How to create a dataframe from data queried from SQL QUERY with the correct data types for each column

I have created a function in python that queries data directly from SQL database using the pyodbc library. The problem is that pd.read_csv in the function generates a dataframe where columns expected to be integers are floats and columns with strings have extra spaces after the column values. How do i resolve this problem?

Here is a sample of the code

import pyodbc
import pandas as pd
def query(sql_query):
    conn = pyodbc(connection details)
    cursor = conn.cursor()
    df = pd.read_csv(sql_query, conn)
    return df

Expected

Output