Why is Kubernetes' pod CPU and memory load uneven?

I'm running a Kubernetes Deployment with four Pods. Each Pod is expected to share the load equally, but I've noticed that one Pod consistently uses significantly more CPU and memory than the others. I suspect this issue might be related to kube-proxy's iptables mode, but I'm not sure of the exact cause.

[root@iac ~]# kubectl top pod --sort-by=memory | grep api
api-5cd64b46cb-b9ffs                        183m         1220Mi   
api-5cd64b46cb-4m8h6                        172m         952Mi      
api-5cd64b46cb-hrnhw                        168m         939Mi           
api-5cd64b46cb-kbm7d                        215m         895Mi

I’ve checked the resource limits and requests for the pods, and they’re all configured identically.