Cancer which Location is Safest, Urban or Rural? Based on a USA Nationwide cohort study. By Dr Chris Barnes, firstname.lastname@example.org
US state cancer profiles have been used in a statistical and geographic meta-analysis study to answer the question ‘which is safest for cancer: urban or rural?’ Urban living is shown to be safer in all but breast, prostate and kidney cancer. For all types of cancer considered, with the exception of kidney cancer, breast cancer and prostate cancer, living in the countryside or a rural area in the USA would appear to present about 1.3 -1.6 times the risk of living in the most urbanised areas. My examination of the data leads me to make the radical suggestion that possibly over 60% of present day breast cancer ( and to some extent prostate as well) could be avoided by sleeping in total darkness or having sufficient melatonin! Geographic spatial meta-analysis by means of simple visual inspection tends to suggest that the risk may be greatest for pancreatic cancer with glyphosphate. Occupational risk has been noted by others but this work may be the first which shows risk to the public at large. Similarly NHL risk for 2,4D also seems to cross –related in a similar manner according to this study. There were insufficient degrees of freedom available in the 56% sunshine data to prove statistical relevance to the dangers of wireless technology. However both wireless and night light are thought to act on the field sensitive pineal gland in reducing melatonin concentration. My study has shown overwhelming proof of an association between nightlight and breast and prostate cancer.
Many kinds of human cancer have proliferated alarmingly in recent years. There are many known chemical carcinogens but few of us are heavily exposed to these. Ionising radiation will vary from region to region due to radon gas and medical exposure but is not particularly on the increase except perhaps in certain regions after Chernobyl or the Japanese Fukushima disaster.
It is logical, then, to seek which other main environmental common denominator(s) to which we are all exposed.
A recent study based on stem cell division/differentiation has suggested up to 2/3rds of cancer may be down to chance mutations. An independent way of testing this is to look at geographic spread in incidence across a huge cohort and further to compare the maximum and minimum values of the said incidence and its data range. Further if for a given type of cancer a firm association with a causative agent can be shown with a high regression factor than possibly the rest is down to chance. Alternatively, there can be more than one carcinogen and/or promoter at work and then less is due to ‘chance’. We must understand, however, that the easiest way for mutations to occur is by impact of high energy radiation photons but chemical agents such as cigarette smoke or asbestos which release free radicals can also do the trick. There are also a whole host of possible mechanisms where radiofrequency radiation could either release free radicals at the gas liquid interface  or perturb the 3D molecular motion of a dividing/differentiating cellular system .
The manifold types of cancer afflicting human kind have had causes ascribed to chemical carcinogens, especially tobacco smoke, to radiation and to disease and inflammation.
Populations cramped into cities are more likely to transmit disease, there is often more roadside and air pollution in city and industrialised regions, there is more light and noise pollution and possibly more instances of electromagnetic pollution in cities thus a possible hypothesis is that city areas might spawn a higher cancer incidence. On the other hand people in rural areas are more exposed to potential chemical carcinogenic action of various pesticides. Also in rural areas, electricity distributing conductors are more likely to be overhead and these have also been associated with the concentration of radon gas. The purpose of this short publication is to test which is the dominant hypothesis.
I have previously considered RF radiation as a new and increasing environmental factor. I have reached the conclusion that RF radiation may potentially be a co-promoter of cancer in conjunction with environmental nano-pollution [3,4] . I have also explained how it can be particularly hazardous for those with faulty p53 genes and further how it is a source of oxidative stress in bio-systems in general. This would suggest it ought to be more hazardous living in urban areas, towns and cities but in this present work I pose the question is this case for all types of cancer.
I have also suggested that city living could be hazardous with respect to cancer due to increases in night noise and light here in the UK . In this present work is the opportunity to test this hypothesis by using the huge U.S. Incidence cohort. Because wireless technology penetration is so saturated in the USA I can explore states where it can be classed as 100% saturated and would not be expected to influence night light results. Indeed even if I were to obtain a negative result i.e. trend towards higher than average cancer rates in rural areas, others have explained this in terms of handset APO ( autopower out adjustment), see Kundi (2009) .
I have not previously considered the epidemiology of cancers and pesticides although others have. I shall make such an investigation in this present work as a result of recent disclosures by others on the severe dangers of the common household and agricultural pesticide known gyphosphate, also know by the weed-killer trade name ‘round-up’.
I have recently considered the effects of sunshine elsewhere, see http://www.drchrisbarnes.co.uk/SUNGOOD.htm .
With regard to rural or urban living there is sufficient demographic detail available to determine which is safest for a number of different cancers. The expectation is expected to be that for cancers which have been reported to show improvement on melatonin administration a positive correlation with urban or city living ought to be seen due to increased night light pollution. I have previously referred to such cancers as ‘RF or field sensitive cancers’. Whereas for cancers which have been described as associated with farming and agriculture, possibly pesticides a negative correlation with urbanisation should be seen.
Experimental and Data Sources
The data pertaining to US cancers is all available at http://statecancerprofiles.cancer.gov/ .
with the exception of total incidence for the state of Nevada which was obtained from http://www.foxnews.com/health/2012/10/29/nevada-has-higher-cancer-rates-than-neighboring-states/ .
The data relating to degree of urbanisation was taken from http://en.wikipedia.org/wiki/Urbanization_in_the_United_States .
The odds ratio is about 1.208 stacked in favour of cancer risk been worse in rural areas. But the result is not statistically significant with R=.19 and P =.168.
A correction can be made for sunshine by regressing against % sunshine and then using the residuals. This brings about a result which is classed as statistically significant.
If this is done, the regression factor is increased to .27 so the odds ratio is stacked slightly higher at about 1.31 in favour of getting any cancer in a the most rural area. Here the P Value Results
r=.27 DF=52 yields a two-tailed P value equals 0.0483
By conventional criteria, this difference is considered to be statistically significant.
All cancers tested showed a risk factor which was stacked worse towards countryside living, with the exception of breast and prostate. One way of assessing the behaviour was to choose states with equal sunshine and regress each specific cancer incidence against degree of urbanisation. A constant annual sunshine rate of 56% was chosen.
Odds ratio is 1.52 in favour of most rural location. The problem with only considering states with 56% sunshine is that for brain cancer it renders the result not statistically significant.
Odds ratio is 1.936 in favour of most rural location. Colon cancer is a different matter, even with only 6 degrees of freedom P=0.074 which is almost significant.
Odds ratio is 1.607 in favour of most rural area.
Odds ratio is 1.487 stacked towards most rural location.
Odds ratio is 2.009 in favour of most urban location. P=.05 i.e. just statistically significant.
Odds ratio is 1.687 in favour of most urban location.
Odds ratio is 1.046 i.e. a very weak urban risk.
Other risk factors
One of the biggest risks of living in the countryside is reckoned to be agriculture, ESPECIALLY pesticides, SEE FOR EXAMPLE , ‘The Ecologist’ . The results of my study certainly appear to confirm this viewpoint. Glyphosphate is one of the most commonly used herbicide in agriculture and as the domestic weed killer ‘roundup’, see Pesticide.org .
Odds ratio is 1.75. Any cancer has 1.75 more risk in a an area with high Glyphosphate application than one with none. P Value Results
r=.5507 DF=15 The two-tailed P value equals 0.0220. By conventional criteria, this difference is considered to be statistically significant. Of course correlation does not necessarily imply causation and there are generally many other pesticides applied in the same regions as is glyphosphate so we could be looking at synergistic effects?
However, after correction for annual sunlight the O/R for gyphosphate increases to 1.89.
A very similar O/R is obtained for the other most popular pesticide which is Atrazine with an O/R =1.894. P=.0259. So perhaps one or other ro both of these pesticides working as promoter and co-promoter are responsible?
Pesticide geographic distribution
One way of elucidating the above is to look at a method of geographic meta –analysis.
Indeed, the above highly statistically relevant result has led me to consider a geographic mapping visual meta-analysis to see if I can tie in any specific cancers with any specific pesticides.
After searching the geographic distribution of a large range of different types of cancers, I noted that one stood out. I see what looks like a plausible association between glyphosphate and cancer of the Pancreas.
Alguacil et al (2000)  have shown significant OR’s for occupational exposure and glyphosphate but I believe my findings may be the first for exposure of the general public. This could, potentially, be a highly significant finding for Pancreatic Cancer is one of the least treatable and one that some groups have recently been ascribing to simple genetic bad luck.
Atrazine is used in virtually the same geographic areas to glyphosphate.
Atrazine Pancreatic Colon
In this case no one cancer correlates perfectly but the two which stand out as having some sort of association by visual mapping analysis are pancreatic and colon cancer.
Rusiecki et al (2004)  reach a similar conclusion in respect that their analyses did not find any clear associations between atrazine exposure and any cancer analysed but did show a suggestion of trend (lung, bladder, non-Hodgkin lymphoma, and multiple myeloma). My conclusion is that with regard to atrazine it could be therefore that other co-applied pesticides are to blame.
2,4 D is again used in more or less the same geographic locations as glyphosphate and atrazine. Upon visual inspection a reasonable general association with all cancer incidence combined can be seen. Similarly perhaps more of a visual correlation with NHL can be seen. Zahm et al (1990)  has noticed a significant occupational risk for NHL with 2,4D applicators, but once again I believe my present study may be a first to reveal an association in the public at large.
Radio Frequency Penetration
O/R =1.28 100% Wireless penetration/15% wireless. There is suggestion of a trend, but there are insufficient degrees of freedom here to confirm absolute statistical significance.
O/R =1.376 Once again there is suggestion of a positively associated trend, but there are insufficient degrees of freedom here to confirm absolute statistical significance.
Nightlight (Pure Urbanisation) Breast/Prostate Cancer at constant RF penetration level.
R=.955 O/R =5.52. P Value < 0.0001
The two-tailed P value is less than 0.0001. By all conventional measurement criteria, this difference is considered to be extremely statistically significant.
This enormous level of significance is also evident in the geographic visual meta-analysis, for breast cancer and night light see below:
My examination of the data leads me to make the radical suggestion that possibly over 60% of present day breast cancer ( and to some extent prostate as well) could be avoided by sleeping in total darkness or having sufficient melatonin!. Besides saving energy and carbon emissions surely this is all the more reason for councils to turn street lights off at night.
For all types of cancer considered with the exception of kidney cancer, breast cancer and prostate cancer, living in the countryside or a rural area in the USA would appear to present about 1.3 -1.6 times the risk of living in the most urbanised areas. This risk appears to be due to pesticides especially glyphosphate and 2,4D. Geographic spatial meta-analysis by means of simple visual inspection tends to suggest that the risk may be greatest for pancreatic cancer with glyphosphate. Occupational risk has been noted by others but this work may be the first which shows risk to the public at large. Similarly NHL risk for 2,4D also seems to cross –related in a similar manner according to this study.
There were insufficient degrees of freedom available in the 56% sunshine data to prove statistical relevance to the dangers of wireless technology. However both wireless and night light are thought to act on the field sensitive pineal gland in reducing melatonin concentration. My study has shown overwhelming proof of an association between nightlight and breast and prostate cancer. Interestingly, there are at least 50 other publications discussing the roles of artificial light, sleeping patterns, night shifts and melatonin levels on cancer, especially those of breast and prostate but also to a lesser extent bowel. Further indeed, Mills et al (2005)  agree with me that Melatonin may have a great potential in treating cancer. There are even papers which discuss night light and kidney cancer namely many by Anisimov and other cancers Parent et al (2012)  even though the present study only showed a very weak association for kidney cancer as did that of Parent .
13. Alguacil et al, ‘Risk of Pancreatic Cancer and Occupational Exposures in Spain’ Annals of Occupational Hygiene, Vol 44 No. 5 , pp 391-403, 2000.
15. Zahm, S. H., Weisenburger, D. D.,Babbitt, P.A., Saal, R.C.,Vaught, J. B.,Cantor, K.P.,and Blair, A. (1990). A case-control study of non-Hodgkin's lymphoma and the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) in eastern Nebraska. Epidemiology 1, pp 349-35
16. Mills and Wu, Seeley and Guyat, Journal of Pineal Research (2005).