If you use the ID of the relevant item directly in the URL, the response will also include the line items of the given documents.
If you use a filter for the ID, for example /skladovy-pohyb/(id = 13).xml?detail=custom:sumCelkem, the output will contain only the document headers, for example:
<winstrom version="1.0">
<skladovy-pohyb>
<id>13</id>
<sumCelkem>3212321.33</sumCelkem>
</skladovy-pohyb>
</winstrom>