Table 1 Summary of publications that did not satisfy inclusion criteria, but were tangentially related to study research question regarding lifestyle intervention effects on HbA1c in children or youth with type 2 diabetes (T2D).

Zdravkovic et al., 2004 [20]

To review the clinical experience of children and teens diagnosed with T2D at a pediatric hospital serving a large urban multi-ethnic population

Retrospective medical chart review (n = 41)

54% of patients were initially treated by lifestyle (diet and exercise); the minority (3%) did not require short-term intensification (i.e., medication) of therapy at some point since initial diagnosis to help achieve glycemia goals

Low quality study design; variable intervention elements (i.e., modality, duration, intensity); inconsistent outcome evaluation

Grinstein et al., 2003 [21]

To report the presentation and 5-year treatment requirements of African-American and Caribbean-Hispanic adolescents with T2D followed at the Montefiore Medical Centre (Bronx, NY, USA)

Retrospective medical chart review (n = 83)

Most patients (63%) were prescribed oral medication (glipizide and/or metformin) and/or insulin. All patients were referred to a nutritionist for dietary counselling and recommendation of appropriate exercise intervention

Low quality study design; undefined lifestyle intervention elements (i.e., modality, duration, intensity); undocumented independent effects of lifestyle variables on glycemia

Zuhri-Yafi et al., 2002 [22]

To study possible treatment modalities for type 2 diabetes in children and adolescents.

Retrospective medical chart review (n = 25); insulin the only initial treatment in 72% (n = 18); weight management strategies were taught and encouraged; insulin was withdrawn as euglycemia was achieved.

Mean change in HbA1c over 2 years: -2.9% for insulin users only and -2.3% for those treated with insulin and metformin, and -4.4.% for those treated by metformin alone. Few patents sustained any weight loss regardless of treatment.

Low quality study design; no description of diet and physical activity elements; inconsistent results with low statistical power to detect intervention effects.

Rothman et al., 2008 [23]

To examine self-management behaviours and glycemic control among adolescents with T2D

Telephone survey + retrospective medical chart review (n = 103)

Minority (11.8%) used lifestyle changes exclusively as treatment. More than 80% of patients reported ≥75% medication compliance. More than 70% of patients reported exercising ≥2×/week; 68% reported viewing ≥2 hours of television daily. Patients reported frequent episodes of overeating, drinking sugar-sweetened beverages, and eating fast food. Many (37%) reported that 'following diet or exercise regime' was the 'hardest thing about having diabetes'

Data derived from cross-sectional, self-report survey and chart abstraction; non-specific intervention details queried

Reinehr et al., 2008 [24]

To study the 2-year course of children and adolescents with T2D in general practice to present (1) the treatment modalities, (2) patient adherence, and (3) the occurrence of comorbidities.

Clinical data collected prospectively from 1995 to 2003 among 129 children and adolescents with type 2 diabetes from 62 specialized diabetes centers in Germany

Reduction in median HbA1c after 2 years; 60% of children dropped out of the study. Lifestyle intervention as sole treatment was usually not useful for achieving long-term metabolic control.

High number of cases lost to follow-up; poorly defined lifestyle intervention elements.

Shield et al., 2009 [25]

To report the 1-year outcome for children newly diagnosed as having T2D across the UK.

Retrospective medical chart review (n = 73); follow-up occurred one year after incident cases were ascertained

Most common treatment at diagnosis was metformin (n = 34; 47%); lifestyle (diet and physical activity) was the initial treatment for a sub-group (n = 12; 17%); necessary lifestyle changes needed to positively affect metabolic health are not occurring; heterogeneity of treatment regimens appears relatively effective in achieving glycemic control.

Low quality study; inadequate detail regarding lifestyle intervention elements; no direct report of lifestyle behaviours.

Allan et al., 2008 [26]

(1) To assess quality of life (QOL) in youth with type 2 diabetes, (2) to compare youth and parent-proxy perceptions of youth QOL, (3) to determine if youth QOL is associated with diabetes control, and (4) to determine if demographic and/or medical history is associated with youth QOL and/or diabetes control.

Cross-sectional survey and clinical data collection among First Nation youth aged 7-18 years and their parents at a regional diabetes program; 39% received lifestyle counselling alone with remainder receiving either insulin monotherapy or combination with oral hypoglycmeic agent; mean HbA1c 9.2 ± 2.9%

Youth reported higher scores in the generic and diabetes related domains compared to parents. Youth not taking diabetes medications reported higher QOL. QOL may be affected by specific demographic and clinical factors to reduce the psychosocial burden of their disease.

Low quality study; inadequate detail regarding lifestyle intervention elements.

Anderson & Dean (1990) [27]

To investigate the effect of a regulated food intake and daily exercise program on blood glucose levels, total glycosylated hemoglobin and weight status over a 3-year period.

Case study (n = 1); subject attended a month-long summer camp for three successive years.

Average HbA1c over 29 months was 18.4% with the lowest value of 16.9% occurring after summer camp. Controlled food intake and daily exercise improves glycemic control and in controlled environment only.

Low quality study design; small sample size limits generalizability; Glycemic target difficult to reach with intensive lifestyle intervention.

Nichol et al., 2008 [28]

(1) To modify the Canadian Diabetes Association Pacific Area's Cooking For Your Life! Program for youth with T2D (or impaired glucose tolerance) and their families; (2) to evaluate program satisfaction

Pilot study (n = 15 adolescents; n = 21 family members); intervention included three 'hands-on' cooking classes + one grocery store tour

Three out of four sessions were attended by 86% of participants; 90% were 'mostly satisfied' or 'completely satisfied' with the program

Low quality study design; study not designed to impact glycemic control; study focussed on feasibility and process-related outcomes