Increased Transmissibility of SARS-CoV-2 Lineage B.1.1.7 by Age and Viral Load: Evidence from Danish Households

Aim The aim of this study was to estimate the household transmissibility of SARSCoV-2 for lineage B.1.1.7 compared with other lineages, by age and viral load. Furthermore, we wanted to estimate whether there is a multiplicative or additive effect of the increased transmissibility of B.1.1.7 compared with other lineages. Background New lineages of SARS-CoV-2 are of potential concern due to higher transmissibility, risk of severe outcomes, and/or escape from neutralizing antibodies. Lineage B.1.1.7 has been estimated to be more transmissible than other previously known lineages, but the association between transmissibility and risk factors, such as age of primary case and viral load is still unknown. Methods We used comprehensive administrative data from Denmark, comprising the full population, all SARS-CoV-2 RT-PCR tests, and all WGS lineage data (January 11 to February 7, 2021), to estimate household transmissibility stratified by lineage B.1.1.7 and other lineages. Results We included 5,241 households with primary cases; 808 were infected with SARS-CoV-2 lineage B.1.1.7 and 4,433 were infected with other lineages. The attack rate was 38% in households with a primary case infected with B.1.1.7 and 27% in households with a primary case infected with other lineages. Primary cases infected with B.1.1.7 had an increased transmissibility of 1.5-1.7 times that of primary cases infected with other lineages. The increased transmissibility of B.1.1.7 was multiplicative across age and viral load. Conclusions The results found in this study add new knowledge that can be used to mitigate the further spread of SARS-CoV-2 lineage B.1.1.7, which is becoming increasingly widespread in numerous countries. Our results clarify that the transmissibility of B.1.1.7 should be included as a multiplicative effect in mathematical models used as a tool for decision makers. The results may have important public health implications, as household transmission may serve as a bridge between otherwise separate transmission domains, such as schools and physical workplaces, despite implemented non-pharmaceutical interventions.

In our data, not all positive cases have a successfully sequenced genome. This can be 126 due to various reasons, e.g., sequencing capacity constraints. Moreover, the probability of 127 successfully sequencing a genome is correlated with the viral load, which is reflected in the 128 Ct value. Therefore, sample selection bias is a major concern. If some cases have a higher 129 probability of being selected for WGS than others, it can lead to false conclusions. In 130 Appendix A, we provide summary statistics to substantiate our choice of study period. As 131 both viral load (Ct values) and age of the primary case are associated with transmissibility 132 (Lyngse et al., 2021;Lee et al., 2021;Marks et al., 2021), we naturally explored this. We defined primary cases as the first identified RT-PCR positive SARS-CoV-2 case in 135 a household, and any cases that were detected in the same household within the following 136 1-14 days were considered to be secondary cases (see also sensitivity analysis of this below).

137
If more than one person tested positive on the first date, the primary case was randomly 138 selected. We utilized two concepts for transmissibility of the primary case: transmission 139 risk and transmission rate. The transmission risk describes the risk of infecting at least 140 one other person within the household, and equals one if any (one or more) secondary 141 cases are identified within the same household, and zero otherwise. The transmission 142 6 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 rate is the proportion of potential secondary cases within the same household that tested 143 positive. The two transmissibility measures are weighted on the primary case level, such 144 that each primary has a weight of one. 145 Furthermore, we utilized one concept for susceptibility of the potential secondary case: 146 attack rate. The (secondary) attack rate is defined as the proportion of potential sec-147 ondary cases that tested positive. The attack rate is weighted on the potential secondary 148 case level, such that each potential secondary case has a weight of one. 149 We estimated the transmission rate and transmission risk for each 10 year age group  To investigate whether the increased transmissibility of B.1.1.7 compared with other 152 lineages was best described as an additive or multiplicative effect, we compared the model 153 fit of both a linear and a logistic regression analysis, using the Akaike Information Criteria 154 (AIC). 155 We used a logistic regression model to estimate the odds ratio of the transmission rate 156 and transmission risk for B.1.1.7 compared with other lineages. As the transmissibility can be dependent on the age of the primary case, the age of the potential secondary case,  To investigate the robustness of the estimated transmissibility across age groups, we 163 supplemented our main analyses of ten-year age groups with five-year age groups. 164 We estimated the transmission rate and transmission risk by Ct value intervals.

165
The estimates are sensitive to the definition of primary and secondary cases. In our 166 approach, it is possible that a co-primary case may be misclassified as a secondary case, 167 if she is tested positive one or more days later than the first identified case. In order to 168 investigate the robustness of the results to the definition of primary and secondary cases, 169 we additionally analyzed the data defining secondary cases as those that tested positive at 170 7 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 38%, compared with 27% when the primary cases were infected with other lineages, and 198 17% when the primary case did not have a successfully sequenced genome. The age specific transmissibility followed a U shaped pattern with the lowest trans- with other lineages (blue) across all ten-year age groups. The transmissibility was lower 204 for primary cases without a successfully sequenced genome (gray).

205
10 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ; https://doi.org/10.1101/2021.04.16.21255459 doi: medRxiv preprint Notes: The transmission rate describes the proportion of potential secondary cases within the household that were infected. The transmission risk describes the proportion of infected primary cases that infected at least one secondary case. Figure S7 provides the same graphs for five-year age groups. The shaded areas show the 95% confidence bands clustered on the household level.
To investigate whether the increased transmissibility of B.1.1.7 compared with other 206 lineages was best described as an additive or multiplicative effect, we compared the model 207 fit of both a linear and a logistic regression analysis. We compared the fit of the two 208 models using the Akaike Information Criteria (AIC) and found that the logit model was 209 a better fit (Appendix C). This supports the hypothesis that the effect of the increased 210 transmissibility is best described as a multiplicative effect.

211
Using a logit model, we estimated the increased transmission rate and transmission 212 risk for B.1.1.7 compared with other lineages. In Table 3, we present the crude estimates 213 as well as models controlling for age of the primary case, age of the potential secondary 214 cases, and Ct value of the primary case. Primary cases infected with B.1.1.7 were 1.5 215 times more transmissible than primary cases infected with other lineages, without any 216 adjustments. When controlling for age and viral load, this effect was 1.6. 217 11 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. 12 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021.  CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 imply that the transmissibility of B.1.1.7 should be modelled as a multiplicative effect and 260 not an additive effect. This is pivotal for the validity and accuracy of simulations models 261 of the current pandemic, which are used as tools for decision makers.  CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)  Figure S1). Therefore, it is not likely that our findings are an artefact generated by a 290 misleading baseline of other lineages.

291
There is a significant proportion of positive RT-PCR positive samples without a suc-292 cessfully sequenced genome that could not be assigned to specific lineages. This can 293 potentially result in sample selection bias. Samples with low viral load (high Ct values) 294 were less likely to be selected for WGS and successfully sequenced ( Figure S3 and S4).   Some limitations apply to this study. This is a retrospective observational study, 314 therefore causality naturally cannot be inferred. Additionally, we did not have access to 315 data on rapid antigen tests, which have been increasingly used in Denmark since December 316 15 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ; https://doi.org/10.1101/2021.04.16.21255459 doi: medRxiv preprint could not include these as positive cases. Despite of these limitations, we believe that the 319 results of this study provide useful new insights into the transmissibility of B.1.1.7.

320
In summary, we found an attack rate of 38% in households with a primary cases CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021.

21
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Notes: This figure shows the ten most abundant lineages for cases with a complete genome in Denmark during the study period. Less abundant lineages are included in the white space. 14-day rolling average.

22
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

23
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 genome is dependent on the Ct value ( Figure S3)

24
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ; https://doi.org/10.1101/2021.04.16.21255459 doi: medRxiv preprint time ( Figure S4). In week 2 TCDK started to sample systematically and to sample on 449 Ct values. From Figure S4, we see that in week 2, TCDK used a Ct value cut-off of 30, 450 32, and 35. In weeks 3-6, TCDK used a Ct value cut-off of 35. Samples with higher Ct 451 values (35<Ct≤38) were included, when WGS capacity allowed for it. Week 7 Week 6 Week 5 Week 4 Week 3 Week 2 Week 1 Week 53 Week 52   18  22  26  30  34  38  18  22  26  30  34  38  18  22  26  30  34  Notes: This figures shows the proportion of cases selected for WGS and the proportion that were successfully sequenced stratified by the Ct value of the sample, across weeks. Only samples from TCDK are included. An RT-PCR test is positive if the Ct value is ≤38. The shaded areas show the 95% confidence bands.

25
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 (blue) are relatively similar, while samples with no genome (gray) have a distribution with 454 higher Ct values ( Figure S5).

455
WGS was mainly obtained for samples with low Ct values compared with the distri-456 bution of Ct values from the whole population (gray dashed line in Figure S5). We found CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021.

27
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 are relatively similar, although B.1.1.7 seems to mainly infect younger people in weeks 28 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  Notes: The transmission rate describes the proportion of potential secondary cases within the household that were infected. The transmission risk describes the proportion of infected primary cases that infected at least one secondary case. This figure is the same as Figure 1, except that it shows five-year age groups. The shaded areas show the 95% confidence bands clustered on the household level.

30
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

31
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

32
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) where Age p,10 is the age (in ten-year groups) of the primary case. β measures the trans-472 mission rate for each ten-year age group of the primary cases. ε p denotes the error term, 473 clustered on the household (event) level.
while varying the link function to compare the model fit of an additive versus a mul-484 tiplicative effect.

485
As the two models include the same parameters, the model fits can be compared 486 using the Akaike Information Criterion (AIC). Furthermore, reduced versions of the linear 487 predictors were tested. Across all three model specifications and for both transmission rate 488 and transmission risk, we found that the logit model had a lower AIC and, thereby, was a 489 33 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted April 19, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 better fit compared with the identity model, implying that the increased transmissibility 490 is multiplicative and not additive (Table S4 and S5).  We also tested whether other explanatory variables had any significant effect on the 492 increased transmissibility, e.g., household size. Moreover, we investigated the interaction 493 effect, e.g., to see whether the effect was different across age groups. (Data not shown.) 494 34 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)