Documentation Index
Fetch the complete documentation index at: https://mintlify.com/JuanSebasSV/healtyhelp/llms.txt
Use this file to discover all available pages before exploring further.
Personalized Nutrition Recommendations
HealtyHelp includes a server-side recommendation engine (motorRecomendaciones.js) that analyses each user’s health profile, body metrics, and the last 30 days of consumption history to produce real-time, personalised nutritional and exercise guidance.
What the Engine Uses
The engine reads only two fields from healthProfile:
| Field | Description |
|---|
healthProfile.condiciones | Active health conditions / dietary patterns |
healthProfile.categorias | Active meal-time categories (desayuno, almuerzo, etc.) |
healthProfile.alergias and healthProfile.preferencias are not read by the recommendation engine. Those fields are only used by NutriBot’s recipe filtering. Allergy information never influences the recommendation score.
| User field | Description |
|---|
weight | Weight in kg |
height | Height in cm |
birthDate | Date of birth (for age and BMR calculation) |
TDEE Calculation (Mifflin-St Jeor)
The engine calculates a gender-neutral Total Daily Energy Expenditure using a modified Mifflin-St Jeor formula with a gender-neutral BMR offset:
// Offset -78 is the mean of the male (+5) and female (-161) constants
const bmr = 10 * weight + 6.25 * height - 5 * age - 78;
Age Correction for Sarcopenia
After age 40, the BMR drops approximately 2% per decade due to muscle mass loss:
const factorEdad = age > 60 ? 0.93 : age > 40 ? 0.97 : 1.0;
Activity Factor from BMI
Because HealtyHelp does not store a user-reported activity level, the engine infers one from the BMI category as a statistical proxy for sedentary behaviour:
const FACTOR_ACTIVIDAD_IMC = {
bajo_peso: 1.45,
normal: 1.55,
sobrepeso: 1.48,
obesidad: 1.375,
};
const tdee = Math.round(bmr * activityFactor * ageFactor);
BMI Categories
function calcularIMC(weight, height) {
const m = height / 100;
const val = weight / (m * m);
// bajo_peso < 18.5 | normal < 25 | sobrepeso < 30 | obesidad >= 30
}
categoria string drives the activity factor, hydration adjustments, exercise recommendations, and several alert messages.
Health Condition Limits (LIMITES_CONDICION)
The engine enforces per-day nutritional ceilings for each health condition. When today’s accumulated intake exceeds a limit, the engine adds a real-time alert to alertasHoy.
| Condition | Nutrient | Daily Limit |
|---|
diabetes | azucar | ≤ 25 g |
diabetes | carb | ≤ 130 g |
hipertension | sodio | ≤ 1 500 mg |
bajo-sodio | sodio | ≤ 1 500 mg |
enfermedad-renal | sodio | ≤ 1 500 mg |
enfermedad-renal | potasio | ≤ 2 000 mg |
enfermedad-renal | fosforo | ≤ 800 mg |
colesterol-alto | grasSat | ≤ 15 g |
colesterol-alto | colesterol | ≤ 200 mg |
keto | carbNetos | ≤ 50 g |
bajo-carbohidratos | carb | ≤ 100 g |
sin-azucar | azucar | ≤ 10 g |
bajo-grasa | gras | ≤ 50 g |
- Conditions
hipertension, enfermedad-renal, or bajo-sodio → 1 500 mg
- Age ≥ 50 (no cardiovascular condition) → 1 800 mg
- All other adults → 2 300 mg (WHO general guideline)
Macro Objectives by Condition (OBJETIVOS_MACRO)
These percentage targets are used to evaluate the macro split in alertasHoy when the user has consumed more than 30 % of their TDEE:
| Condition | Protein % | Carb % | Fat % |
|---|
default | 20 % | 50 % | 30 % |
keto | 20 % | 5 % | 75 % |
bajo-carbohidratos | 30 % | 20 % | 50 % |
bajo-grasa | 30 % | 55 % | 15 % |
paleo | 30 % | 25 % | 45 % |
vegano | 18 % | 55 % | 27 % |
vegetariano | 18 % | 52 % | 30 % |
colesterol-alto | 22 % | 50 % | 28 % |
diabetes | 20 % | 45 % | 35 % |
30-Day Consumption History
The engine fetches all Consumo documents for the authenticated user where fechaBogota >= today - 30 days. From this history it computes:
- Daily averages for
cal, prot, sodio, fiber, carb, gras, azucar (via promedioNutri)
- Skipped meal detection — if a meal type (desayuno, almuerzo, cena) appears in fewer than 50 % of logged days over the period, it is flagged as a
comidasSaltadas entry
function detectarComidasSaltadas(consumos) {
const fechas = [...new Set(consumos.map(c => c.fechaBogota))];
if (fechas.length < 3) return []; // requires at least 3 days of data
return ['desayuno', 'almuerzo', 'cena'].reduce((acc, tipo) => {
const pct = fechas.filter(f =>
consumos.some(c => c.fechaBogota === f && c.tipo === tipo)
).length / fechas.length;
if (pct < 0.5) acc.push({ tipo, porcentaje: Math.round(pct * 100) });
return acc;
}, []);
}
Timezone Handling
Like the consumption controller, the recommendations route uses Bogotá UTC−5 for all date comparisons:
const offsetMs = -5 * 60 * 60 * 1000;
const bogota = new Date(ahora.getTime() + offsetMs);
const horaActual = bogota.getUTCHours(); // 0–23, drives contextual meal tips
const fechaHoy = bogota.toISOString().split('T')[0];
const hace30Dias = new Date(bogota.getTime() - 30 * 24 * 60 * 60 * 1000);
horaActual is passed into generarRecomendaciones to produce time-sensitive contextoHorario messages (e.g. late-night eating warnings, skipped breakfast alerts).
API Endpoints
Get Recommendations
GET /api/recomendaciones
Authorization: Bearer <token>
healthProfile.condiciones / healthProfile.categorias, queries the last 30 days of Consumo records and today’s records in parallel, then calls generarRecomendaciones.
Response structure:
{
"ok": true,
"nombre": "María",
"recomendaciones": {
"imc": { "valor": 24.1, "categoria": "normal" },
"caloriasObjetivo": 1890,
"caloriasHoy": 1420,
"caloriasRestantes": 470,
"progresoHoy": 75,
"alertas": [...],
"alertasHoy": [...],
"contextoHorario": [...],
"alimentacion": [...],
"ejercicio": [...],
"ejercicioHoy": [...],
"hidratacion": { "litros": 2.1, "vasos": 8, "nota": "..." },
"comidasSaltadas": [...],
"condicionesDetectadas": ["diabetes"],
"categoriasActivas": ["desayuno"],
"nutriPromedio": {
"calPromedio": 1750, "protPromedio": 62,
"carbPromedio": 210, "grasaPromedio": 58,
"sodioPromedio": 1800, "fibraPromedio": 18
},
"macrosHoy": { "carbPct": 48, "protPct": 22, "grasPct": 30 }
}
}
Get Recommendation Filters
GET /api/recomendaciones/filtros
Authorization: Bearer <token>
condiciones and categorias — not allergies or preferences, which are irrelevant to the recommendation engine:
{
"condiciones": ["diabetes", "hipertension"],
"categorias": ["desayuno"]
}
The PanelRecomendaciones Component
The front-end panel (PanelRecomendaciones.jsx) is embedded in the tracking view and re-fetches from /api/recomendaciones whenever its versionFiltros prop increments. This prop is bumped by the parent component whenever the user changes their health condition filters on the home page — ensuring the panel stays in sync without polling.
The panel renders these sections:
| Section | Source field |
|---|
| IMC badge + calorie target | imc, caloriasObjetivo |
| Caloric progress bar today | caloriasHoy, progresoHoy |
| Macro split bars today | macrosHoy |
| Real-time contextual tips | contextoHorario |
| General profile alerts | alertas |
| Today’s micronutrient alerts | alertasHoy |
| Suggested activity for today | ejercicioHoy |
| Hydration goal | hidratacion |
| Skipped meal warnings | comidasSaltadas |
| Nutrition tips | alimentacion |
| General exercise guidance | ejercicio |
All recommendation text is generated purely server-side. The client panel is a display layer only — it never applies local filtering logic to the recommendation data.