There is no Hypothesis formulated as such. It is a retrospective cohort analysis which gives a summary of the trends. It postulates though that studies of this nature would help in postulating the as yet not known causes of such persistent incidence besides the reactivation of old infections and disease spread by normal transmission.
The authors include Annelies M Vos who is the first author, a PhD candidate at Erasmus MC at Rotterdam accompanied by Abraham Meima. On further research it is found that this is the first publication of Ms.Annelies whereas Abrahm Meima has done substantial work on Leprosy which could be discovered at (1) Another co-author Vivian Bos has worked on mortality in various socio-economic and ethnic groups in Netherlands. Borgdorff is associated with the KNCV Tuberculosis Foundation and has many publications regarding the etiology and epidemiology of Tuberculosis to credit.
Research sample included incident cases of tuberculosis which occurred more than six months after immigration so as to completely exclude the prevalent cases. According to the statistics mentioned, immigrant population formed a whopping 1/8th of the total population of Netherlands. Data was obtained from standard sources thus ascertaining its authenticity. The period of data collection was four years that between 1996-2000. The study combines "data on all immigrant patients in whom tuberculosis was detected and all legal immigrants present in a 5-year period in a low incidence country, enabling detailed analysis with a long follow-up period" as quoted by the author. Thus the representativeness of the sample is strong as it includes all the incident cases. There was no scope for random allocation in this study as the cohort was defined by including all the cases. Bias was however excluded by not directly comparing the data obtained from Organisation of Asylum Seekers (COA) and from Municipal Population Registrars (GBA). Also bias in the missing values of country of origin and time since migration was avoided by multiple imputation and substitution of values by average in the datasets. The sample size was 2661 patients whichis actually the entire affected population.
Observations include highest rate of incidence among African immigrants especially Somalis. While Somalis had an incidence rate of 379/100,000, the indigenous population had only 3/100,000. Univariate and multivariate poisson regression is used as a tool which enables a log-linear transition and allows to get a large sample size data on a linear scale.
As mentioned, only 42% showed lower incidence rates 9.5-19.4 years after immigration. Children showed lower incidence of infection. "In univariate analysis, incidence rate ratios in adults decreased with age, whereas in multivariate analysis the oldest age group had an increased risk." This is a problem of confounding. In confounding the exposed/infected population is compared to a substitute population which is not totally representative. Thus as shown by the author itself, off the African immigrants, older generation formed a small population which too had migrated recently.
Thus the study discusses that pulmonary tuberculosis is highly incident inpopulations even a decade after migration which is consistent with previous works published on the same matter. The author has put forth various reasons for persisted incidence and no sharp decline after immigration.