Abstract:Insulin is a cornerstone in the treatment of diabetes, being life-saving for patients with type 1 diabetes and a therapeutic option for patients with type 2 diabetes. In the latter, insulin treatment is usually implemented either to treat metabolic decompensation at disease onset or in patients with long-standing disease in whom insulin secretory capacity is unresponsive to therapy.1 Insulin, a hormone with a considerable number of pleiotropic actions beyond its ability to regulate glucose levels, is lipogenic, exerts a powerful anabolic activity, and stimulates sodium reabsorption by the distal nephron. As occurs in the majority of patients with type 2 diabetes, when obesity develops beta cells are forced to secrete more insulin to keep blood glucose levels normal. This situation, mostly determined by the expansion of fat tissue, is identified as insulin resistance.2 In this condition, circulating insulin levels are significantly higher and become even higher during exogenous insulin treatment. Despite this, and when necessary, insulin therapy becomes essential to fully accomplish the metabolic effects of the hormone. Thus, paradoxically, human beings have converted a physiologic hormone into a pathologic one with their wrong lifestyle and the subsequent expansion of fat mass.3 This results in weight gain, sodium retention, and noradrenergic activation, which all, along with insulin-induced hypoglycaemia, are important risk factors for heart failure (HF). It is therefore not surprising that insulin treatment may be associated with an increased odd of congestion and hospitalization for HF, especially if the patient’s metabolic control is not at target and repeated hypoglycaemic episodes occur.4 With this background in mind, Cosmi and colleagues assessed in four randomized controlled trials (RCTs) and in a real-world survey whether insulin treatment was associated with an increased risk of hospitalization for HF.5 Given that patients receiving insulin are usually older, co-morbid, and with longer duration of the disease than those who are not on insulin, the authors used propensity

Are Training Samples Correlated? Learning to Generate Dialogue Responses with Multiple References

Deep Reinforcement Learning for Dialogue Generation

Adversarial Learning for Neural Dialogue Generation.

Generating More Interesting Responses in Neural Conversation Models with Distributional Constraints

Towards Robust Online Dialogue Response Generation

Two Are Better Than One: an Ensemble of Retrieval- and Generation-Based Dialog Systems.

Open Domain Response Generation Guided by Retrieved Conversations

Multi-Domain Dialogue Acts and Response Co-Generation

CORE: Cooperative Training of Retriever-Reranker for Effective Dialogue Response Selection

A Survey on Response Selection for Retrieval-based Dialogues.

Can Neural Generators for Dialogue Learn Sentence Planning and Discourse Structuring?

Personalized Dialogue Response Generation Learned from Monologues

Learning to Know Myself: A Coarse-to-Fine Persona-Aware Training Framework for Personalized Dialogue Generation

Structured and Natural Responses Co-generation for Conversational Search

Chinese Emotional Dialogue Response Generation via Reinforcement Learning

More is Better: Enhancing Open-Domain Dialogue Generation via Multi-Source Heterogeneous Knowledge

There Is No Standard Answer: Knowledge-Grounded Dialogue Generation with Adversarial Activated Multi-Reference Learning

Know Deeper: Knowledge-Conversation Cyclic Utilization Mechanism for Open-domain Dialogue Generation

Insulin treatment in patients with diabetes and heart failure: defendant on the stand

Training Neural Response Selection for Task-Oriented Dialogue Systems

Exploring Dense Retrieval for Dialogue Response Selection