may change considerably, as illustrated in Figure 6 in Bavaria.
Fig. 6. Mortality rates for all forms of tuberculosis by sex and age for Bavaria (1938 and 1946) and United States of America (1939–1941). a Males, b females. Fig. 2 from [9].
Age-specific rates for each sex for the years 1938–1946 were compared with the corresponding rates for the white population of the United States averaged for 1939–1941 [9]. “In 1938, Bavaria’s age distribution of mortality in both sexes was roughly parallel to, although somewhat higher than that in the United States. In 1946 the TB mortality for males in Bavaria was greatly increased in infancy and in all adult age groups (Fig. 6a). The most marked excess was in the 20–34 year age span and it was assumed that the excessive mortality among repatriated prisoners has contributed to this high rate. Roentgen surveys of such groups have revealed a high prevalence of significant TB. Among females, on the contrary, the only population groups showing an increase in 1946 as compared with 1938 were infants and young children, among whom a rise similar to the male rise was recorded (Fig. 6b). There was practically no difference between prewar and postwar mortality in adult females. Presumably had there been no war, the 1946 rates would have shown a decrease, but at least there was no rise. The increase in infant mortality from TB in both sexes suggests that, at least in this age group, the frequency of infection was greater in 1946 than before the war.” One reason may have been the imperfect or absent pasteurization of the milk [32].
As an additional explanation, it was assumed that there was a sex selection process during the war among the civilian population. Healthy males were drafted to the army, were killed or returned sick – in particular, after war captivity – whereas the non-drafted males had prevalent TB [14].
Tobacco Smoking
A potentially large population-attributable fraction of TB might be related to smoking. However, smoking cannot be clearly established as a risk factor for TB because of the multitude of potentially confounding factors (notably socio-economic ones which are co-shaping the TB epidemic, independent of smoking habits) that tend to interfere with a clean etiologic analysis. An early case-control study nested into the large British cohort study on smoking examined the relationship between smoking and TB and found a dose-relationship, a finding that always enhances the credibility of and strengthens the case for postulating causality [33]. Since this early study, a multitude of cohort studies have added to the evidence base and established increasingly the veracity of establishing smoking as an independent risk factor for the progression from latent infection with M. tuberculosis to TB. The potential of increased smoking burden contributing to increased TB incidence is thus discussed in the following. However, it should be kept in mind that the overall differential population-attributable fraction might have been fairly modest.
Tobacco smoking is a risk factor for both pulmonary and extrapulmonary TB [34, 35]. Several pathogenetic mechanisms have been discussed, though the main cause is most likely the suppression of the immune defense against M. tuberculosis[36]. Cigarette consumption increased considerably during WWI and even more so in WWII, in particular in the military [37, 38]. Thus, it is possible that active and passive cigarette smoking played an additional role in the TB epidemic during WWII, although this is not mentioned specifically in the literature.
Indoor Air Pollution
Indoor air pollution may be a further risk factor for TB and, although often linked with passive smoking, emerges as an independent risk factor in some epidemiological studies [39]. However, in a recent systematic review and meta-analysis, the level of evidence for the association between domestic use of solid fuels and TB was found to be very low [40]. Authors from India observed that using solid cooking fuel was associated with a 1.8 higher risk for bacteriologically positive pulmonary TB [41]. Although this again is not mentioned specifically in the literature, it can be assumed that indoor air pollution in limited settings may have played an additional role in the TB exacerbation during WWII.
Alcohol Consumption
A further, difficult to quantify, potential risk factor for TB is alcohol abuse. The epidemiologic difficulties are similar to those of smoking, that is, the direct and indirect potential confounding with socio-economic factors. Such factors and specific behaviors of substance abusers in general may put them at an increased risk of acquisition of infection with tubercle bacilli rather than at an increased risk of progression to TB subsequent to infection. Experimental animal studies seem to establish alcohol as a risk factor for the latter mechanism, although animal experiments with substance abuse are to be viewed particularly critical: most animals quite in contrast to humans – albeit with some notable exceptions – have little “natural” inclination for intoxication. In this context, alcohol abuse is also discussed in the following. Again, the overall differential population-attributable fraction must be fairly small as a substantial quantitative change in massive alcohol abuse at the population level is much less likely and widespread than for instance famine and protein energy malnutrition.
The association of alcohol abuse with TB is well known [42]. However, again an increased consumption of alcohol is not described as a specific risk factor for TB during WWII, but may have been an additional one. Daniels mentions that a fall in alcohol consumption in France due to the introduction of rationing wines and spirits during the war may have been responsible in part for the decline in TB mortality. He underlines this assumption by adding “a marked fall in morbidity and mortality from diseases attributable to alcoholism, a fall not paralleled in any disease other than TB” [3].
Other Factors
Some other factors that increase the risk of progression from latent infection to TB disease are immune deficiencies from any medical cause including diabetes mellitus and immunosuppressive therapy, plus other medical conditions such as chronic renal failure, silicosis, and gastrectomy [7]. Whether these factors played an additional role in the epidemiology of TB during WWII is not specifically mentioned in the literature.
Human Immunodeficiency Virus
Human immunodeficiency virus (HIV) was not yet known during WWII, but the link between TB and HIV is now well established: “The risk of active TB doubles in the first year of HIV co-infection, and the risk of developing active disease in those who have LTBI is on average 10% per year in the course of an untreated HIV infection, HIV-TB co-infected individuals have reduced survival and are at higher risk for subsequent opportunistic infections. In overcrowded