Query document with very large text field

Hi,

We have indexed 10k documents which have very large text field. The text field of a document is an union of the text in the element pages.
The mapping looks like that

"docs":{
"properties":{
"text":{
"type":"string",
"index":"analyzed",
"term_vector":"yes",
"index_options":"offsets"
}

It takes 4s when searching with text field in the query:
{
"fuzzy_like_this_field":{
"text" : {
"like_text" : query,
"max_query_terms" : 12
}
}
}

And it takes less than 1 s when searching without text field in the query.

Could you help in optimising the queries against the very large text field in this case.

Thanks a lot