed below.

Kidney cancer

antihypertensive drugs at baseline (HR 1.78 (95% CI 0.99–3.22)) also remained in the model (Table 4). This was not the case in men, where only age of the variables included in the study remained in the model to predict the incidence of pancreatic cancer. The lower incidence of lung cancer among hypertensive women was not associated with blood pressure-related variables. The best-fitting model for lung cancer in women included age and smoking (HR for smokers of one to 10 cigarettes per day 3.08 (95% CI 0.71–13.4) and more than 10 cigarettes per day 17.8 (95% CI 5.13–61.8), compared to nonsmokers).

Discussion In our large and population-based cohort of hypertensive patients, the overall cancer incidence did not differ from that of the general population. The incidence of cancers of the kidney and liver, and in women cancers of endometrium and pancreas, elevated, whereas the incidence of lung cancer was lower compared to the general population. Previously, elevated blood pressure has been found to be associated with a moderately increased risk of cancer incidence or mortality in various studies,2–9 although reports of no association also exist.10–12 Apart from cancers of the kidney and endometrium, other sites with a suggested relationship to hypertension include breast cancer,17 lung Journal of Human Hypertension

The body of evidence associating hypertension with increased incidence of kidney cancer, in both men and women, is convincing.14,15,31–34 Various studies have also reported association with the use of diuretics and kidney cancer risk.18,31,35 The question remains, however, whether both of these are independent risk factors or if one merely reflects the effect of another. A theory was recently proposed that lipid peroxidation would be the mechanism responsible for the increased risk of kidney cancer risk associated with both hypertension and obesity.36 In our population of hypertensive patients, in women, the incidence of kidney cancer was associated with the use of antihypertensive drugs at baseline and obesity, but not with blood pressure levels. There was also no association with a functional diagnosis of hypertension at baseline. This speaks in favour of the independent effect of the antihypertensive drugs. However, it cannot be ruled out that at least part of the detected effect of the antihypertensive drugs reflects the effect of some aspect of hypertension that we were not able to sufficiently measure in our study. These include the duration of hypertension and its severity, since the blood pressure measurements present the situation after the start of using antihypertensive drugs, and the functional diagnosis of hypertension, based on the presence of the end-organ damage, can be seen as a rather coarse measure of the severity of the hypertension.


Our results were altered only slightly when the analysis was repeated after the exclusion of subjects whose kidney cancer was diagnosed within 5 years after the beginning of the follow-up. Thus, it is unlikely that there is remarkable bias resulting from preclinical kidney cancer causing the hypertension. We did not have enough individual information about the history of the type of antihypertensive drugs used by the patients to distinguish between different antihypertensive drugs in our analyses. However, from the patients registered in 1978–1979 in this cohort, we know that the most common types of antihypertensive drugs used were diuretics (in 57% of women and 54% in men, either alone or in combination) and betablockers (54 and 43%, respectively).37 The Swedish study28 found no significant differences in the risk of kidney cancer, or cancer of any other site between the users of conventional antihypertensive drugs, ACE inhibitors and calcium antagonists. However, the numbers of cancer cases were rather small in their study. In hypertensive men in our cohort, the incidence of kidney cancer was 20% higher than that in the general population, even though this was not statistically significant. We also failed to show any association between blood pressure levels, use of antihypertensive drugs or functional diagnosis of hypertension and the subsequent risk of kidney cancer among hypertensive men. The interaction term between the sex and the use of antihypertensive drugs at baseline for the risk of kidney cancer was significant in our analyses. The reasons for this difference between sexes remain thus far unknown. The possibilities include hormonal factors and differences in the patterns of drug use (see above). At least in Finland, studies show that women are prescribed diuretics more frequently and beta-blockers and ACE inhibitors less frequently than men.38

Pancreatic cancer

Smoking and diabetes mellitus are the only wellconfirmed risk factors for pancreatic cancer, while recent studies have also suggested that obesity plays a role.39,40 We know two studies suggesting that elevated blood pressure might also be related to an increased risk of pancreatic cancer.41,42 However, in one of these studies, the effect disappeared after adjustment for obesity,41 and in another, the adjustment for weight or obesity was not performed.42 We are not aware of any previous study relating the treatment for blood pressure to the risk of pancreatic cancer. As discussed above in relation to kidney cancer, in our study population, the treatment for blood pressure can reflect various factors, like the severity or duration of hypertension, apart from the direct effect of the drugs themselves. As for kidney cancer, the reason for the difference

between sexes remains unclear. It is also possible that our results are due to chance. Endometrial cancer

The association between obesity and the risk of endometrial cancer is well documented, whereas the role of hypertension as an independent risk factor for endometrial cancer remains unclear. Various studies have shown an association between hypertension and the risk of endometrial cancer,16,17,43–46 but in some reports,43,44 although not all,16,17,45 the association has become nonsignificant after adjustment for obesity. Recent studies have found hypertension to be associated to the endometrial cancer risk only in obese women, supporting a theory that metabolic syndrome would be a key factor in the endometrial cancer risk.43,47 In our study, the incidence of endometrial cancer was increased among hypertensive patients, and obesity was an important predicting factor of the cancer risk inside the cohort. The interaction between obesity and blood pressure was not significant in our data. Liver cancer

Previously, Batty et al13 reported from the Whitehall study that in men both SBP and DBP were directly associated with the increased risk of liver cancer mortality and that this association was independent of obesity. In our study, the incidence of liver cancer was elevated, although nonsignificantly, among both hypertensive men and women. However, this seemed neither to be related to hypertension per se nor to the use of antihypertensive drugs, but rather to obesity. Recently, epidemiological studies have suggested obesity to be a risk factor for liver cancer.39,48 One possible confounding factor is alcohol use, which we could not control for in the present study. However, the association between hypertension and liver cancer has previously been shown to persist after adjustment for alcohol use.39 The mechanism of this association is proposed to be through the progression of nonalcoholic fatty liver disease.49 Lung cancer

The incidence of lung cancer was lower among hypertensive women of our cohort than in the general population, but was not related to blood pressure levels or the use of antihypertensive drugs inside the cohort. The decreased incidence of lung cancer merely reflects the smaller proportion of smokers among our cohort than in the general population. The cross-sectional survey in 1972 showed that the proportion of the current smokers among women of 30–59 years of age of North Karelia Journal of Human Hypertension


was 10%,24 while in our data, the prevalence of smoking among hypertensive women registered in 1972 at the age of 30–59 years, was 5%. Also, in hypertensive men, the incidence of lung cancer was slightly reduced, reflecting the reduced frequency of smoking among Finnish hypertensive patients.24 However, for men, we have earlier reported that there is actually a direct association between DBP and lung cancer risk, especially in smokers.26 General considerations

Our cohort consists only of patients with hypertension of different severity. A significant proportion of the general population or Eastern Finland, used as the reference, also has elevated blood pressure. Thus, the SIRs are underestimates of the effect of hypertension. As there were multiple cancer sites studied in the first phase of this study (Table 1), there is a possibility that some of the associations observed are due to chance. The possible confounding factors that we could not control in the present study include, for instance, nutritional factors, physical activity and social factors. However, these factors should not play major roles in the aetiology of cancers of the pancreas, kidney, liver or endometrium. The cancer registration system in Finland is virtually complete, and the computerised record linkage procedure is precise.50 Therefore, technical incompleteness did not cause bias in our results.

Conclusions The overall cancer incidence in hypertensive patients does not differ from that in the general population. Some cancer types, such as kidney cancer and endometrial cancer, are however more common among hypertensive patients. This may be at least partly due to obesity or the use of antihypertensive drugs, and not necessarily due to hypertension per se.

Acknowledgements The research was funded with a grant from the Finnish Cultural Foundation and the Finnish Cancer Foundation.

References 1 Dyer AR et al. High blood-pressure: a risk factor for cancer mortality? Lancet 1975; 1: 1051–1056. 2 Svardsudd K, Tibblin G. Mortality and morbidity during 13.5 years’ follow-up in relation to blood pressure. The study of men born in 1913. Acta Med Scand 1979; 205: 483–492. Journal of Human Hypertension

3 Raynor Jr WJ et al. High blood pressure and 17-year cancer mortality in the Western Electric Health Study. Am J Epidemiol 1981; 113: 371–377. 4 Khaw KT, Barrett-Connor E. Systolic blood pressure and cancer mortality in an elderly population. Am J Epidemiol 1984; 120: 550–558. 5 Goldbourt U et al. Elevated systolic blood pressure as a predictor of long-term cancer mortality: analysis by site and histologic subtype in 10,000 middle-aged and elderly men. J Natl Cancer Inst 1986; 77: 63–70. 6 Filipovsky J et al. Abdominal body mass distribution and elevated blood pressure are associated with increased risk of death from cardiovascular diseases and cancer in middle-aged men. The results of a 15- to 20-year follow-up in the Paris prospective study I. Int J Obes Relat Metab Disord 1993; 17: 197–203. 7 Rosengren A et al. Hypertension and long-term cancer incidence and mortality among Swedish men. J Hypertens 1998; 16: 933–940. 8 Wannamethee G, Shaper AG. Blood pressure and cancer in middle-aged British men. Int J Epidemiol 1996; 25: 22–31. 9 Peeters P et al. Hypertension, antihypertensive drugs, and mortality from cancer among women. J Hypertens 1998; 16: 941–947. 10 Taylor JO et al. Blood pressure and mortality risk in the elderly. Am J Epidemiol 1991; 134: 489–501. 11 Clausen J, Jensen G. Blood pressure and mortality: an epidemiological survey with 10 years follow-up. J Hum Hypertens 1992; 6: 53–59. 12 Grove JS et al. The association of blood pressure with cancer incidence in a prospective study. Am J Epidemiol 1991; 134: 942–947. 13 Batty GD, Shipley MJ, Marmot MG, Davey SG, Whitehall Study. Blood pressure and site-specific cancer mortality: evidence from the original Whitehall study. Br J Cancer 2003; 89: 1243–1247. 14 Shapiro JA et al. Hypertension, antihypertensive medication use, and risk of renal cell carcinoma. Am J Epidemiol 1999; 149: 521–530. 15 Chow WH et al. Obesity, hypertension, and the risk of kidney cancer in men. N Engl J Med 2000; 343: 1305–1311. 16 Parazzini F et al. Population attributable risk for endometrial cancer in northern Italy. Eur J Cancer Clin Oncol 1989; 25: 1451–1456. 17 Soler M et al. Hypertension and hormone-related neoplasms in women. Hypertension 1999; 34: 320– 325. 18 Zhang ZF et al. Adenocarcinomas of the esophagus and gastric cardia: medical conditions, tobacco, alcohol, and socioeconomic factors. Cancer Epidemiol Biomarkers Prev 1996; 5: 761–768. 19 Grossman E, Messerli FH, Goldbourt U. Does diuretic therapy increase the risk of renal cell carcinoma? Am J Cardiol 1999; 83: 1090–1093. 20 Lindblad P et al. Risk of kidney cancer among patients using analgesics and diuretics: a population-based cohort study. Int J Cancer 1993; 55: 5–9. 21 Pahor M et al. Do calcium channel blockers increase the risk of cancer? Am J Hypertens 1996; 9: 695–699. 22 Lever AF et al. Do inhibitors of angiotensin-I-converting enzyme protect against risk of cancer? Lancet 1998; 352: 179–184. 23 Nissinen A, Tuomilehto J, Puska P. The hypertension register of the North Karelia project. Clin Sci Mol Med Suppl 1978; 4: 355s–358s.


24 Puska P et al. The North Karelia Project: 20 Years’ Results and Experience. Helsinki, Finland: National Public Health Institute, 1995. 25 Lindgren A, et al. Cancer incidence in hypertensive patients in North Karelia, Finland. Hypertension 2001; 37: 1251–1255. 26 Lindgren A et al. Blood pressure, smoking and the incidence of lung cancer in hypertensive men in North Karelia, Finland. Am J Epidemiol 2003; 158: 442–447. 27 World Health Organization. Arterial Hypertension: Report of a WHO Expert Committee. Geneva, Switzerland: World Health Organization, Technical Report Series 628, 1998. 28 Lindholm LH et al. Relation between drug treatment and cancer in hypertensives in the Swedish Trial in Old Patients with Hypertension 2: a 5-year, prospective, randomised, controlled trial. Lancet 2001; 358: 539–544. 29 Dyer AR, Elliott P. The INTERSALT study: relations of body mass index to blood pressure. INTERSALT Cooperative Research Group. J Hum Hypertens 1989; 3: 299–308. 30 Jalkanen L et al. Changes in body mass index in a Finnish population between 1972 and 1982. J Int Med 1989; 226: 163–170. 31 Yuan JM et al. Hypertension, obesity and their medications in relation to renal cell carcinoma. Br J Cancer 1998; 77: 1508–1513. 32 McLaughlin JK et al. International renal-cell cancer study. VIII. Role of diuretics, other anti-hypertensive medications and hypertension. Int J Cancer 1995; 63: 216–221. 33 Mellemgaard A et al. Risk factors for renal cell carcinoma in Denmark: role of medication and medical history. Int J Epidemiol 1994; 23: 923–930. 34 Muscat JE, Hoffmann D, Wynder EL. The epidemiology of renal cell carcinoma. A second look. Cancer 1995; 75: 2552–2557. 35 Mellemgaard A, Moller H, Olsen JH. Diuretics may increase risk of renal cell carcinoma. Cancer Causes Control 1992; 3: 309–312. 36 Gago-Dominguez M et al. Lipid peroxidation: a novel and unifying concept of the etiology of renal cell carcinoma (United States). Cancer Causes Control 2002; 13: 287–293.

37 Nissinen A et al. Drug treatment of high blood pressure in the community—experience in eastern Finland. J Hum Hypertens 1989; 3: 165–171. 38 Wallenius S, Peura S, Klaukka T, Enlund H. Who is using antihypertensive drugs? A prescription analysis from Finland. Scand J Prim Health Care 1996; 14: 54–61. 39 Calle EE et al. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of US adults. N Engl J Med 2003; 348: 1625–1638. 40 Berrington de Gonzalez A, Sweetland S, Spencer E. A meta-analysis of obesity and the risk of pancreatic cancer. Br J Cancer 2003; 89: 519–523. 41 Friedman GD, van den Eeden SK. Risk factors for pancreatic cancer: an exploratory study. Int J Epidemiol 1993; 22: 30–37. 42 Stolzenber-Solomon RZ et al. A prospective study of medical conditions, anthropometry, physical activity, and pancreatic cancer in male smokers (Finland). Cancer Causes Control 2002; 13: 417–426. 43 Weiderpass E et al. Body size in different periods of life, diabetes mellitus, hypertension, and risk of postmenopausal endometrial cancer (Sweden). Cancer Causes Control 2000; 11: 185–192. 44 Brinton LA et al. Ethnicity and variation in breast cancer incidence. Int J Cancer 1997; 73: 349–355. 45 Inoue M et al. A case–control study on risk factors for uterine endometrial cancer in Japan. Jpn J Cancer Res 1994; 85: 346–350. 46 Parazzini F et al. The epidemiology of endometrial cancer. Gynecol Oncol 1991; 41: 1–16. 47 Furberg AS, Thune I. Metabolic abnormalities (hypertension, hyperglycemia and overweight), lifestyle (high energy intake and physical inactivity) and endometrial cancer risk in a Norwegian cohort. Int J Cancer 2003; 104: 669–676. 48 Moller H, Mellemgaard A, Lindvig K, Olsen JH. Obesity and cancer risk: a Danish record-linkage study. Eur J Cancer 1994; 30A: 344–350. 49 Caldwell SH, Crespo DM, Kang HS, Al-Osaimi AM. Obesity and hepatocellular carcinoma. Gastroenterology 2004; 127(5 Suppl 1): S97–S103. 50 Teppo L, Pukkala E, Lehtonen M. Data quality and quality control of a population-based cancer registry. Experience in Finland. Acta Oncol 1996; 33: 365–369.

Journal of Human Hypertension