Table 2.
Summary of the studies on the effect of weight change in micro and macro vascular complications in T2DM patients
| First author | Year | Type of study | Follow-up duration | Participants: T2DM patients | Gender | Mean age (years old) | Weight change/definition | Weight change assessment method | Outcomes | Adjusted variables | Quality score | Ref. |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Chaturvedi | 1995 | Cohort study | 13 years | 2,960 | 52% female | 47 ± 2 | Gain > 2 kg/m3 Loss > 2 kg/m3 |
Measured | Weight loss at BMI < 26 kg/m2: mortality rate (RR: 3.05; 95% CI: 1.26, 7.36) Weight loss at BMI > 29 kg/m2: mortality rate (RR: 0.84; 95% CI: 0.40, 1.74) |
Age-blood pressure-cholesterol-duration of diabetes-retinopathy-smoking status-blood glucose-insulin therapy | 7 | 16 |
| Cho | 2002 | Cohort study | 20 years | 5,897 | 100% female | 55 ± 1 | Weight loss ≥ 5 Loss = 4.9-Gain = 4.9 Weight gain = 5–7.9 Weight gain = 8–10.9 Weight gain = 11–19.9 Weight gain = 20–29.9 Weight gain = 30–39.9 Weight gain ≥ 40 |
Self-reported and measured | CHD incidence risk increased as follows: Weight loss ≥ 5: RR≃1.4 Loss = 4.9-Gain = 4.9: RR≃1 Weight gain = 5–7.9: RR≃1.4 Weight gain = 8–10.9: RR≃1.6 Weight gain = 11–19.9: RR≃1.5 Weight gain = 20–29.9: RR≃1.8 Weight gain = 30–39.9: RR≃2 Weight gain ≥ 40: RR≃2.5 |
Smoking status-age-weight-height-history of myocardial infarction, reproductive history-parental history of myocardial infarction, reproductive history, use of oral contraceptives or postmenopausal hormones, and personal history of coronary heart disease stroke, hypertension, diabetes, elevated serum cholesterol level and cancer | 8 | 17 |
| Chung | 2017 | Cohort study | 7 years | 881 | 52% female | 55 ± 9 | Weight gain > 10% Weight gain > 5% Stable (5%) Weight loss > 5% Weight loss > 10% |
Measured | CKD incidence risk increased as follows: Weight gain > 10%: RR: 1.45; 95% CI: 1.07, 1.97; P = 0.016 Gain > 5% RR: 1.08; 95% CI: 0.79, 1.46; P = 0.634 Loss > 5% RR: 1.08; 95% CI: 0.86, 1.36; P = 0.519 Loss > 10%: RR: 0.91; 95% CI: 0.57, 1.47; P = 0.710 |
Diabetes duration, gender, education (≤ 6, > 6 years), smoking status (never, past, current smoker) and drinking habit (yes, no), HbA1c, triglycerides and hypertension (high blood pressure or antihypertensive medication use), diabetic retinopathy and albuminuria | 7 | 10 |
| Strelitz | 2021 | Cohort study | 5 years | 3,057 | 42% female | 60.2 | i) Gained weight > 5% ii) Gained weight > 2%–≤ 5% iii) Maintained weight (gained ≤ 2% or lost < 2%) iv) Lost weight ≥ 2%–< 5% v) Lost weight 5%– < 10% vi) Lost weight ≥ 10% |
vi | CVD hazard ratio increased in groups accordingly: i): 0.92 (0.50, 1.70) ii): 1.35 (0.81, 2.24) iii): 1.0 iv): 1.44 (0.87, 2.39) v): 1.05 (0.62, 1.80) vi): 1.50 (0.85, 2.66) All-cause mortality HR increased in groups accordingly: i): 1.27 (0.72, 2.22) ii): 1.31 (0.79, 2.20) iii): 1.0 iv): 1.12 (0.66, 1.92) v): 0.85 (0.47, 1.54) vi): 2.04 (1.17, 3.55) Patients with weight loss of ≥ 10% show significant more all-cause mortality rate |
Age, gender, baseline weight, education, smoking, trial group, study center, baseline antihypertensive or lipid-lowering medication use, changes in medication use between baseline and 5 years, and having a CVD event within 5 years after diabetes diagnosis | 9 (Out of 11) | 19 |
| Moazzeni | 2021 | Cohort study | 14.4 years | 763 | 60.7% female | 53.6 ± 11 | Lost > 5% weight; lost 3%–5% weight; stable (± 3%) weight; gained 3%–5% weight; gained > 5% weight | Measured | CVD incidence hazard ratios are as followed: Lost > 5% (11%), 3%–5% (11% decrease), gained 3%–5%: (24% decrease), > 5% (30% decrease): not significant; HR and 95% CI, respectively: 1.11 (0.79, 1.56); 0.89 (0.60, 1.33); 0.76 (0.46, 1.23); 0.70 (0.48, 1.01) Age, year: 5% increase BMI: no change (not significant) Current smoker, yes: 64% increase GLD use, yes: 62% increase Family history of premature CVD, yes: 15% increase (not significant) Hypertension, yes: 73% increase Hypercholesterolemia, yes: 77% increase CKD, yes: 22% decrease (not significant) FPG: $4 increase (not significant) HR and 95% CI, respectively; 1.05 (1.03, 1.07); 1.00 (0.97, 1.03); 1.64 (1.03, 2.61); 1.62 (1.21, 2.16); 1.15 (0.84, 1.57); 1.73 (1.32, 2.26); 1.77 (1.27, 2.48); 0.78 (0.58,1.05); 1.04 (1.00,1.08) |
Age, sex, body mass index, educational level, current smoking (at first follow-up), glucose-lowering drug use (at baseline or first follow-up), family history of CVD, hypertension, hypercholesterolemia, chronic kidney disease, FPG | 9 (Out of 11) | 14 |
| Casanova | 2020 | Cohort study | 3.14±0.21 years | 154 | 72.3% female | 67.9 (66.6, 69.2) | Weight loss: ≥ 5% Weight gain: ≥ 5% |
Measured | Weight loss: 1.2 (95% CI: 13.2, 15.7) AU × min Stable weight: 15.8 (−10.5, −21.0) AU × min Weight gain: 37.8 (−19.4, −56.2) AU × min |
Age/sex/SD change | 9 (Out of 11) | 12 |
| Strelitz | 2019 | Cohort study | 10 years | 725 | 38.5% female | 61.1 ± 7.1 | i) Gained > 2% weight ii) Maintained weight iii) Lost ≥ 2%–< 5% iv) Lost ≥ 5% weight |
Measured | CVD incidence HR: i): 0.41 (0.15, 1.11) ii): 1.00 iii): 0.79 (0.43, 1.46) iv): 0.52 (0.32, 0.86) All-cause mortality HR: i): 1.63 (0.83, 3.19) ii): 1.00 iii): 1.08 (0.60, 1.93) iv): 1.12 (0.52, 2.37) Patients who gained weight were at more risk of all-cause mortality |
Age, sex, baseline SES, baseline BMI, smoking at 1 year, use of antihypertensive, lipid- or glucose-lowering medication at 1 year, and trial arm | 10 (Out of 11) | 15 |
| Polemiti | 2021 | Cohort study | 10.8 years | 1,083 | 46% female | 59.1 (52.2–64.4) | i): > 1% BMI loss ii): Stable BMI iii): > 1% BMI gain |
Self-reported | HRs and 95% CIs for microvascular and macrovascular complications. Total vascular complications: i): 0.69 (0.54, 0.89) ii): 1.00 (Ref.) iii): 0.86 (0.65, 1.14) Macrovascular complications: i): 1.04 (0.62, 1.74) ii): 1.00 (Ref.) iii): 0.82 (0.42, 1.63) Microvascular complications: i): 0.62 (0.47, 0.80) ii): 1.00 (Ref.) iii): 0.90 (0.67, 1.21) Kidney disease: i): 0.57 (0.40, 0.81) ii): 1.00 (Ref.) iii): 1.03 (0.71, 1.50) Neuropathy: i): 0.73 (0.52, 1.03) ii): 1.00 (Ref.) iii): 0.82 (0.56, 1.20) Patients with increased BMI were at more risk of CVD compared to those with decreased BMI. |
Adjusted for age, sex and pre-diagnosis BMI education, smoking status change, smoking duration at pre-diagnosis, smoking duration change, physical activity at pre-diagnosis, physical activity change, alcohol consumption at pre-diagnosis, alcohol consumption change, MedPyr score, lipid-lowering medication, antihypertensive medication and glucose-lowering medication | 9 (Out of 11) | 6 |
| Cui | 2021 | Cohort study | 7 years | 1,774 | 55.1% female | 60.32 ± 8.88 | One-unit increase in the BMI level | Measured | OR and P-value for one-unit increase in BMI in different populations for stroke are as followed: Whole population: 1.133 (1.046, 1.242); 0.004 Men: 1.153 (1.045, 1.313); 0.008 Women: 1.12 (0.977, 1.292); 0.12 Increase in BMI was slightly associated with stroke in both sexes |
Age, sex (if not stratified), education level, marriage status, residence, exercise, smoking status and drinking hypertension, dyslipidemia, systolic blood pressure, fasting blood glucose, glycosylated haemoglobin, triglycerides, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, uric acid and eGFR. | 9 (Out of 11) | 18 |
| Aucott | 2016 | Cohort study | 5.2 years | 2,9316 | 45.5% female | 58 ± 12 | i) Loss: 10% or more, < 10%–5%, < 5%–2.5%; ii) Stable: Loss of 2.5% up to gain of 2.5%; iii) Gain: > 2.5%–5%, > 5%–10%, 10% or more. |
Measured | HRs for all-cause mortality and cardiovascular outcomes for each weight category: All-cause mortality: i) 1 ii) 0.86 (0.55, 1.33) iii) 0.98 (0.69, 1.37) MI: i) 1 ii) 0.98 (0.62, 1.54) iii) 0.94 (0.64, 1.39) CHF: i) 1 ii) 0.97 (0.54, 1.77) iii) 0.96 (0.59, 1.55) PVD: i) 1 ii) 1.61 (0.87, 2.98) iii) 0.81 (0.43, 1.55) Weight gain and weight loss showed strong association with various kinds of cardiovascular complications. |
Age, BMI, sex, smoking status and deprivation | 9 (Out of 11) | 11 |
| Liu | 2020 | Cohort study | 38 years | 173,229 | 70% female | 62.01 | Weight gain: < 0 kg 0.1 kg–5 kg > 5 kg |
self-reported | Among all recent quitters: HR: 0.83 (95% CI: 0.70, 0.99) Among recent quitters without weight gain: HR: 0.77 (95% CI: 0.62, 0.95) Among recent quitters with weight gain of 0·1 kg–5.0 kg: HR: 0·99 (95% CI: 0.70, 1.41) Among recent quitters with weight gain of > 5·0 kg: HR:0·89 (95% CI: 0·65, 1·23) Among longer-term quitters: HR:0·72 (95% CI: 0.61, 0·84) Among long-term quitters without weight gain: HR: 0.69 (95% CI: 0.58, 0.82) Among long-term quitters with weight gain of 0.1 kg–5.0 kg: HR: 0.57 (95% CI: 0.45, 0.71) Among long-term quitters with weight gain of > 5.0 kg: HR: 0.51 (95% CI: 0.42, 0.62) |
Age, diabetes duration, sex, white ethnic origin, BMI assessed in the cycle before diabetes was diagnosed, physical activity, alcohol consumption, Alternative Health Eating Index score, family history of myocardial infarction before age 60 years old, family history of cancer, current aspirin use, current multivitamin use, presence of hypertension, presence of hypercholesterolemia and use of diabetes medication (insulin, oral medication or others) | 9 (Out of 11) | 13 |
Note: CVD = cardiovascular disease; HR = hazard ratio; CI = confidence interval; GLD = glucose-lowering drug; BMI = body mass index; RAS = renin-angiotensin system; SBP = systolic blood pressure; CKD = chronic kidney disease; BW = body weight; FPG = fasting plasma glucose; T2DM = type 2 diabetes mellitus; HbA1c = haemoglobin A1c; eGFR = estimated glomerular filtration rate; ACR = albumin to creatinine ratio; CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; MI = myocardial infarction; LDL = low-density lipoprotein; aHR = adjusted hazard ratio; SD = standard deviation; CHF = congestive heart failure; TM = total mortality; MVE = microvascular events defined as nephropathy, neuropathy or retinopathy; OR = observed ratio; MI = myocardial infarction; PVD = peripheral vascular disease