This is the first Annual Report to the Nation to document a decline in both the incidence and the death rates from all cancers combined in both men and women. These declines occurred in most racial and ethnic groups and partly reflect decreases in the three most common cancers in men (lung, colorectum, and prostate) and two of the three most common cancers in women (breast and colorectum), as well as the leveling off of lung cancer death rates in women. These cancers account for about half of all cancer cases and deaths in both men and women. The sustained declines in cancer death rates overall and for the three major cancer sites in men and two major cancer sites in women have been discussed previously (2
). Declines in cancer death rates indicate real progress in cancer control, reflecting a combination of primary prevention, early detection, and treatment.
Trends in incidence are more difficult to interpret, because both increasing and decreasing trends can result from changes in screening and diagnostic practices as well as changes in exposure to risk factors. Incidence declines attributed to reductions in risk factors include the decrease in lung cancer in men caused by historical patterns of smoking cessation (32
) and sharp declines in breast cancer incidence in 2002–2003 following reduced use of hormonal replacement therapy (HRT) (34
). The decline in breast cancer incidence attributed to HRT use is particularly notable because of the short lag time between changes in exposure and resulting changes in incidence. A similarly rapid change in a hormonally related cancer following changes in hormonal therapy was seen in the 1970s when the incidence of endometrial cancer first increased and then decreased with the rise and fall of HRT formulations containing estrogen (38
Changes in incidence rates can also be related to changes in use of diagnostic and screening modalities. The accelerated decline in the colorectal cancer incidence rate since 1998 may be associated with increased use of colorectal cancer screening, which prevents cancer through removal of precancerous adenomatous polyps (7
). Between 2000 and 2005, the percentage of adults aged 50 years and older who reported having had colonoscopy increased from 20% to 39%, whereas the percentage reporting testing for fecal occult blood decreased from 17% to 12% (42
). Overall, the use of colorectal screening among adults 50 years and older increased from 27% in 1987 to 50% in 2005 (42
Changes in use of mammography may have also contributed to recent declines in breast cancer incidence trends that began in 1999. The prevalence of recent mammography began to stabilize or decline in the late 1990s after increasing for many years (44
); this trend may have contributed to the decline in incidence, due to decreased detection or reduced number of undiagnosed prevalent cancers (35
). Long-term declines in cervical cancer incidence in women are likely related to widespread dissemination of cervical cancer screening (46
In contrast to mammography and colorectal cancer screening, the benefits of prostate-specific antigen (PSA) screening in reducing morbidity and mortality from prostate cancer have not yet been established (49
). Trends in use of PSA screening have undoubtedly influenced prostate cancer incidence trends over the last several decades (50
), and the leveling off of PSA screening may be contributing to the recent decline in prostate cancer incidence because of decreased detection, or reduced number of undiagnosed prevalent cancers. According to the National Ambulatory Medical Survey (51
), the frequency of PSA testing during visits for a general medical examination among American men increased from 1995 through 2002 and then leveled off through 2004.
The increasing incidence of several other cancers is related, at least in part, to increased detection and use of diagnostic and imagining technology. These cancers include melanoma of the skin, cancer of the kidney and renal pelvis, and thyroid cancer (52
With respect to trends in lung cancer, tobacco use, and tobacco control, this report documents large geographic variation in tobacco smoking that, together with generational differences in past smoking behavior, is delaying the decrease in lung cancer death rates in women and slowing the decrease in men. Cigarette smoking alone still accounts for approximately 30% of all cancer deaths in the United States, despite reductions in smoking prevalence (57
). Most (80%) of these smoking-attributable cancer deaths involve lung cancer, although smoking also causes cancers of the oral cavity, pharynx, larynx, esophagus, stomach, bladder, pancreas, liver, kidney, uterine cervix, and myeloid leukemia (58
). Lung cancer is commonly perceived by public health professionals as the sentinel health consequence of cigarette smoking because although smoking causes more deaths from cardiovascular and respiratory diseases than from lung cancer (58
), those conditions are less strongly associated with smoking than is lung cancer.
Sex differences in lung cancer incidence and death rates, and particularly the delayed increase and then leveling off of lung cancer risk in women compared with men (), have been described repeatedly elsewhere (32
). These temporal differences reflect the later uptake of cigarette smoking among women, who began smoking predominantly during and after World War II, compared with men, who began cigarette smoking in the early 20th century, with large peaks of initiation during the two World Wars (60
). Because of the historical differences in smoking patterns, the sustained decrease in lung cancer incidence and death rates in men has been a major contributor to the overall decrease in male cancer incidence and death rates (62
), whereas the leveling off of the lung cancer death rate among women has only recently facilitated the downturn in the overall female cancer death rate.
Less attention has been paid to the prominent state and regional variations in the trends in lung cancer and tobacco use in men and women, particularly as these relate to various indices of state tobacco control activity. Although the lung cancer death rate among men has been decreasing nationally since the early 1990s, the rate of this decrease varies substantially by state and geographic region. For example, the average percentage decrease in the lung cancer death rate among men in California from 1996 through 2005 (2.8% per year) is more than twice that of many states in the Midwest and South. The geographic variation is even more extreme among women, for whom the lung cancer death rate increased from 1996 through 2005 in 13 states and decreased in only three. Although fewer data on trends at the state level are available for lung cancer incidence than for mortality, in five states (Pennsylvania, Illinois, Minnesota, Nebraska, and Idaho), lung cancer incidence among women was increasing and mortality rates were stable during the same time interval. Our findings help to explain why the lung cancer incidence and death rates among women nationally have not decreased, despite reductions in smoking prevalence. At least three factors contribute to the rates and trends among women (–). First, age-specific incidence and death rates continue to increase for age groups 70 and above ( and ), and rates occurring in these age groups heavily influence the trend in the age-standardized rates because they contribute to more than 50% of the age-standardized rates. Second, smoking cessation rates are historically lower in women than men, especially at older ages (63
). Based on data from the National Health Interview Surveys, the “quit ratio,” or fraction of ever smokers who had stopped smoking, was more than 50% higher in men than women aged 65 years and older in 1965 and 1970, and remained 15%–19% higher in 1990 and 1994 (63
A third factor that is delaying a decrease in female lung cancer incidence and death rates nationally is that incidence and death rates continue to increase in certain regions of the United States. All of the 13 states in which the lung cancer death rates increased in women from 1996 through 2005 are located in the South and Midwest, where, on average, the prevalence of smoking is higher and the annual percentage decrease in current smoking among adult women is lower than in the West and Northeast. State variations in smoking prevalence are influenced by several factors, which include public awareness about the harmful health effects of tobacco use, social norms about tobacco use, educational levels within the state, racial and ethnic variations among the states, tobacco control activities at the state and local level (64
), and industry promotional activities (66
). California was the first state in the United States to implement a comprehensive state-wide tobacco control program (70
) and has made the greatest progress in reducing tobacco use (71
), although Utah continues to have the lowest smoking prevalence. Adult smoking prevalence among women in California decreased from 14.5% in 1997 to 11.4% in 2006 (). Many states in the South and Midwest have only recently achieved a reduction in female smoking prevalence and have not yet experienced a leveling off or decrease of lung cancer incidence and death rates among women. For example, the percentage of adult female current smokers in Kentucky changed little from 1997 (28.7%) to 2006 (28%) ().
Most Southern and Midwestern states continue to have a high prevalence of smoking and low excise tax (), despite compelling evidence that excise taxes and other components of comprehensive tobacco control can achieve substantial reductions in tobacco use (75
). Many Southern states have historically been economically dependent on tobacco farming and production (77
). The exceptionally low lung cancer rate in Utah reflects the religious prohibition against smoking among Mormons (78
Nationally, the anticipated future decline in age-standardized lung cancer rates in women will be offset by the generation of women born between 1950 and 1960 who passed through adolescence and young adulthood at the time when cigarette brands such as Virginia Slims were introduced and marketed intensively to women (26
). A sharp increase in smoking initiation occurred among these women between 1965 and 1975 (). These same birth cohorts account for the interruption in the decline in female lung cancer incidence and death rates between ages 30 and 49 years (–). Women who were born in this era will likely continue to offset future decreases in lung cancer incidence and death rates in other birth cohorts over the next 50 years.
As mentioned above, tobacco smoking causes many cancers in addition to lung cancer. However, only three of these cancers (cancers of the oral cavity, esophagus, and larynx) have a smoking-attributable mortality of greater than 50% (57
). Incidence and death rates for cancer of the oral cavity and larynx and incidence rates for squamous cell carcinoma of the esophagus (the histological type most strongly associated with smoking) have decreased in both men and women over the past 15 years (data not shown), following reduction of smoking prevalence over the past 40 years.