The is likely still the top word.
It's impossible to find this out for certain, since not all spoken language is recorded and corpuses tend to capture established usage. Furthermore, filler (like, oh, um, you know) may be underrepresented in corpuses, and spoken usage in general can vary widely depending on the context. Still, there are two good sources for American and British usage.
Method 1: Corpus search on the spoken subcorpus of COCA.
Result: The. It isn't close, folks. (5000 words lists these words in an accessible format but doesn't separate written and spoken English.)
Limitation: COCA's spoken corpus comes from TV, radio, and sources that privilege standard American dialects in a professional register.
Method 2: Consult Geoffrey Leech, Paul Rayson, Andrew Wilson, authors of Word Frequencies in Written and Spoken English: based on the British National Corpus, 2001.
Result: The (spoken English in quantity, though they cleverly find that "oh" and "yeah" are the most distinctively conversational versus task-oriented speech, and the most common interjections/discourse particles.)
Limitations: British English. Spoken corpus was ~10 million words.