This is true for most of the ACS tables, they are often broken down by an additional variable and have to be aggregated to create a desired analysis table.
To take your example, searching for "Median Age" led me to table B01002, "Median Age by Sex":
| B01002 | | | MEDIAN AGE BY SEX |
| B01002 | | | Universe: Total population |
| B01002 | 0.5 | | Median age -- |
| B01002 | 1 | B01002_001 | Total: |
| B01002 | 2 | B01002_002 | Male |
| B01002 | 3 | B01002_003 | Female |
This gives you a median age value for the entire population for a given geography (B01002_001) along with the breakdown by sex. In some cases, such as Table B01001, you have to add variables together to get a total. For example, to get the total number of children under 5, you need to add B01001_003E + B01001_027E. You also have to calculate a revised margin of error --Look in the technical docs to see how to do this.
Unfortunately, THERE IS NO WAY AROUND THIS TYPE OF DATA PROCESSING. Learning how to manipulate the table cells published by the Census Bureau is an essential skill.
Good luck with your project.
AB