Cytomegalovirus seropositivity is associated with reduced risk of multiple sclerosis—a presymptomatic case–control study, 2021, Grut et al

forestglip

Senior Member (Voting Rights)
Staff member
Cytomegalovirus seropositivity is associated with reduced risk of multiple sclerosis—a presymptomatic case–control study

Viktor Grut, Martin Biström, Jonatan Salzer, Pernilla Stridh, Daniel Jons, Rasmus Gustafsson, Anna Fogdell-Hahn, Jesse Huang, Nicole Brenner, Julia Butt, Noemi Bender, Anna Lindam, Lucia Alonso-Magdalena, Martin Gunnarsson, Magnus Vrethem, Tomas Bergström, Oluf Andersen, Ingrid Kockum, Tim Waterboer, Tomas Olsson, Peter Sundström

First published: 09 June 2021


Background and purpose
Epstein–Barr virus (EBV) and human herpesvirus 6A (HHV-6A) are associated with increased risk of multiple sclerosis (MS). Conversely, infection with cytomegalovirus (CMV) has been suggested to reduce the risk of MS but supporting data from presymptomatic studies are lacking. Here, it was sought to increase the understanding of CMV in MS aetiology.

Methods
A nested case–control study was performed with presymptomatically collected blood samples identified through crosslinkage of MS registries and Swedish biobanks. Serological antibody response against CMV, EBV and HHV-6A was determined using a bead-based multiplex assay. Odds ratio (OR) with 95% confidence interval (CI) for CMV seropositivity as a risk factor for MS was calculated by conditional logistic regression and adjusted for EBV and HHV-6A seropositivity. Potential interactions on the additive scale were analysed by calculating the attributable proportion due to interaction (AP).

Results
Serum samples from 670 pairs of matched cases and controls were included. CMV seropositivity was associated with a reduced risk for MS (OR = 0.70, 95% CI 0.56–0.88, p = 0.003). Statistical interactions on the additive scale were observed between seronegativity for CMV and seropositivity against HHV-6A (AP 0.34, 95% CI 0.06–0.61) and EBV antigen EBNA-1 (amino acid 385–420) at age 20–39 years (AP 0.37, 95% CI 0.09–0.65).

Conclusions
Cytomegalovirus seropositivity is associated with a decreased risk for MS. The protective role for CMV infection in MS aetiology is further supported by the interactions between CMV seronegativity and EBV and HHV-6A seropositivity.

Link | PDF (European Journal of Neurology) [Open Access]
 
They created matched pairs for "biobank, sex, date of blood sampling and date of birth (in order of priority)":
The absolute mean differences for sampling date and sampling age between cases and controls were 6 days and 152 days, respectively.

They also looked at strata based on age (<20 and 20-39) and divided by sex. While it's not significant in age<20, that might be due to the much smaller sample size, and it's at least still the same effect direction.
ene14961-fig-0001-m.png

Screenshot_20250414-192316.png

The association wasn't significant when just looking at males, but they say it may be due to the low number of males.
Sex stratified subgroup analyses were performed, but the low number of male cases (n = 108) limits the possibility of drawing conclusions from such assessments.

It's possible the blood sampling isn't actually from before the MS disease process begins.
Still, some limitations must be acknowledged. The mean time from sampling to MS symptom onset was 8 years. Even this latency might not be sufficient to ensure that samples are presymptomatic, considering the emerging evidence of a long prodromal phase of MS [25, 26]. However, the negative association of CMV serostatus and MS risk remained in the sensitivity analysis of samples drawn more than 8 years before MS onset.

They say their results align with previous research:
The relationship between CMV and MS has been addressed previously in a few presymptomatic serological studies, but none with significant results [13-16]. The previous studies were all relatively small, consisting of 18 to 305 presymptomatic cases, and their lack of significant results might be due to insufficient power. A meta-analysis using raw data from three presymptomatic studies [13, 15, 16] showed a negative association of CMV and MS risk with OR = 0.73, 95% CI 0.59–0.91, p = 0.005 [29]. This is similar to the results from a Swedish study on samples drawn after MS diagnosis, which also reported a negative association of CMV and MS risk with OR = 0.73, 95% CI 0.58–0.92, p = 0.005 [10]. Both results are consistent with that of the present study, where the OR for CMV as a risk factor for MS was 0.70, 95% CI 0.56–0.88, p = 0.003.

Speculation on why CMV might be protective:
The latent CMV infection is suggested to preoccupy a large proportion of the immune system, thereby reducing the adverse immune reaction against EBV that could lead to MS. Alternative mechanisms have also been suggested. For example, CMV infection promotes expansion of a subset of mature natural killer cells, which could modulate the control of EBV [36, 37].
 
The Ascherio study also looked at CMV. I think they found no association. Can that data not be used to analyse some of the ideas present here?
 
The Ascherio study also looked at CMV. I think they found no association. Can that data not be used to analyse some of the ideas present here?
Ascherio 2001 was reference 13 in the above quote. Not significant, possibly due to small sample size, but included in a meta-analysis that did come out significant.

Edit: Added link
 
Last edited:
Here is the previous meta-analysis:

Cytomegalovirus and multiple sclerosis risk (2013, Journal of Neurology)

Ascherio, Levin, and DeLorenze are the three prospective studies here:
Screenshot_20250414-205259.png

In total, 16 studies of 1,341 MS patients and 2,042 control subjects were included [3–18]. [...] The overall OR for MS in CMV seropositive individuals was 1.15 (95 % CI 0.83–1.60), p = 0.39, Fig. 1. Significant heterogeneity was identified, s2= 0.24; v2= 47.48, df = 15 (P < 0.0001); I2= 68 %, but sensitivity analysis through removal of the study by Sanadgol et al. [14] rendered heterogeneity insignificant, s2= 0.05; v2= 20.59, df = 14 (P = 0.11); I2= 32 %, and changed the OR to 0.92 (95 % CI 0.74–1.14), p = 0.42.

However, subgroup analysis of the three prospective studies which together constitute a sample size of 495 MS patients and 1,035 controls generated an OR of 0.73 (95 % CI 0.59–0.91), p = 0.005 and no significant heterogeneity (P = 0.78) [4, 9, 18].
This well powered meta-analysis found no significant difference in the likelihood of an MS patient being CMV seropositive compared to a healthy control upon pooling of all studies to date. However, some evidence for a protective role for CMV on MS risk was found upon inclusion of only prospective studies. Limitations exist: no assay is 100 % sensitive or specific, and there is some remaining heterogeneity and the possibility of bias.
 
Last edited:
Back
Top Bottom