A great deal of study has gone into the assessment of the epidemiology of NTM infection and disease in many different parts of the world. Review of the available studies provides insight into the frequency of this clinical problem as well as important limitations in current data. Study methods have varied greatly, undoubtedly leading to differing biases. In general, reported rates of infection and disease are likely underestimates, with the former probably less accurate than the latter, given that people without significant symptoms are not likely to have intensive investigations to detect infection. Pulmonary NTM is a problem with differing rates in various parts of the world. North American rates of infection and disease have been reported to range from approximately 1-15 per 100,000 and 0.1-2 per 100,000, respectively (see Table 1). Rates have been observed to increase with coincident decreases in TB. MAC has been reported most commonly, followed by rapid growers and M kansasii. Generally similar rates have been reported in European studies, with the exception of extremely high rates in an area of the Czech Republic where mining is the dominant industry (see Table 2). These studies have also shown marked geographic variability in prevalence. The only available population-based studies have been in South Africa and report extremely high rates of infection, three orders of magnitude greater than studies from other parts of the world (see Table 3). This undoubtedly reflects the select population with an extremely high rate of TB and resultant bronchiectasis leading to NTM infection. Rates in Japan and Australia were similar to those reported in Europe and North America and also show significant increases over time (see Table 3). Specific risk factors have been identified in several studies. CF and HIV, mentioned above, are two important high-risk groups. Other important factors include underlying chronic lung disease, work in the mining industry, warm climate, advancing age, and male sex. Aside from HIV and CF, mining with associated high rates of pneumoconiosis and previous TB may be the most important historically, reported in studies worldwide [63]. A recurring observation is the increase in rates of infection and disease. The reason for this is unclear but may be caused by any of several contributing factors. The possibility exists that the apparent increase is either spurious or less significant than studies would suggest. Changes in clinician awareness leading to increased investigations, or laboratory methods leading to isolation and identification of previously unnoticed organisms, could play a role in this trend, and studies have been published that support [67] and refute [31] this argument. We believe such factors may contribute to but do not explain the significant increases that have been observed. A true increase could be related to the host, the pathogen, or some interaction between the two. Host changes leading to increased susceptibility could play an important role, with increased numbers of patients with inadequate defenses from diseases such as HIV infection, malignancy, or simply advanced age [31]. An increase in susceptibility could also relate to the decrease in infection with two other mycobacteria. It has been speculated that infection with TB [29,38] and Bacillus Calmette-Guerin (BCG) [19,68] may provide cross-immunity protecting against NTM infection. Many investigations have observed decreasing rates of TB concomitant with the increases in NTM. In addition, studies from Sweden [68] and the Czech Republic [19] have found that children who were not vaccinated with BCG had a far higher rate of extrapulmonary NTM infection. Potential changes in the pathogens include increases in NTM virulence, and it has been argued that this should be considered as a possible contributing factor [69]. Finally, an interaction between the host and pathogen could involve a major increase in pathogen exposure or potential inoculum size. This may be occurring secondary to the increase in popularity of showering as a form of bathing [66], a habit that greatly increases respiratory exposure to water contaminants. Several limitations of our review should be noted. We reviewed English-language reports and abstracts, probably leading to fewer data from non-English speaking regions, which may explain the paucity of studies from Africa, Eastern Europe, and most Asian nations. The heterogeneity of study methods in identifying cases and the lack of a uniformly applied definition of disease makes it difficult to compare rates between studies. Finally, the lack of systematic reporting of NTM infection in most nations limits the ability to derive accurate estimates of infection and disease. Regardless, there are more than adequate data to conclude that NTM disease rates vary widely depending on population and geographic location. NTM disease is clearly a major problem in certain groups, including patients with underlying lung disease and also in individuals with impaired immunity. The rates of NTM infection and disease are increasing, so the problem will likely continue to grow and become a far more important issue than current rates suggest.
