Evidence-gap map · non-drug interventions

Type 2 Diabetes Mellitus evidence gaps

677 registered non-drug trials, mapped by approach and outcome.

  1. Research Tracker
  2. Evidence-gap maps
  3. Type 2 Diabetes Mellitus

Which non-drug approaches have been tested for Type 2 Diabetes Mellitus, and which outcomes they were tested on. An empty cell means no matching trial was found among those searched: a possible gap, not proof that nobody has asked the question.

677registered non-drug trials
61with results posted
215recruiting as of 2026-11-16
114group × outcome cells with at least one trial
⚠️
Research information, not medical advice. These pages index what has been studied in registered trials. A trial existing, completing or posting results says nothing on its own about whether an approach works or is safe for you. Supplements and devices can interact with medicines. Discuss any change with your own clinician.

Evidence-gap matrix

Each number is the count of trials; "with results" is how many of them have posted results on ClinicalTrials.gov. A trial is counted once per group and per outcome domain it covers. The number in brackets after each group is its total trials for this condition. The registry search returned 4,934 trials, more than the 1,000-trial cap, so this map covers the most recently updated 1,000.

none1–23–910–2930 or more
Registered non-drug trials for Type 2 Diabetes Mellitus by intervention group and primary outcome domain
Intervention groupFatigue and energyPost-exertional symptomsCognitionMood, anxiety and wellbeingPainSleepPhysical function and exercise capacityHeart rate, blood pressure and autonomicBreathing and lung functionQuality of lifeWeight and metabolic markersInflammation and other biomarkersSymptom severity and disease activityClinical events, survival and healthcare useFeasibility, adherence and safetyNot classified
Exercise and physical activity (96)1·81 with results61 with results21281 with results81 with results43381 with results195111142 with results
Rehabilitation and therapy services (3)····1·2····11···
Breathing and respiratory training (3)1···1·11········
Mind-body and psychological approaches (38)···22142·4101 with results41 with results21 with results·102 with results51 with results
Cognitive training and education (118)··2112191 with results3·5572 with results303 with results721 with results262 with results17
Diet, fasting and nutrition (89)1·621··51 with results11623 with results271 with results62113 with results4
Sleep and circadian approaches (6)····12····42··2·
Neuromodulation and electrical stimulation (4)··2·······21····
Oxygen, light, temperature and manual therapies (16)1·1·3·1···91 with results1··41
Supplements and nutraceuticals (117)1·102 with results81 with results5·4111 with results11704 with results412 with results331 with results5111 with results
Other devices and digital tools (80)··321·11··627 with results922121 with results121 with results
Care delivery and health services (18)···2··11··71 with results5·124
Other behavioural, not yet classified (181)2·3171 with results4·9151 with results261 with results9119 with results343 with results121 with results82 with results404 with results232 with results

The table scrolls sideways: 16 outcome columns.

Where the gaps are

This condition has 4,934 matching trials, more than the 1,000 this map covers, so it cannot show that an outcome was never measured. Only studied-but-unreported questions are listed.

Questions studied but with no results posted

  • Exercise and physical activity × inflammation and other biomarkers: 19 trials registered, none with results posted (7 still recruiting)
  • Exercise and physical activity × feasibility, adherence and safety: 11 trials registered, none with results posted (5 still recruiting)
  • Cognitive training and education × mood, anxiety and wellbeing: 11 trials registered, none with results posted (3 still recruiting)
  • Other devices and digital tools × inflammation and other biomarkers: 9 trials registered, none with results posted (2 still recruiting)
  • Cognitive training and education × symptom severity and disease activity: 7 trials registered, none with results posted (4 still recruiting)
  • Diet, fasting and nutrition × cognition: 6 trials registered, none with results posted (2 still recruiting)
  • Diet, fasting and nutrition × symptom severity and disease activity: 6 trials registered, none with results posted (4 still recruiting)
  • Exercise and physical activity × symptom severity and disease activity: 5 trials registered, none with results posted (3 still recruiting)
  • Cognitive training and education × quality of life: 5 trials registered, none with results posted (5 still recruiting)
  • Supplements and nutraceuticals × pain: 5 trials registered, none with results posted (2 still recruiting)
  • Supplements and nutraceuticals × feasibility, adherence and safety: 5 trials registered, none with results posted (1 still recruiting)
  • Care delivery and health services × inflammation and other biomarkers: 5 trials registered, none with results posted (3 still recruiting)
  • Exercise and physical activity × breathing and lung function: 4 trials registered, none with results posted (1 still recruiting)
  • Mind-body and psychological approaches × physical function and exercise capacity: 4 trials registered, none with results posted (2 still recruiting)

Design a trial with DecenTrial (beta) Fund a component of this work

A gap here means no matching trial was found among those searched, not that an approach would work. About one trial in six has primary outcomes the rules cannot classify, so a few gaps may be measured under another name. Some gaps are gaps for good reason.

Most-studied approaches for Type 2 Diabetes Mellitus

Download this matrix (CSV) Non-drug evidence map Evidence brief RepurpOS condition page Condition page

How outcomes are grouped

Each trial's primary outcome measures are sorted into domains by keyword rules. 82 trials for this condition have primary outcomes the rules could not classify; they appear in the last column rather than being dropped. Checked against 90 hand-labelled outcomes; on a held-out sample of 30, the rules agreed with the hand label for 20 before adjustment, and every outcome they labelled was labelled correctly (misses were left unclassified).

Trial selection and exclusions follow the non-drug evidence map method. Source: ClinicalTrials.gov, searched 2026-11-16.