Late in its life, Mir suffered significant failures, a lot.
Designed for only a five-year life, the aging Mir suffered a series of equipment failures and accidents in 1996–97 but remained in service.
Around the STS-86 time frame, Mir was prone to failures of its attitude control computer.
This made Mir start to rotate and the solar panels lose tracking. Then a massive powerdown was done to keep it from totally losing battery power until the panels could be manually pointed at the sun to charge the batteries. A slow manual recovery process ensued.
a sensor in the Spektr
module failed, prompting the motion control
computer to switch to a backup system. During
the three-minute swap-over, all attitude control
was lost. The crew placed the station in what called "free drift" and then used onboard thruster
jets to stabilize its attitude. For most of that day,
Mir remained in a "gravity gradient," which basically means that the most massive part of Mir
naturally pointed toward Earth. Because this
attitude did not keep tile solar arrays pointed at the Sun, the crew turned off the gyrodynes and
Other equipment to conserve power. Late in the
day, flight controllers uplinked a new attitude
maneuver to the motion control system computer,
and the crew restarted the gyrodvnes. The station's
primary attitude sensor, called Omega, was still
inoperable; so control was managed by a backup
unit until the crew installed a spare Omega sensor
and rerouted cables.
Official NASA Shuttle/Mir History
See also When Mir's attitude control computer failed, why did the station immediately start rotating?
This was happening so much that right before STS-86 launched they had special simulator sessions with the crew and flight control team to see how fast a loss-of-control Mir could be tumbling and still allow the shuttle to dock.
Since the Mir space station had been having
trouble—losing attitude control and drifting
into slow tumble—the STS-86 crew had trained
until almost the last minute, practicing simulations of various docking situations. The crew was
also bringing to Mir a new attitude control computer.
On November 13, Mir suffered a temporary
power loss during a test of the newly installed
solar array on Kvant- I . This caused the shutdown
of the motion control system computer and interrupted Wolf's scientific investigations. The crew
transferred fully charged batteries from the Kristall
module to the Base Block, and they restored
power to five of the II gyrodynes that provided
attitude control to Mir. However, the batteries
alone could not support the many hours required
to stabilize the station. So throughout the week-
end, the three crewmates alternated shifts to monitor systems; and whenever Mir drifted into an
attitude favorable for solar energy collection, they
temporarily powered up the battery chargers.
So while it's not necessarily true that leaving Mir uncrewed would have meant the end, it would have increased the chances of losing it a lot. (There were other problems too, not just the attitude control system. That's just one I'm most familiar with. The Burrough's book Dragonfly is a great history of this period.)