*3.3. "Metabolically Healthy" Cohort*

No differences in age, sex or pubertal distribution between the MH and No-MH groups were observed. However, the MH group showed a slightly lower mean standardized BMI (+3.78 ± 1.30) than No-MH (+4.15 ± 1.59 SDS; *p* < 0.001).

The degree of metabolic impairment categories observed in the entire cohort based on a fasting serum sample determinations and using the criteria previously described (12) are shown in Figure 5.

**Figure 3.** Evolution of patient retainment and their standardized BMI (expressed as SDS) in the five years following weight loss in the excellent responder group. *Abbreviations:* BMI-SDS: Standardized body mass index (Z-score); WL: Weight loss.

**Figure 4.** Evolution of BMI Z-score at every time-point throughout the 5 years following intense weight loss in the "excellent responder" group. Shown as the mean of a paired comparison with each patient's BMI Z-score at each timepoint compared to baseline (taking into account exclusively those retained at each follow-up) differentiating between those further reducing their BMI (F-WR) or regaining it (Regain) after the initial intense weight loss. *Abbreviations:* 95% CI: 95% Confidence interval; F WR: Further weight (and BMI) reduction; Regain: Showing weight (and BMI) increase after initial intense loss. The percentages of patients showing F-WR after initial weight loss among those retaining follow-up were 71.1% at 6 months, 63.9% at 1 year, 53.7% at 2 years, 61.3% at 3 years, 38.9% at 4 years, and 69.2% at 5 years.

**Figure 5.** Prevalence of metabolic comorbidities in the entire study cohort at baseline based on fasting determinations. *Abbreviations:* ADA: American Diabetes Association (criteria for IFG: ≥100 mg/dL); HDL-c: High density lipoprotein cholesterol; IFG: Impaired fasting glucose; IR: Insulin resistance (fasting insulin ≥ 15 μU/mL); LDL-c: Low density lipoprotein cholesterol; T2DM: Type 2 diabetes mellitus (ADA criteria: ≥126 mg/dL, confirmed); WHO: World Health Organization (criteria for IFG: ≥110 mg/dL).

The prevalence of MH patients was lower in Latinos (30.4%) than in Caucasians (41.0%; χ<sup>2</sup> 9.358; *p* < 0.01) and was also lower in IR (28.1%) compared to No-IR patients (48.0%; χ<sup>2</sup> 46.003; *p*< 0.01). Consequently, the prevalence of IR (37.0% in the whole cohort) was higher in the No-MH group (44.8%) than in the MH patients (25.5%). Among MH patients, those with IR (*n* = 129) were older, more severely obese, and had higher systolic and diastolic blood pressure and glucose, uric acid and triglyceride levels and lower HDL-c levels than the MH patients without IR (Table 5).


**Table 5.** Clinical features of metabolically healthy (MH) patients and comparison according to the presence or absence of fasting hyperinsulinemia (Insulin ≥ 15 μU/mL vs. Insulin < 15 μU/mL).

*Abbreviations:* BMI: body mass index; DBP: diastolic blood pressure; SBP: diastolic blood pressure; SDS: Standard deviation score. Data are shown as mean ± SD.

Metabolic data at the end of follow-up were available for 152 of the 505 MH patients at the onset of the study (30%). Among these, 73.7% remained MH whereas 26.3% had developed at least one metabolic comorbidity (independently of the presence or absence of IR at diagnosis). Among the No-MH patients (metabolic data at the end of follow-up were available in 29.8% [237/795]) 8.4% became MH by the end of follow-up, whereas 91.6% still showed at least one metabolic comorbidity.

A significant decrease in BMI from baseline to the end of the study was observed both in those patients who remained MH (+3.50 ± 1.02 to +2.76 ± 1.36 SDS, *p* < 0.001) as well as in those who resolved their initial comorbidities (+3.58 ± 1.57 to +2.69 ± 2.28 SDS, *p* < 0.05), but not in those developing comorbidities throughout follow-up or remaining No-MH.
