Our study found a reduced risk of bladder cancer among hypertensive subjects who did not use antihypertensives or diuretics regularly and this reduction in risk was limited to smokers and carriers of the GSTM1
-null genotype. Consistent with these results, untreated hypertensive control subjects exhibited a lower level of hemoglobin adducts of 4-ABP, a known human bladder carcinogen present in cigarette smoking (30
), and this reduction was limited to smokers with the GSTM1
The present study reports a thorough investigation of the effects of hypertension and use of antihypertensive drugs on bladder cancer risk. Very few studies have examined the associations between hypertension, antihypertensive medications and risk of bladder cancer. Hypertension was not associated with bladder cancer in a case–control study conducted among nonsmokers in the USA (OR not reported) (32
) and in the Iowa Women’s Health Study among postmenopausal women (OR: 0.89; 95% CI: 0.60–1.31) (33
). Similarly, no significant association with bladder cancer risk was found for history of hypertension in a Korean case–control study (OR: 1.63; 95 CI%: 0.79–3.34) (34
). Interestingly, a case–control study conducted in Japan found a significantly lower risk for bladder cancer in men who had a past history or complications of hypertension (35
). As for antihypertensive drugs, an elevated risk of bladder cancer was observed in a Danish cohort among men who received diltiazem (a class of calcium channel blocker used in the treatment of hypertension) exclusively (standardized incidence ratio = 2.1; 95% CI: 1.2–3.4) or combined with other calcium channel blockers (36
The precise mechanism whereby hypertension reduces the risk of bladder cancer is unclear. We speculate that one mechanism could be related to the oxidative stress-induced apoptosis pathway, although the underlying biological basis of this pathway as an anticancer mechanism is still unclear and controversial. Damage to DNA by ROS has been widely accepted as a major cause of cancer (7
). However, evidence is also accumulating that ROS are essential mediators of apoptosis, which eliminates precancerous and cancerous, virus-infected and otherwise damaged cells (37
). Thus, it seems that ROS may have divergent cellular effects depending on their origin, extent of their production and the enzymatic or nonenzymatic mechanisms available for their dismutation in a given cell type. It has been shown that untreated hypertensive subjects have significantly increased levels of ROS (6
), and this increase is reversible upon treatment with antihypertensive medications (45
). In bladder cancer, the generation of ROS and oxidative stress have been proposed as the major mechanism for the induction of apoptosis and inhibition of cancer growth by the cancer chemopreventive or chemotherapeutic agents N
-(4-hydroxyphenyl) retinamide (46
), isothiocyanates (37
), arsenic trioxide (48
), interferon-γ (49
) and cis-dichlorodiammineplatinum (52
). In light of this evidence, it is plausible that increased ROS from untreated hypertension may induce apoptosis of bladder cancer cells and thus reduce bladder cancer risk. However, as expressed above, despite these experimental data, the underlying biological basis of this pathway as an anticancer mechanism is still uncertain.
Because glutathione-associated metabolism is a major mechanism against agents that generate oxidative stress by eliminating ROS, we hypothesized that individuals possessing the low activity genotypes of antioxidant GSTM1, GSTT1 and or GSTP1 (i.e. the GSTM1
null and GSTP1
AB/BB genotypes, respectively) may exhibit a stronger untreated hypertension–bladder cancer inverse association than those possessing the high-activity genotypes (53
). This hypothesis was supported by our results that the reduction in bladder cancer risk from untreated hypertension was mostly confined to carriers of the GSTM1
Our observation of a lower level of 4-ABP-Hb adducts among untreated hypertensive healthy subjects may indicate a possible connection between hypertension and metabolism of 4-ABP. Arylamines, such as 4-ABP, are present in tobacco smoke and are the leading putative constituents responsible for bladder cancer development in smokers (54
). One could hypothesize that if the mechanism of protection involves reducing exposure to cigarette-derived carcinogens, smokers would have benefited more from untreated hypertension than nonsmokers. Consistent with this hypothesis, our study found that untreated hypertension appeared most protective among smokers. However, no direct evidence for this proposed mechanism is currently available. In addition, the interpretation of the 4-ABP-Hb adducts results needs to be cautious since the blood samples used to measure the 4-ABP-Hb adducts were collected at the time of study recruitment, whereas the hypertension and treatment status were lifetime histories.
It is also possible that factors associated with the decision of treatment for hypertensive patients may explain the association between untreated hypertension and bladder cancer risk. In our study, untreated hypertensive subjects were relatively young with a mean age of 56 years. To achieve blood pressure control, these individuals may have adopted healthy lifestyles, such as eating more fruits and vegetables and engaging in regular aerobic physical activity (55
), instead of taking antihypertensive medications. Information on physical activity was not collected in our study. Nonetheless, physical activity is not an established risk/protective factor for bladder cancer with most studies showing a lack of association (56
). We did collect information on dietary intake of fruits and vegetables and found that dietary intake of carotenoids was associated with reduced risk of bladder (27
). Similarly, we previously reported an inverse association between alcohol intake and bladder cancer (25
). In our study, hypertensive control subjects, especially those who did not regularly use diuretics or antihypertensives, reported a slightly higher dietary intake of carotenoids and a significantly higher consumption of alcoholic beverages than control subjects without hypertension, therefore it is possible that the association between hypertension and bladder cancer was confounded by any of these two factors. However, the hypertension–bladder cancer association remained unchanged after we adjusted our analyses for dietary intake of carotenoids and alcohol consumption, and the untreated hypertension-associated reduction in risk was also observed among subjects with low intake of carotenoids and alcohol nondrinkers. It is possible that people who were not treated for hypertension may have been different in many aspects from people who were, including in hypertension itself. According to the Third National Health and Nutrition Examination Survey (NHANES III), in the 45- to 64-year-old age group (to which majority of our study subjects belonged), untreated hypertensive patients had a mean blood pressure of 152/89, with 54% having isolated systolic hypertensive (systolic blood pressure ≥140 mm Hg and diastolic blood pressure <90 mm Hg) (57
). For a long time, guidelines, clinical trials and clinical practice on blood pressure control have placed greater importance on diastolic blood pressure levels (58
) (the first Joint National Committee report in 1977 (59
) defined diastolic blood pressure, not systolic blood pressure, as the basis for detection and treatment), so it is possible that an individual with isolated systolic hypertension may have not been eligible for treatment under the old guidelines. Consistent with this notion, studies that documented physicians’ behavior have also confirmed that physicians are unlikely to diagnose isolated systolic hypertension as hypertension or to treat this condition aggressively (57
). We did not directly measure the study participants’ blood pressure and hence were unable to quantitatively measure the effect of hypertension by its severity and different types, so it is still possible that confounding by indication may have risen due to different prognostic factors that have influenced treatment decisions.
Our study has limitations. One important limitation is the relatively small sample size of untreated hypertensive subjects (104 cases and 150 controls), particularly in the subgroup analyses and 4-ABP-Hb analyses, increasing thus the likelihood of spurious associations and precluding firm conclusions. Large studies are needed to confirm these findings. In addition, the diagnosis of hypertension was self-reported and the retrospective nature of our case–control design did not allow us to obtain accurate and complete history of hypertension and use of medications over long periods, leading to potential misclassification. However, we believe the misclassification is non-differential for cases and controls, therefore unlikely to explain our findings. Hypertension is often associated with other medical conditions such as central obesity, diabetes mellitus and hyperlipidemia. Some of these conditions (56
) and their associated medications such as statin (61
) have been associated with bladder cancer risk in some studies. Our study did not collect detailed information on all these conditions and their associated medications; therefore, it is possible that our observed associations may be confounded by these factors. However, we did obtain some information on diabetes and BMI and these two factors did not seem to affect the inverse association between untreated hypertension and bladder cancer risk. Untreated or poorly treated hypertension is associated with increased mortality and stroke (62
), indicating the possibility that untreated hypertensive subjects may be more probably to die from competing risks and therefore less probably to live long enough to develop bladder cancer. However, such possibility is most probably to be low since the NHANES III found that subjects with acknowledged untreated hypertension were more probably to be young male current-smokers (average age: 55.3 years old), similar to the findings from our study (average age among controls: 56.1 years old), and majority of them were subjects who had access to medical care (57
). In addition, our observation of a similar effect of untreated hypertension for bladder cancer diagnosed at different stages indicates that survival bias is unlikely. Finally, there is also the possibility that the GSTM1 genotyping method that we used, which did not distinguish between homozygous wild-type +/+ and heterozygous +/− individuals, may have led to an underestimation of the genetic effect.
In summary, we observed a statistically significant inverse association between untreated hypertension and bladder cancer risk. Future studies are needed to confirm our findings and explore possible mechanisms.