5 RESULTS
5.3 Enterovirus RNA in Blood Is Associated with the Initiation of Islet
A possible risk association between the detection of enterovirus RNA in serum and later development of T1D was evaluated in a nested case-control study carried out in the prospective birth cohort study in Finland (DIPP study). The study cohort consisted of 38 children who developed T1D during the follow-up and 140 autoantibody negative and non-diabetic control children who were pair-wise matched with case children. The proportion of children who were positive for enterovirus RNA in serum at least once during the follow-up was significantly higher in the case group, compared to the control group (32% vs. 14% of children were at least once positive for enterovirus, respectively; p<0,002). In the group of case children a total of 5,1% of the samples (17/333) were enterovirus positive compared to 1,9% of the samples (19/993) in control children (p<0,01).
The study design made it possible to analyze the frequency of enterovirus RNA positivity in different stages of the autoimmune process. In the case children, enterovirus RNA was detected more frequently in the time period ranging from birth to onset of T1D, compared to the corresponding period in the control group (OR=4,7; 95% CI 1,9-12,0). Overall, the frequency of enterovirus positivity was the lowest in the time period ranging from birth to the 6 months long period preceding the first detection of autoantibodies when 2,4% of the samples were positive in the case and 0,7% in the control group (OR=3,1; 95% CI 0,4-22,4). On the other hand, enterovirus positivity peaked in the case group during a period of six months preceding the first autoantibody positive sample, and at the time point 15,2% of the samples in the case group and 3,3% of samples in the control group were enterovirus positive (OR 7,7; 95% CI 1,9-31,5). During the time period ranging from autoantibody seroconversion to diagnosis of T1D, altogether 3,9% of the samples were enterovirus positive in case children and 2,2% in the control children (OR=3,2 [95% IC 1,1-8,9]) (Fig. 9).
Figure 9. The association between the detection of enterovirus RNA in serum and the development of clinical T1D in different stages of the disease process. The figure shows odds ratios (black squares) and 95% confidence intervals (error bars). “Birth-T1D” = whole time from birth to onset of T1D; “Birth-before AAb” = time from birth to autoantibody seroconversion; “Birth-before 6 months period prior AAb” = time from birth to six months before autoantibody seroconversion; “6 month period prior AAb” = time window six months before autoantibody seroconversion; “AAb-T1D” = period from autoantibody seroconversion to onset of T1D. Figure A represents the whole cohort, figure B shows the data for boys and figure C for girls. (Adapted from Table 1 in Report III)
0,1 1,0 10,0 100,0 1000,0
Birth-T1D Birth-before Aab Birth-before 6 months
window 6 months window Aab-T1D
0,1 1,0 10,0 100,0 1000,0
Birth-T1D Birth-before Aab Birth-before 6 months
window 6 months window Aab-T1D
0,1 1,0 10,0 100,0 1000,0
Birth-T1D Birth-before AAb
Birth-before 6 months period prior AAb
6 months period prior
AAb
AAb-T1D OR
OR OR
A
B
C
The sliding window analysis was applied to carry out time-dependent analyses of the association between enterovirus RNA positivity and appearance of islet autoantibodies. The analysis with the zero point anchored to the first autoantibody positive sample indicated the same peak in enterovirus RNA, around 6 months before autoantibody seroconversion among the case children (OR=5,6; 95% IC 1,8-18,0). At other time points such clustering of infections was not seen (Fig. 10).
Analogously, when the zero-point was anchored to the onset of clinical T1D, a clear peak in enterovirus positivity was seen in case children about two years before the onset of diabetes. No such peaks occurred in the control group.
Figure 10. The occurrence of enterovirus RNA in serum across different stages of the autoimmune process in the “sliding window” analysis. Enterovirus positivity peaked six months prior to the detection of first autoantibodies. X-axis indicates time from the first autoantibody positive sample (time point 0) and y-axis the proportion of enterovirus positive samples by calculating moving average for each six-month window. Figure (A) presents the whole cohort, figure (B) boys and figure (C) girls. Black squares = case children. White squares = control children.
0 5 10 15 20
-24 -18 -12 -6 0 6 12 18 24 30
Proportion of EV+ samples (%)
Time (months)
A
0 5 10 15 20
-24 -18 -12 -6 0 6 12 18 24 30
Proportion of EV+ samples (%)
Time (months)
B
0 5 10 15 20
-24 -18 -12 -6 0 6 12 18 24 30
Proportion of EV+ samples (%)
Time (months)
C
Enterovirus positive samples were equally frequent in boys and girls (2,5% vs. 3,1%) but the risk association between enteroviruses positivity and T1D was stronger among boys than among girls (OR=10,5; 95% CI 2,2-51,2 vs. OR=2,1; 95% IC 0,8-5,3, respectively). The stronger risk association among boys was similarly seen in infections occurring before and after autoantibody seroconversion. The highest risk was related to infections occurring in boys during the 6-month period prior to the first autoantibody positive sample (OR=18,2; 95% CI 2,0-164,5 p=0,01, in boys vs.
OR=3,1; 95% IC 0,4-21,8 P<0,26, in girls).
The age of the child influenced the frequency of enterovirus RNA in serum samples. The lowest frequency of positive samples (1,0%) was found among children younger than six months, while the rate of positivity increased to 3,5% at the age of 6-18 months and to 5,0% in children older than 18 months old. At the age of two years the frequency of virus positive samples dropped to 4,3% and further to 2,0%
in children older than two years (Fig. 11). The first samples which were enterovirus positive were taken at a younger age in the case children compared to the control children (median age of first enterovirus positive sample was 10 vs. 16 months, respectively).
Figure 11. The proportion (%) of enterovirus positive serum samples according to the age of the child. The number of analyzed samples is shown above the columns. (Adapted from Figure 2A in Report III)
0 1 2 3 4 5 6 7 8
0 1 2 3 4 5
Enterovirus positive samples (%)
Age (years)
Samples (N)
209 201 210 108 207 93 182 95 185 89 105 51 75 36 37
Four children (three cases and one control) had more than one enterovirus positive sample (two of these children had two positive samples and another two children had three positive samples). No signs of persistent infection were observed, because none of these samples were subsequent and also virus negative samples occurred between the positive ones. In addition, different virus genotypes were present in serially positive samples in each of these children.
Enterovirus RNA was detected during different seasons of the year, but the frequency showed some variation. The samples taken during the spring months (February-May) were characterized by low frequency of the virus (1,4 - 2,2%), while the rate of positivity started to increase in August peaking to 5,6% in September (Fig.
12).
Figure 12. Proportion (%) of enterovirus positive serum samples according to the calendar month of sample collection.