EF Core overhead?

I have a code-first EF Core codebase running on SqlLite with one entity with about 20 columns and 10 rows. One page requests this entity in code like this:

var start = DateTime.Now;
var entA = await _context.entA.FromSql("SELECT * FROM entA "+
    "where (Name like '%_Demo' or EXISTS(SELECT NULL FROM entB WHERE StudyId = entA.ID and email = @email)) AND IsDeleted = 0 "+
    "order by colC COLLATE NOCASE",
            new SqliteParameter("@email", User.Identity.Name)) 
    .ToListAsync();
log.Info($"User {User.Identity.Name} requested entA, in {DateTime.Now.Subtract(start).TotalMilliseconds}");

I have EF Core logging switched on and I see the query with

Executed DbCommand (2ms) [Parameters=[@email='?'], CommandType='Text', CommandTimeout='30']

However in the logs (even after runnning it multiple times) I see

2018-06-28 15:16:49,442 INFO Controller - User user@domain.com requested entA, in 166,0095 

What is happening for the ~160ms that are not spent executing the query?

Update: As suggested by Xanatos I've removed the await and async and luckily I'm seeing more or less the same numbers

INFO Controller - User user@domain.com requested entA, in 207,0118 
INFO Controller - User user@domain.com requested entA, in 164,0094 
INFO Controller - User user@domain.com requested entA, in 225,0129 
INFO Controller - User user@domain.com requested entA, in 180,0103 

It would be pretty sad if .net was waiting for 100ms before returning results of a Task

EDIT: I've found out that switching tracking off speeds things up considerably for read-only queries:

_context.ChangeTracker.QueryTrackingBehavior = QueryTrackingBehavior.NoTracking;