Table 2.
Study characteristics and results.
| Author | Country | Design | Sample size |
Follow-up time | Median/mean age | Weight status measure |
Self-report/measured | Outcome parameter | Results | Quality | Limitations |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Zamora et al. (2013) [47] | Spain | Prospective cohort | 504 | 6.1 years (IQR 2.2–7.8) |
68 years (IQR 58–74) |
BMI | Measured | Hazard ratio | Lower risk of death associated with obesity. Hazard ratios were 2.08 (1.16–3.75), 0.88 (0.54–1.43), and 0.49 (0.28–0.86) for underweight, overweight, and obese patients, respectively. |
Strong | |
|
| |||||||||||
| Casas-Vara et al. (2012) [42] | Spain | Prospective cohort | 244 | 984 days (2.7 years) |
(83.2 ± 0.5) | BMI | Measured | Survival | Overall median survival was 984 days with better survival as BMI increased; log ranks were 5.26, 8.45, and 10.95 for underweight, overweight, and obese patients, respectively. |
Moderate | Baseline characteristics of study population were not fully reported. |
|
| |||||||||||
| Lavie et al. (2013) [43] | USA | Prospective cohort | 2066 | 5 years (25.0 ± 17.5 months) |
BMI | Measured | 5-year survival | Overall better prognosis in patients with higher BMI. On stratification on high FIT versus low FIT, obesity paradox persists in the high fit group but not in the low fit group. | Moderate | Primary objective was comparison between low fit versus high fit and not obese versus lean. | |
|
| |||||||||||
| Gastelurrutia et al. (2011) [44] | Spain | Prospective cohort | 979 | 44 months | 65 ± 12 years | BMI | Measured | Hazard ratio | Lower risk of death associated with higher BMI when compared to normal BMI = 0.94, (0.91–0.97). | Strong | |
|
| |||||||||||
| Clark et al. (2011) [36] | USA | Retrospective cohort | 344 | 2 years | 53.3 ± 13.1 years | BMI, WC | Measured | 2-year survival | Higher 2-year survival in high versus normal WC: 77.9% versus 64.3%, and high versus normal BMI: 89.8% versus 58.2%. |
Moderate | Retrospective analysis of data of patients seen at a specialist center for heart transplant patients selected based on disease severity and prognosis and thus not likely to be representative of all heart failure patients. |
|
| |||||||||||
| Clark et al. (2012) [35] | USA | Retrospective cohort | 2718 | 2 years | 53.0 ± 12.4 years | BMI, WC | Measured | 2-year survival | Higher 2-year survival with increasing BMI in both sexes. Men: high versus normal, BMI (63.2% versus 53.5%), WC (78.8% versus 63.1), women: BMI (67.1% versus 56.6%), WC, no difference. | Moderate | Not many women were included (~25%), bringing into question the power to detect differences in that subgroup. |
|
| |||||||||||
| Schwartzenberg et al. (2012) [48] | Israel | Prospective cohort | 2323 | 15 months | 71.6 ± 12.5 years | BMI | Not stated | Hazard ratio | Lower risk of death in persons with higher BMI. Normal versus higher BMI, though not statistically significant after adjustment. Hazard ratio of 0.79 (95% CI 0.59–1.05). | Moderate | Important baseline characteristics not shown. |
|
| |||||||||||
| Kapoor and Heidenreich (2010) [49] | USA | Retrospective cohort | 1236 | 426 ± 461 days | 71 ± 12 years | BMI | Measured | Hazard ratio | Lower risk of death as BMI increased until BMI > 45. Hazard ratios were 1.68 (95% CI, 1.04–2.69), 0.99 (95% CI, 0.71–1.36), 0.58 (95% CI, 0.35–0.97), 0.79 (95% CI, 0.44–1.4), and 1.38 (95% CI 0.74–2.6) for underweight, overweight, obese, and morbidly obese patients and BMI > 45 kg/m2, respectively. Thus, underweight and BMI > 45 kg/m2 were associated with increased risk of death. |
Strong | |
|
| |||||||||||
| Curtis et al. (2005) [45] | USA | Clinical trial | 7767 | 37 months | 63.9 ± 10.9 years | BMI | Measured | Hazard ratio | Lower risk of death associated with increasing BMI. Using normal BMI as reference, hazard ratios were 1.21 (0.95–1.53), 0.88 (0.80–0.96), and 0.81 (0.72–0.92) for underweight, overweight, and obese patients, respectively. |
Moderate | Results were primarily for a Digitalis clinical trial; thus, results were reanalyzed with a new hypothesis that was specified after study. |
|
| |||||||||||
| Zuchinali et al. (2013) [46] | Brazil | Prospective cohort | 344 | 30 ± 8.2 months | 59 ± 13 | TSF | Measured | Hazard ratio | TSF > 20 mm was a strong predictor of all-cause mortality. Hazard ratio = 0.36 (0.13–0.97, p = 0.03). Results for other measures of body fat composition, BMI, WC, arm circumference (AC), Arm Muscle Circumference (AMC), and TSF, were not statistically significant in any direction. |
Strong | |