of Geography Chronic Disease in Addison County, Vermont
Hallee Branin Holly Carlson Betsy Cramer Eliza Johnston James Munro Tejas Parikh
Examining recent trends in population distribution reveals an increase in the percentage of people over age 65 living in Vermont. The epidemiological transition fostering an aging population and the replacement of fatal infectious diseases with chronic diseases explains the current leading causes of death in Vermont. Current Vermont data shows that heart disease, cancer, stroke, and chronic obstructive pulmonary disease (COPD) are the leading killers in this state, followed by pneumonia, injuries, and diabetes (Healthy Vermonters 2000, 2). Based on these trends we have chosen to examine the risks and death rates of heart disease, COPD, lung cancer, breast cancer, and diabetes in order to better understand the measures necessary for healthier life and prevention of such diseases in Vermont and within Addison County. In order to analyze various trends in the distribution of the diseases, we chose to focus on town and county scales. County level data were readily available from the Vermont Department of Health. Town death rates in this report are calculated using the number of deaths per town each year and the population estimates of 1998. Due to low numbers of deaths and small town populations, our calculated town death rates do reflect the distribution of deaths over the eleven year time span but are based on too small a sample size to reflect any statistically significant trends. Data on age, income and incidence of smoking, inactivity, obesity, and poverty are used to suggest spatial correlation on the county level, although income is the only risk factor data available on the town level. Heart disease is the leading cause of death both nationally and in the state of Vermont. In 1997, 466,101 lives were lost due to this disease in America (American Heart Association 2000). In Vermont, an average of 1210 people die per year due to heart disease and as a result, nearly 50.2 million dollars are spent on hospital charges
2
(Health Status Report 1998, 14). It is clear that heart disease is a major problem, but in order to reduce this severity it is important to understand the meaning of heart disease and the associated risk factors before any recommendations to reduce the problem can be made. Heart disease is a term that is used to describe any pathological condition of the heart. It is defined as decreased blood flow to the heart muscle also known as ischemia, which results in a diminished supply of oxygen to the heart (Thomas 1989, 416). The cause of heart disease is usually atherosclerosis of the coronary arteries, but the amount of damage to the heart depends on the demand of oxygen by the myocardium, the degree of narrowing of the arterial lumen, and the duration of the ischemia (Thomas 1989, 789). The number of deaths due to heart disease varies by county throughout Vermont. Between 1993 and 1997, 88.2 people per 100,000 died due to heart disease in Addison County, the lowest rate of all the counties in Vermont. Caledonia, Chittenden, and Windham counties had similar death rates ranging from 88.7 to 91.5. Lamoille, Washington, and Windsor counties experienced slightly higher death rates ranging from 93.5 to 100.4. Bennington, Orange, and Orleans counties had even higher rates that ranged from 103.8 to 113.1. The highest rates were found in the northern counties of Franklin and Essex, as well as the southern county of Rutland. These rates extend from 116.8 to 131.5. (Health Status Report 1998, 14) (Figure 1) Within Addison County, heart disease death rates are measured per 10,000 people over the past 11 years. The data points to divergent rates throughout the county, with the higher rates being mostly concentrated in the central towns. (Figure 2) The highest rates of heart disease deaths are found in Middlebury, Vergennes, New Haven, Ripton, and Orwell, which range from 18 to 28.7 deaths per 10,000 people.
3
There are a variety of risk factors, either modifiable or nonmodifiable, that increase the chances of developing heart disease. Of the modifiable risk factors, smoking, physical inactivity, obesity, diabetes, high blood pressure and high cholesterol are the most important. Smoking is the leading cause of preventable illness and death in this country. According to the American Heart Association (2000), a smoker's risk of heart attack is more than twice that of nonsmokers. Nearly 400,000 deaths per year are due to smoking and at least one in three of these deaths is related to heart disease (Bellenir 2000, 162). Smoking increases the risk of heart disease by promoting atherosclerosis. In Vermont's counties, the percentage of adult smokers ranged from 19.4 to 27.4 between 1993 and 1997 (Health Status Report 1998, 30). (Figure 3) The high percentage of smoking among adults in Franklin, Orleans, Orange, and Rutland counties may be one explanation for high death rates due to heart disease in these counties. However, Essex and Bennington counties have low percentages of smoking and high death rates due to heart disease. Therefore, smoking is not the only risk factor associated with heart disease. Physical activity is very important in preventing heart disease. It is necessary to exercise for at least 30 minutes continuously three times a week. Being active increases the heart's ability to pump blood, promotes weight loss, and helps protect against high blood pressure and diabetes. Exercise also increases the heart's strength, endurance, and efficiency (Bellenir 2000, 174). However, being overweight and obese are direct consequences of physical inactivity. These conditions also occur due to excessive consumption of foods high in fat, saturated fat, and cholesterol. Obesity has become an epidemic in this country; one in three Americans is overweight and one in ten is obese, which is defined as being 20 percent or more above the normal weight for that height.
4
(Bellenir 2000, 169). It increases the risk of heart disease by elevating other risk factors such as high triglycerides, high blood pressure, and diabetes. In Vermont, 50 percent of adults were inactive and 26.5 percent were overweight during 1993 and 1997. Specifically within the counties, the range for inactive adults was 43.3 to 60.2 percent and for overweight adults it was 22.5 to 33.0 percent. (Health Status Report 1998, 24-25) (Figure 4 and 5) The counties that had relatively high rates of heart disease deaths such as the northern counties, Orange, and Bennington, also had high percentages of physical inactivity. This trend was not as evident with obesity as a risk factor. Only Franklin, Orange, and Bennington had both a high rate of death and a high percentage of obesity. However, obesity is still a major risk factor for heart disease. Diabetes is also a major contributor to heart disease. It increases the risk of heart disease in men by 2-fold and in women by 5-fold (Bellenir 2000, 171). According to the American Heart Association (2000), nearly two-thirds of all diabetics die from heart disease. Diabetes increases cholesterol and triglyceride levels, which ultimately causes heart disease, and most diabetics are overweight, which is another contributing factor. The counties within Vermont had a prevalence rate of diabetes ranging from 3.3 to 6.1 percent of adults between 1993 and 1997 (Health Status Report 1998, 13). Like smoking, diabetes can also account for many of the deaths due to heart disease, which is evident in those counties in which both the death rate and diabetes percentage is high. High blood pressure and high cholesterol are also major modifiable risk factors of heart disease. High blood pressure increases the heart's workload causing the heart to enlarge and weaken over time (American Heart Association 2000). Eventually, the workload becomes excessive and the heart is unable to function efficiently. It also facilitates the development of atherosclerosis, which further increases the risk of heart
5
disease. High cholesterol levels are also problematic because they can result in the buildup of atherosclerotic plaques, which inevitably leads to heart disease. In Vermont, 22 percent of adults have high blood pressure and 30 percent of heart disease deaths may be due to elevated blood pressure (Healthy Vermonters 2010 2000, 22; Healthy Vermonters 2000 1992, 11). Thirty-one percent of Vermont adults do not get their cholesterol checked and as a result, 10 percent of heart disease deaths are attributed to high cholesterol levels (Health Status Report 1998, 15; Healthy Vermonters 2000 1992, 10). In addition to these risk factors, there are many non-modifiable factors that also increase the risk of heart disease. These include age, family history, race, and gender. Heart disease becomes very common with increasing age, with the highest risk occurring at the age of 65 or older. According to the American Heart Association (2000), about four out of five people who die from heart disease are 65 or older. Older people tend to exercise less on a regular basis, and thus are more likely to be overweight. They are also more likely to develop other modifiable risk factors such as high blood pressure and diabetes. Nearly 85 percent of heart disease deaths that happen in Vermont occur in people 65 or older (Healthy Vermonters 2000 1992, 10). Based on the 1996 population estimates, the percentage of people 65 or older ranges from 9 to 15.5 percent within the counties of Vermont. (Figure 6) Of these counties, Orleans, Essex, Rutland, and Bennington have a high percentage of elderly and a high death rate due to heart disease. This indicates that the mortality rate due to heart disease is characteristic of a high percentage of the population being 65 or older. Furthermore, the tendency of developing heart disease seems to be hereditary. The risk of developing heart disease is much greater if a close relative, such as a mother,
6
father, or sibling, developed the disease before the age of 55. This holds true even if other risks such as smoking or high cholesterol are not present in the family. There is also a genetic influence on heart disease. Recent studies have shown a direct correlation between clogged coronary arteries and high levels of an amino acid called homocysteine. Therefore, individuals that have inherited high levels of homocysteine may attain heart disease at an early age. Race also plays a role in the development of heart disease. It is the leading cause of death for all ethnic groups (VT Health Plan 1999, 76). Compared to whites, AfricanAmericans have a higher risk of developing heart disease because they tend to develop high blood pressure at an earlier age. Also, the disease tends to be more severe in blacks than in whites. Heart disease is also higher among other ethnic groups such as Asian Americans, American Indians, Mexican Americans, and native Hawaiians. This is partly due to higher rates of diabetes and obesity in these ethnic groups (American Heart Association 2000). Because of the small minority population within Vermont, analyses of these health issues are not possible. "Heart disease is an equal opportunity killer" (Bellenir 2000, 33). This is a common view of heart disease based on gender differences even though it is hard to separate the role of gender from the other risk factors. Although men are at greater risk of developing heart disease, more women die from this disease in America. For women, estrogen protects them from heart disease, but after menopause the risk of heart disease approaches that of men. Post menopause also increases the risk of high blood pressure and cholesterol, which when combined, further increases the risk of heart disease in older women.
7
Although poverty rate is not a risk factor for heart disease, it does correlate with the death rate due to heart disease. The northern counties and Bennington County have high death rates for the disease and have a greater percentage of people living below the federal poverty level. (Figure 7) The reason for this correlation may be that low income prevents individuals from visiting the doctor to get their blood pressure or cholesterol levels checked. Also, joining a health club may be too expensive, therefore they may not be exercising enough. Both reasons indicate that these individuals are at greater risk of developing heart disease. The income data within Addison County indicates variation of income levels between the towns. (Figure 8) Of the towns with the higher mortality rates, Middlebury, Granville, Bristol, and Orwell have relatively low median household income. This correlation is also confirmed by Monkton, Waltham, Cornwall, Leicester, and Goshen, which have high median household income levels and consequently low heart disease death rates. As previously mentioned, the death rate due to heart disease was the lowest in Addison County between 1993 and 1997 at a rate of 88.2 deaths per 100,000 people. Based on the risk factors measured in Addison County, the percentage of smokers was 22.5 percent, physical inactivity was 45.7 percent, overweight adults was 26.9 percent, diabetes prevalence was 4.3 percent and population age 65 or older was 10.5 percent. The poverty rate was 9.7 percent. (Vermont Status Report 1998, 13,24,25,30,40) All of these percentages were relatively low compared to many of the other counties. Therefore, it may not be one risk factor that is contributing to the low death rates, but rather a combination of these risk factors. Within Addison County, a shortage of data on risk factors for these towns prevents us from making any conclusions.
8
In order to reduce the heart disease death rate throughout Vermont and Addison County, a number of recommendations must be considered. These recommendations are based on altering all of the modifiable risk factors associated with heart disease and thus require changes in lifestyle. The biggest change is to quit smoking, as it is the most preventable cause of heart disease. Although quitting is not easy, it is imperative that physicians continue to reinforce the hazards associated with smoking. Also, smokers usually quit over time, therefore an increase in counseling services will help facilitate and expedite this process especially in counties such as Franklin, Orleans, Orange, and Rutland counties, in which the smoking percentage and heart disease death rate are high. Quitting smoking decreases blood pressure, which also reduces the risk of heart disease. Other smoking recommendations made throughout the paper also apply here. Physically inactive people are almost twice as likely to develop heart disease compared to those who are physically active. Some of the barriers to physical activity include cost, access, transportation, and health problems (VT Health Plan 1999, 40-41). One way to promote more exercise is to increase the number of low or no cost health facilities or community programs especially in Franklin, Orleans, Essex, Orange, and Bennington counties because of the high heart disease death rate and high percentage of inactivity. There should also be an increase in public transportation to these health centers in order to accommodate those who have limited or no means of transportation. Hopefully these recommendations in unison will encourage more people to exercise. In addition to exercise, people need to eat properly. A healthy diet consists of low fat intake and an increase in fruits, vegetables, grains, and legumes. Together, exercise and a healthy diet will improve many other risk factors such as obesity, high blood pressure, high cholesterol and diabetes.
9
Although treatment options including medications and surgical procedures such as angioplasty are available, taking preventative measures, such as quitting smoking and increasing exercise, are the best ways to decrease the risk of heart disease. It is important to remind health care providers to continue to emphasize the risks associated with heart disease and counsel their patients to make the necessary changes in their lifestyle. Making such changes will not only reduce the disability associated with heart disease, but also improve the quality and duration of life. Following heart disease, COPD and lung cancer are two of the most prevalent chronic diseases affecting Vermonters today. COPD, including emphysema, chronic bronchitis, and asthma, accounts for 8.8 million dollars in hospital charges and kills about 258 Vermont residents annually. (Health Status Report 1998, 20) With COPD, airways narrow and excess secretion of mucus may occur due to chronic inflammation. Common symptoms of COPD include difficulty breathing, persistent cough, wheezing, and chest pain. In the later stages of the diseases, victims will often need supplementary oxygen to help with breathing. (Thurlbeck 1990, 6-11) Lung cancer accounts for 3.3 million dollars in hospital charges and is the leading cause of death from cancer in men and women in Vermont today with an average of 129 women and 185 men dying from lung cancer in Vermont each year (Health Status Report 1998, 20). While many patients are asymptomatic during the earlier stages of the disease, some will experience symptoms similar to COPD, such as a persistent cough, wheezing, and chest pain, while others may experience more general symptoms such as fatigue, weightloss or swelling in the arm and neck area (Dest 2000, 32). Both lung cancer and COPD may largely be preventable illnesses. Steps to prevent these diseases include avoiding smoking, and decreasing exposure to environmental risk factors such as second-
10
hand smoke and other air pollutants. Eating a diet rich in fruits and vegetables, which contain vitamin A precursors, carotenoids, and other antioxidants such as vitamins C and E may also help protect against lung cancer and COPD (Diet and Lung Cancer 1990, 135-137; Antinoro 2000, 1). Within Vermont, the number of deaths from COPD varies from county to county as seen in Figure 9. In the years from 1993 to 1997, two counties in the northeastern corner of Vermont, Essex, and Caledonia, as well as Bennington County in the south, had slightly higher death rates from COPD ranging from 33.8 to 40.7 deaths per 100,000 people each year. Excluding Bennington, most of the southern counties, including Windham, Windsor, and Rutland, as well as the more northern counties of Washington and Chittenden had relatively lower rates of 23.5 and below. Addison County had an average rate of 27.1 attributed to COPD, which is similar to the rates of Orange and Orleans counties (26.7 and 32.1 respectively) and in the relatively high range for Vermont counties in general. (Health Status Report 1998, 20) There is variation of death rates from COPD within Addison County as well as indicated by Vermont Health Department data from 1988-1998, although it should be noted that due to very small populations in the towns within Addison, rates of death are variable even over this eleven year span. As seen in Figure 10, many of the southeastern towns in Addison County such as Ripton, Hancock, and Salisbury, as well as Bristol, experienced the highest death rates of 4.88 to 6.76 deaths per 10,000 people per year. Middlebury, Leichester, Ferrisberg, and Shoreham also had relatively high death rates of 4.01 to 4.88. These towns contrast with Whiting, Bridport, Panton, New Haven, and Starksboro, which were in the low range of death rates, of 0.72 to 2.69, and the town of Goshen (pop. 226 in 1998) which saw no deaths from COPD.
11
Lung cancer deaths are distributed slightly differently than COPD deaths within Vermont as shown in Figure 11. During the same time period (1993-1997), notably high rates of lung cancer deaths (51 to 55.5 deaths per 100,000 people) were seen in Windham, Bennington, Washington, Chittenden, and Franklin counties, with Franklin and Washington the highest at 53.6 and 55.5 respectively. Lamoille, Essex, Caledonia, and Windsor counties saw relatively low rates of death from lung cancer of 41.8 or lower, and Addison County experienced one of the lowest death rates of 38. (Health Status Report 1998, 20) Towns within Addison also experience different rates of lung cancer. Death rates from lung cancer (Vermont Department of Health data from 1988-1998) as seen in Figure 12, are highest (7.17 to 8.04) within the county in Lincoln, Salisbury, Leichester, and Shoreham, and also relatively high (4.54 to 7.17) in Cornwall, Addison, Panton, Vergennes, Bristol, and Starksboro. It is interesting that the more southern towns of Salisbury, Leichester, and Shoreham have relatively high death rates from both COPD and lung cancer. There are several possible demographic factors that may affect the distribution of these diseases in Vermont. Men have a statistically higher death rate from both COPD and lung cancer in Vermont. However, while the trend in incidence has been relatively stable for men, the rate in women has been increasing and may be attributed to rising rates of smoking in women (Health Status Report 1998, 20). This is especially unfortunate since recent research suggests that women may be more susceptible to the effects of tobacco-related carcinogens (Henderson 1999, NA). Additionally, COPD and lung cancer display an incidence that increases with age and is more common in people of middle or lower income. One reason that some of
12
northeastern counties display a higher mortality rate from COPD may be attributed to an older age distribution in these areas (Figure 6). Orleans, Essex, Caledonia, and Bennington counties, which all had higher rates of COPD deaths, also estimated 14.1 to 15.7 percent of the population as 65 years or older, a relatively high statistic compared to other areas across the state (based on 1996 population data). However, other counties with relatively older age distribution such as Rutland (14.8%) and Windsor (15.1%) were in the lower range of COPD and lung cancer death rates, indicating that age is not necessarily a determining factor in county rates. (Health Status Report 1998, 40) Income rates may also correlate with deaths from COPD and lung cancer in Vermont, as in some of the same counties which saw higher death rates for the diseases, including Orleans, Essex, Caledonia, and Bennington, a higher percentage of people live below federal poverty level (Figure 7). Several consequences of lower per capita income may explain the higher disease rates, including less availability of preventative care and treatments. Other behavioral aspects may include a poorer diet lacking in fruits and vegetables, as well as a general tendency for smoking cessation rates to be 27-33 percent lower for lower income smokers (Blueprint for a Tobacco Free VT 2000, 8). There seems to be fairly little correlation between the lung cancer and COPD death rates and income rates for towns within Addison County (Figure 8), except perhaps in the towns of Bristol, Middlebury, and Shoreham, which had relatively low household incomes and relatively high COPD and lung cancer rates. Both COPD and lung cancer have a number of common environmental and behavioral risk factors which may also show some correlation with death rates in Vermont. Eighty to ninety-percent of COPD and lung cancer deaths have been attributed to smoking (Health Status Report 1998, 20). Not only do smokers increase their own
13
risk, but they may place non-smokers around them at home or in the work place at risk, as each year an estimated three thousand non-smoking Americans die of lung cancer (Raloff 1998, 251; VT Health Plan 1999, 55). In Vermont over the past decade, the overall rate of smoking among adults, twenty years or older, declined from thirty to twenty-one percent, and now seems to be relatively stable, 21.3 percent in 1990 and 22.1 percent in 1995. In 1998, the percentage of the population smoking in Vermont was 22.6 percent for adults and 44 percent for grade 12 students, although the latter statistic seems to be starting to drop (Health Status Report 1998, 31). As seen in Figure 3, smoking rates among Vermont counties are relatively high across the state, as the percentage of adults who smoke ranges between 19.4 and 27.4 percent. Addison falls approximately in the middle of this range at 22.5 percent. (Health Status Report 1998, 30) Some of the higher rates of smoking that occur in Franklin, Orleans, Caledonia, Washington, Orange counties do seem to correlate with higher rates of death from COPD and/or lung cancer. However the slightly lower rates of smoking in other counties in Vermont does not mean smoking is not a predominant cause of disease in those places as well. According to the Surgeon General, radon is considered the second leading cause of lung cancer. In Vermont, about forty cases of lung cancer each year may be attributed to radon exposure (VT Health Plan 1999, 57). Radon is a naturally occurring gas, a byproduct of uranium decay. Radon gas is released from soil below houses and seeps in though the foundation and walls. Levels vary from area-to-area depending on the ground the houses are built on, and can even vary from house-to-house. Testing for radon is an easily obtained service but most likely underused by Vermonters. (Viera 2000, 950) Air pollutants, other than tobacco smoke components and radon, constitute other risk factors for COPD and lung cancer. Data from the Vermont Health Department
14
attributes as many as 14 percent of cases of COPD to environmental and occupational exposures. Vehicle exhaust and other industrial emissions, especially, may drastically reduce air quality and exacerbate airway diseases (MacNee and Donaldson 2000, 390). Other air pollutants linked to COPD and lung cancer include talc and asbestos, often found in buildings or near mines (Kilburn 2000, 104). There are many aspects of preventative care that should be a focus in Vermont. Smoking rates are high across the state, and should be a primary concern in policies for prevention. Lowering smoking rates begins with education of the dangerous health consequences of the habit. Educational programs are most helpful when they reach youths before they ever start smoking, but can also be useful to get people to stop at any age. The Vermont Department of Health has suggested a five-element comprehensive program to reduce smoking rates. This program includes an emphasis on communitybased programs, treatment of tobacco addictions, school-based programs, the enforcement of tobacco laws, regular monitoring of programs, and assessment of their effectiveness (Blueprint for Tobacco-Free VT 2000, 9-10). In addition, the patient-physician relationship is a crucial factor in the prevention and care of COPD and lung cancer. A recent study found that while 75 percent of Vermont physicians ask adult patients about smoking and tobacco use, only 44 percent routinely advise patients about strategies to quit (VT Health Plan 1999, 88). Consistency in questioning patients about smoking habits will serve to make patients more aware of the risks and less likely to start, and encourage patients to quite sooner rather than later. The patient-physician relationship is also important in early detection of lung tissue damage, a possible indication of the early stages of COPD or lung cancer. Screening is
15
especially important for people who are at high risk for the diseases such as those people who smoke, are exposed to other air pollutants, or have a family history of the disease. Once a person falls ill with COPD or lung cancer, the patient-physician relationship continues to be a key to treatment. There are several treatment options for both COPD and lung cancer, with varying success depending on the stage and type of disease. Each patient must be individually assessed and must choose the right option for him or her, as different patients will respond differently to treatments. COPD is usually treated with medication to reduce symptoms and ease breathing, while these medications may be long- or short-term therapies. Inhaled corticosteroids, anticholinergics, and bronchodilators are included in the list of medications that are used for various types of the disease (Rudolf 2000, 29). Conventional treatment for lung cancer includes surgery, radiation and chemotherapy. Some patients with lung cancer may be candidates for an operation to remove a tumor, however if the cancer has spread, or the tumor is otherwise determined inoperable, other options must be explored. As of now, radiation and chemotherapy are two of the most aggressive treatments for lung cancer (Kahana 2000, 840). New advances include thoracoscopic surgery in which surgeons remove tumors through a small incision in the chest cavity, and radioprotectors, which reduce the damage done to healthy tissue during radiation (Dest 2000, 32-33). People with COPD or lung cancer need to be educated about the steps they can take to control their disease, improving their diet, and reducing damaging behavior such as smoking. A strong support system of community, family and physician can help in this process. The future impact of COPD and lung cancer in Addison County and across the state of Vermont depends on the programs and resources available to work on the prevention and treatment of the diseases. Since damage the to lungs cannot be reversed,
16
prevention is obviously the key to reducing the incidence of COPD and lung cancer. Policy efforts should be spread across the state, but should also focus on high-risk populations such as in areas previously noted where high death rates of COPD and lung cancer occur, and in elderly and low-income populations. Breast cancer is another chronic disease which has become prominent in people's minds both nationally and in Vermont. It is a very important disease to study not only because it is the second leading cause of premature death for women and the most common cancer for American women, but also because early detection of the disease dramatically increases chances for survival (Cook, et. al 1996, 367). Studying and publicizing the way breast cancer affects people's lives will bring needed attention to the disease and preventive care methods that can provide early detection. In addition to the importance of general study, it is useful to analyze where breast cancer is the biggest problem so that resources can be focused on specific geographic areas and distinct populations. To isolate particular areas or populations on which to focus resources, the first step is to determine the spatial distribution of incidence and mortality rates for breast cancer, and then to research what risk factors contribute to this distribution. Incidence and mortality rates on the county level will be given as the rate per 100,000 women, and while this paper focuses on Addison County, it is important to analyze these rates at different scales because patterns may exist at one scale but not another. Therefore, breast cancer incidence will be presented at three scales: national, Vermont state, and Addison County. Rates in towns within Addison County are not available. In 1995, the national breast cancer incidence was 111.3 (per 100,000 women), compared to a Vermont average from 1994-1996 of 102.4 and an Addison County average over those same years of 92-155 with a 90 percent confidence interval. While
17
the upper range of this interval is third highest in Vermont, the confidence intervals from every county overlap with one another, suggesting that one county is not statistically better or worse than any other county. (Vermont Health Plan 1999, 110) Mortality rates fall in a pattern similar to incidence rates, with a 1995 national mortality rate of 25 (per 100,000 women), a Vermont rate of 23.9 averaged from 19931997, and an Addison County average from those same years of 33.2. This level of 33.2 deaths per 100,000 women gives Addison County the highest mortality rate in Vermont, followed by Orange, Windham, and Orleans counties (Figure 13). Essex County has the lowest breast cancer mortality rate. In 1994, Vermont's mortality rate from breast cancer dipped below the nation's for the first time since 1981, and both rates appeared to continue a downward trend in 1995. (Health Status Report 1998, 8-9) We were able to obtain mortality data for towns within Addison County, and found that townships vary considerably in their rates. In calculating mortality per 10,000 people, we found that Hancock had the highest rate at approximately 8 women per year (Figure 14). The largest town, Middlebury, had a very low rate at just over one individual per 10,000 each year. However, this data is highly unreliable because sample populations from which calculations were made were very small, and in some cases one death boosted a small town's mortality rate to a rather high level. Further, the mortality rates can be deceiving, because none of the towns for which we calculated rates had populations over 10,000 people. This means that while Hancock was statistically found to have 8 breast cancer deaths per 10,000 people each year, the actual population in the town is only 340 people and therefore the rate is translated to only ~0.3 deaths from breast cancer per year. In addition, the three deaths from breast cancer Hancock experienced in ten years which produced its high mortality rate could very well have
18
occurred out of coincidence, boosting the mortality rate in a town that may not actually contain high levels of risk factors. Unfortunately, we are unable to test this theory because of a shortage of data. Now that we have an overall picture of the spatial distribution of breast cancer incidence and mortality, we must begin to analyze potential factors contributing to this spatial pattern. There are no definitive causes of breast cancer, but there are risk factors which may increase a woman's chance of developing the disease. Perhaps the two most well known risk factors are age and genetics. Most cases of breast cancer occur in women over 55, while the highest risk occurs in women over 60 (Cook, et. al 1996, 25). While Addison County has a relatively low proportion of its population over 65, Orleans County has a high percentage and Windham and Orange counties have medium levels (Health Status Report 1998, 40) (Figure 6). While this suggests there may be some correlation between mortality rates and population age, Essex County has the state's largest percentage of population over 65, yet as mentioned above it has the lowest breast cancer mortality rates in the entire state. For genetics, while no breast cancer gene has been identified, an individual does seem to have an increased risk if they have a family history of the disease, especially if a first-degree relative has been affected. Other risk factors which may contribute to an increased risk for breast cancer include smoking, alcohol use, obesity, poor diet or nutrition, ethnicity, hormonal variations, and low-dose radiation. Some studies suggest that smoking may double a woman's chance of developing breast cancer, although distribution of smoking habits in Vermont do not closely correspond to breast cancer incidence or mortality (Figure 3). While Orleans and Orange counties have high rates of smoking, Addison and Windsor
19
both have low rates (Health Status Report 1998, 30). It is likely that the scale of an individual's smoking problem contributes to their scale of risk from this factor. Heavy alcohol consumption (>3 drinks/day) may contribute to 1.5 times the normal risk for breast cancer, and studies have found the highest correlation here with beer rather than wine or other alcoholic beverages (Eades 1991, 40). This was found by conducting studies on a broader scale concerning countries such as France, Italy, and Portugal whose populations consume a lot of wine but have low rates of breast cancer. Once again, the spatial distribution of alcohol consumption for Vermont counties does not match breast cancer incidence or mortality rates, and in fact the Porter health care area (which roughly corresponds to Addison County) has one of the lowest rates of problem drinking in the state (Health Status Report 1998, 39). Obesity may incur as much as double the risk of breast cancer because fatty tissue produces its own estrogen and can overstimulate sensitive organs such as the breasts, but once again the pattern of obesity by county does not correlate with breast cancer incidence or mortality (Eades 1991, 44). In fact, the four Vermont counties with the highest breast cancer mortality rates are all among the lowest for percentage of obese adults (Health Status Report 1998, 24) (Figure 5). Factors that may present smaller risks include fatty diets low in vitamins and minerals, hormonal variations due to early menstruation, late onset of menopause or late motherhood, and low-dose radiation from sources including nuclear power plants (Eades 1991, 53; and Clorfene-Casten 1996, 61). Certain ethnic groups also have higher incidence rates including native Hawaiians, Caucasians, and African Americans versus Native Americans and Japanese (Cook, et. al 1996, 371). As mentioned before, populations of the different minority groups are so small in Vermont that no significant analysis of them can be conducted.
20
While all of the risk factors discussed above may play some role in determining which women are at high risk for developing breast cancer, the analysis above shows that identifying high risk populations and explanations for high incidence or mortality is far from easy. None of the risk factors mentioned above directly correspond to the relatively high mortality rate and incidence rate of breast cancer in Addison County, which illustrates why no clear-cut cause for breast cancer has been identified. However, what this may mean is that these risk factors are likely working in combination to determine which women are predisposed to developing the disease. In other words, one woman may employ three of these risk factors through behavioral choices or physical characteristics, which would thereby put her at a high risk for breast cancer. If this were the case for many women, it would be difficult to find clear-cut trends leading to high incidence or mortality by analyzing each risk factor separately. In short, this means that women should be on the safe side and try to reduce all possible risks, for many risks are unproven and are surrounded by conflicting studies. The good news is that while a woman reduces these potential risks for breast cancer, she will also be improving her general health. Aside from the physical risks presented above, breast cancer is also detected at higher rates in more affluent populations. However, the key word here may be detected, for rates of screenings (mammograms and clinical breast exams) are much higher for high-income populations than low-income populations (Eades 1991, 60). From 19931997, 83 percent of high-income women were screened, 67 percent of medium-income women were screened, and 43 percent of low-income women were screened (Health Status Report 1998, 9). These statistics provide us with a potentially useful tool, for they show us exactly which populations should be targeted for screening programs. Income is
21
the only factor for which we were able to obtain data at the town level, and interestingly Hancock, the town with the highest breast cancer mortality rate, happens to have one of the lowest mean incomes in Addison County (Figure 8). Perhaps this illustrates the direct relationship between income and availability of screening. Early detection is so important because prognosis for a cure is highly dependent upon the scale or stage of breast cancer when it is first diagnosed. If the cancer is detected when it is known as carcinoma in situ, the cure rate is greater than 95 percent. However, if detected in stage I the rate drops to 85 percent, for stage II it is 65 percent, stage III is 40 percent, and stage IV has a 10 percent cure rate. As the stage of detection becomes later, treatment escalates from breast-sparing surgery (lumpectomy) to removal of the breast (mastectomy) along with radiation and chemo/hormone therapy. (Eades 1991, 107) In the late 1980's Vermont was identified as having one of the highest death rates from breast cancer in the nation, so the state undertook actions to improve screening and preventive care which includes mammograms, clinical breast exams, and education concerning self-examinations (Vermont Health Plan 1999, 111). In 1995 the "Ladies First" project was implemented, supplying low-income women over age 50 with free screening. The formation of this program was based upon research showing that 93 percent of women with an annual income over $35,000 obtain regular screening while only 39 percent of women with income less than $10,000 are screened (Healthy Vermonters 2000: Progress Report 1996, 5). Physicians have also been encouraged to instruct women on self-exams, as it is estimated that 90 percent of breast cancers are discovered through this process (Eades 1991, 75). Vermont's breast cancer deaths have
22
steadily fallen since 1987, corresponding to a rise in screening (Vermont Department of Health 1999, 112; Health Status Report 1998, 9). To continue this trend of decreasing breast cancer mortality, the state of Vermont and Addison County can work to educate the public about the importance of screening and work to increase the availability of that screening. This type of action may be best implemented by advertising and organizing on a state level while giving initiative to local health organizations. In addition, health officials can be encouraged on the local level to inform patients about the importance of reduction in all behavioral risk factors, especially smoking and alcohol consumption along with improvements in diet and nutrition, as this will not only work to potentially reduce breast cancer incidence but will also promote general good health. On a national scale, research efforts are steadily growing and studies are tending to focus on issues pertaining to hormonal risk factors and the drug Tamoxifen which may potentially prevent breast cancer in high-risk women (Assersohn 1999, 165). With a decrease in known risk factors, increased education and screening for breast cancer, and continued research, there is no reason not to think that we can maintain the current downward trend in breast cancer mortality. We now turn to diabetes, chosen because it is the seventh leading cause of death in Vermont (Health Status Report, 1998, 12). The severity of this disease stems from the fact that multiple other health problems may occur as a result of a person having diabetes such as high blood pressure, heart disease, stroke, and kidney failure among others, all of which can pose serious health problems and potentially lead to death (Healthy Vermonters Progress Report, 1996, 9). Nationally, there are 15.7 million confirmed cases of diabetes, plus an estimated 5.4 million people who remain undiagnosed. Within the state of Vermont it is estimated that 28 thousand people suffer from the disease, with
23
one third of them remaining undiagnosed. In particular, the prevalence rate in Addison County hovers around 4.3 percent from year to year, which places it as the fifth highest in the state (Healthy Status Report, 1998, 12). The distribution of diabetes related death rates in the counties of Vermont can be seen in Figure 15. Within Addison County, the towns of Middlebury, Ripton, and Whiting have the highest mortality rates due to diabetes as seen in Figure 16. This disease is separated into two types according to the age group it affects, and the bodily chemicals in which it causes the imbalance. Type I diabetes is the insulin related type that affects primarily children; currently approximately 300-400 Vermont children suffer from type I diabetes. Type II diabetes is the non-insulin related type that afflicts overweight adults who are 40 years of age and older. This latter type accounts for 90-95 percent of all diagnosed cases, and will be our primary concern here (Health Status Report, 1998, 13). It is also thought that people of Native American/Alaskan, Asian, Hispanic, and African American dissent are at higher risk of developing diabetes than are people of Caucasian ethnicity. This trend might be explained by a genetic disposition to the disease found more frequently in these minority gene pools as opposed to the Caucasian pool. It may also be due in part to the fact these groups tend to be associated with lower incomes than are Caucasian Americans. The possible effects of income and poverty on diabetes will be discussed further. Probably the most severe risk factor for type II diabetes is obesity. Obesity affects glycemic control and increases blood pressure, both factors have been linked to the onset of diabetes (Hillier, A25). As of 1998 212,500 or 26.5 percent of all
24
Vermonters were overweight according to the definition. Similarly, 26.9 percent of Addison county residents are overweight (Vermont Status Report, 1998, 24). Exercise is clearly closely related to obesity and therefore a major risk factor to acquiring diabetes is an inactive life style. Since exercise helps avoid obesity, it also is found to decrease the risk of diabetes. People who exercise at moderate and intense levels up to three times per week are at a drastically lower risk of developing the disease (Morin et al, A25). Studies have also been performed which were controlled against body mass, age, smoking, alcohol use, and cholesterol level and still the direct correlation exists between an active lifestyle and a decrease in the risk of diabetes (Lippincott 2000, 112). Exercise is also considered a very effective treatment for people with diabetes. It is thought that these people should exercise daily and burn at lease 1000 calories per week to keep blood sugar levels under control (Albright et al 2000, 1345). The distribution of the prevalence of obesity and inactivity around the state of Vermont can be seen in Figures 15 and 5. In the case of inactivity, it is clear that the northernmost counties are the worst off. Also, Bennington County in southwestern Vermont appears to have a problem with lack of exercise. The data concerning obesity correlates with this to a certain extent. Clearly there is agreement between these two risk factors in Bennington County but obesity seems to be slightly less common than might be expected in the north given the low percentage of people who exercise regularly. Generally, these high rates of inactivity and obesity can partially be attributed to the weather. The long winters and cold, wet springs will tend to discourage people from exercising during the majority of the year. The fact that the northern counties have the most unhealthy lifestyles in general supports our ...