PVST and Spanning Tree Load Balancing are related.
Traditional Spanning Tree implementations from other vendors commonly allow only one Spanning Tree for the entire network, no matter how many VLANs there are. That means that for any switch with a pair of redundant links, one link is active and carries all the data traffic; while the other link is on standby and carries no data traffic at all, just STP BPDUs. Effectively, the bandwidth of that second link is wasted.
With Cisco's PVST, you can have a different Spanning Tree for each VLAN. If you configure the Spanning Trees on a VLAN-by-VLAN basis, you can engineer it so that the first redundant link carries data traffic for half the VLANs on the switch, and uses the second link as a standby; while the second link can be active and carry data traffic for the remaining VLANs, and can use the first link as its standby. This allows you to get back the use of the bandwidth on that second link.
Distributing the VLANs and Spanning Trees for each across the two redundant links results in a crude but effective form of load balancing. While the traffic loads will never be exactly 50/50 distributed, it's certainly better than 100/0.
If the two redundant links are point-to-point connections between two switches, you can get closer to the ideal 50/50 load balancing by using Cisco EtherChannel or IEEE 802.3ad link aggregation technology. Here, the traffic is sorted out by MAC address, not by VLAN number; so the load balance is more likely to approach 50/50, although there's no guarantee.
Hope this helps.
