Explainability has gained attention to ensure fairness and transparency in machine learning models, providing users with an understanding of artificial intelligence (AI) models. Machine Learning as a Service (MLaaS) platforms offer several methods to explain model outputs. Patel et al. proposed DPGD-Explain, model explanations with differential privacy. They proved that the DPGD-Explain provides secure model explanations in the sense of central differential privacy. Nevertheless, it remains unclear if model explanations with a differential privacy guarantee are vulnerable against the record reconstruction attack given some pairs of input data and model explanations. In this study, we investigate the record reconstruction risk of DPGD-Explain, in terms of privacy budget and quality.