This paper proposes an innovative nabla fractional distributed optimization algorithm to address the constrained distributed optimization problem in multi-agent networks. The proposed algorithm addresses the unbalanced directed communication topology. Compared to traditional discrete-time distributed optimization algorithms, the algorithms in this paper extend the integer update order to fractional order, replace the delta difference with the more advantageous nabla difference, and utilize the recently developed nabla fractional order system theory for algorithm analysis. The effectiveness of the developed algorithm in finding the optimal solution is validated through simulation results.