Latency handling with Istio Mesh

We have deployed Istio 1.10 on Kubernetes Cluster using the helm chart.
After the implementation of Istio mesh, we saw that Latency has increased for each API call.
Let me know the best practices to minimize the latency.