In a randomized trial of 255 participants with early-stage T2D, participants will be randomized to 1 of 3 groups: Standardized, Personalized, or a Usual Care Control (UCC). In the first phase, participants will be randomized with equal allocation to these 3 groups. In the second phase (current phase), the remaining participants will be randomized with equal allocation to the Standardized and UCC groups.

Type 2 diabetes (T2D) is an epidemic in the United States, affecting 30.2 million people. Vascular complications of T2D (heart and kidney disease, stroke, retinopathy, and amputation) are associated with poor glycemic control. However, the results of recent clinical trials (ACCORD, ADVANCE, VADT) to aggressively reduce HbA1c with medications have not resulted in cardiovascular benefits and may even be harmful due to risks of polypharmacy. Although HbA1c is directly associated with vascular complications, there is growing evidence that glycemic variability (or GV, defined by postprandial excursions and hypoglycemic nadirs) may be a better treatment target.

Postprandial glycemic excursions are primarily driven by diet. However, the results of dietary intervention studies intended to control postprandial excursions in T2D patients have been mixed and mostly negative. Importantly, these studies used one-size-fits-all dietary strategies that did not consider the fact that individuals vary greatly in their glycemic response to same foods. One-size-fits-all approaches are frustrating to T2D patients who have glycemic excursions despite their best efforts. And failure to adequately manage glycemia in the early stages of the disease has long-term vascular consequences that are not entirely remedied by subsequent, better glycemic control.

Personalized medicine is defined as "the right treatment for the right person at the right time" and has grown out of dramatic advances in genetic testing, molecular profiling, and mobile health (mHealth) technology. Personalizing dietary recommendations to the patient's unique glycemic response to food, using an algorithm derived from the gut microbiome to predict postprandial glycemic response (PPGR), is a proactive approach to early T2D dietary management that could increase mastery and self-management success beyond what can be achieved through a one-size-fits-all diet. The investigators also argue that such an approach could reduce glycemic exposure, preserve β-cell function, and reduce downstream metabolic consequences of T2D.

The purpose of this clinical trial is to determine the efficacy of a Personalized behavioral approach for dietary management of early-stage T2D, versus a Standardized behavioral intervention (which uses one-size-fits-all dietary recommendations), versus a UCC.

Primary aims. Compared to UCC, the study will determine the incremental benefits of Standardized and Personalized interventions on mean amplitude of glycemic excursion (or MAGE, the most frequently used measure of GV).

Hypothesis 1: MAGE-Personalized< MAGE-Standardized < MAGE-UCC at 6 months.

Secondary aim. At each time point, the study will describe the impact of the intervention on HbA1c.

Exploratory aims. The study will describe between-group differences in β-cell deterioration (HOMA-β) and the need to escalate the medication regimen. The study also will describe the mediating effects of changes in self-efficacy on the relationship between randomization assignment and GV, HbA1c, and HOMA-β. The study will describe the impact of the interventions on alternative measures of GV (standard deviation, coefficient of variation, Continuous Overall Net Glycemic Action, area under the curve, and frequency of out-of-range and seriously out-of-range glucose values).

Study Population

The study population includes adult men and women, 21-80 years of age, living in the New York City Metropolitan area and who have early-stage type 2 diabetes (defined as an HbA1c<8.0% and managed on lifestyle alone or lifestyle+metformin).

Description of Sites/Facilities Enrolling Participants/Sample size and Randomization

The study will enroll participants from NYU Langone Faculty Group Practices and NYU Langone affiliates located in the New York City metropolitan area. The study expects to enroll 300 individuals, and randomize 255. Participants will be stratified by medication regimen (metformin versus no metformin) and randomized within strata.

Description of Study Intervention/Experimental Manipulation

Eligible participants will be stratified by treatment with metformin (yes or no) and randomized with equal allocation to UCC, Standardized, or Personalized. All 3 groups will receive usual care plus education regarding the nature of T2D, consequences of uncontrolled glycemia, performance and interpretation of capillary glucose monitoring, and the importance of medication adherence. Participants randomized to Standardized will also experience Social Cognitive Theory (SCT)-based behavioral counseling to follow an isocaloric Mediterranean diet. Participants randomized to Personalized will receive the same SCT-based counseling as Standardized, plus personalized guidance to minimize PPGR, using a gut microbiome-based algorithm.

Measurements will occur at 0, 3, and 6 months. [screening A1c usually takes place before T=0]

Inclusion Criteria:

• must be an adult 21-80 years of age
• must be diagnosed with early-stage T2D defined as HbA1c<8% and managed for the past 3 months on a diabetes regimen that included lifestyle plus metformin.
• those who are willing and able to use a smart phone to self monitor their diet and to attend WebEx sessions

Exclusion Criteria:

• those who have conditions or treatments likely to alter the underlying function of the microbiome, an insulin response that is driven by factors other than glycemic response to food, conditions/treatments that make it difficult to isolate the true nature of the relationship between randomization assignment and weight loss, characteristics or preferences that would preclude meaningful participation in the study
• those who are unable or unwilling to adhere to an intervention that requires dietary self-monitoring
• those who have been prescribed: (1) antibiotics or antifungals in the past 3 months, (2) diabetic medications other than metformin, (3) weight loss medications, (4) chronic use of steroids or immunosuppressants, (5) atypical antipsychotics, and (6) chemotherapy.
• those who are pregnant, planning to become pregnant during the study period, or become pregnant during the study
• those who have a chronic disease that affects energy/glucose metabolism (e.g., Cushing's syndrome, acromegaly, hyperthyroidism)
• those who require special dietary management (end-stage kidney disease, cirrhosis, HIV)
• those who are unable or unwilling to provide informed consent
• those who are unable to participate meaningfully in an intervention that involves self-monitoring using software available in English (e.g., uncorrected sight impairment, illiterate, non-English-speaking, dementia)
• those who are unwilling to accept randomization assignment
• those who have limited control over diet (e.g., are homeless or institutionalized, in a nursing home or personal care facility, or incarcerated)
• those who have previously had bariatric surgery, or are unwilling to delay bariatric surgery for the next 7 months
• those who are unable to walk without a walker or cane for 2 city blocks
• those who have been diagnosed with a chronic active inflammatory or neoplastic disease in the past 3 years
• those who have been diagnosed with a chronic gastrointestinal disorder (e.g., inflammatory bowel disease or celiac disease)
• those who have an active substance use disorder

Individual participant data that underlie the results reported in this article, after deidentification (text, tables, figures, and appendices) will be shared.

Beginning 9 months and ending 36 months following article publication or as required by a condition of awards and agreements supporting the research.

The investigator who proposed to use the data will have access to the data upon reasonable request. Requests should be directed to mary.sevick@nyulangone.org. To gain access, data requestors will need to sign a data access agreement.

ll randomized participants that attended at least one intervention session were included in the baseline data analysis. The data were analyzed based on intent-to-treat as specified by the protocol.

6 months

Collection approach for for potential SAE, AE, and all-cause mortality in this study was both systematic (assessed at each timepoint) and non-systematic (self-report by participants between assessment timepoints).

Randomization is performed by the statistician blind to participant identity and baseline data. It is not possible to blind interventionists to randomization group. Outcome assessors will be blind to randomization assignment to the extent possible.