All the studies reporting overall self-management scores were low-quality62 67 69 73

Although with weak design, all four consistently reported higher overall self-management scores with higher PAct scores

• Inconclusive evidence

All the studies available reported a more favourable diet with higher PAct scores,20 48 49 62 73 which indicate consistent results

These studies measured both general diet (healthy eating)20 48 49 62 73 and specific diet (fruits, vegetables and high-fat foods).62 73

Only one study was large and high-quality with a moderate design49

All the others were low-quality.

• Very limited evidence of favourable association between diet and higher PAct scores

All studies that investigated diet also reported higher physical activity with higher PAct scores20 62 65 72–74 with the same spread of results

One addition is a small, low-quality study with a weak design that showed no association.68

• Very limited evidence of favourable association between PAct scores and physical activity

All the studies that investigated the association between smoking and PAct scores had weak designs20 65 72–74

Two large high-quality studies reported no association between smoking and PAct scores.65 72

Two large studies, one high-quality74 and one low-quality62 showed lower levels of smoking with increased PAct scores.

• Very limited evidence of no association between smoking and PAct scores

Only one large high-quality study with a weak design72 reported no association between alcohol consumption and PAct scores.

• Inconclusive evidence

All the studies that investigated medication adherence had weak designs.21 48 62 71 74

The methods used to measure medication adherence were proportion of days covered,71 74 Morisky scale21 and the Summary of Diabetes Self-Care Activities (SDSCA) tool.48 62

Among the high-quality studies, two (one large and one small) reported higher medication adherence with higher PAct,21 71 whereas one large study reported no association.74

• Inconclusive evidence