If AI Macros or AI Nutrition Targets are calculating incorrect values, or not working at all, here's what to check.

How the AI Calculates Macros

The AI calculates BMR (Basal Metabolic Rate) using one of three equations you can choose in your AI Nutrition Targets settings:

Mifflin-St Jeor (default, recommended)
Harris-Benedict
Katch-McArdle (uses lean body mass — requires body fat percentage)

The BMR is then multiplied by an activity level multiplier to get TDEE (Total Daily Energy Expenditure), and finally adjusted based on the client's goal.

Activity level multipliers:

Sedentary — 1.2x
Lightly Active — 1.375x
Moderately Active — 1.55x
Very Active — 1.725x
Extra Active — 1.9x

Required data for calculation:

Current weight (from Trainerize data or starting weight)
Height (from Trainerize profile)
Age (from birth date in Trainerize)
Gender (must be male or female — "other" is not supported by the BMR equation)
Goal (the client must have a goal set, such as fat loss or muscle gain)

If any of these are missing, the calculation returns 0 and the AI cannot suggest macros.

AI Fallback and Reliability

The system is designed to always produce a result, even if something goes wrong.

Multiple AI providers — If the primary AI model fails, the system automatically tries backup models from different providers. You do not need to do anything.
Rule-based fallback — If all AI models fail, the system falls back to a rule-based calculation using the BMR equation, your configured macro preferences, and the client's goal.
Minimum calorie safety — The fallback calculation enforces a minimum of 1,200 calories regardless of deficit settings, to ensure client safety.

Common Issues

1. "No weight data" or Calculation Returns 0

The client must have at least one of:

A recent weight entry in Trainerize
A starting body weight set on their profile

If neither exists, the AI has nothing to calculate from. Ask the client to log their weight in the Trainerize app, or set their starting weight manually.

2. Wrong Calorie Target (Too High or Too Low)

Check these settings:

Activity Level — Is the client's activity level set correctly? "Very Active" (1.725x multiplier) or "Extra Active" (1.9x multiplier) gives significantly more calories than "Sedentary" (1.2x). This is pulled from the client's Trainerize profile.
BMR Method — Different BMR equations produce different base values. Katch-McArdle can produce different results than Mifflin-St Jeor, especially for clients with high muscle mass. Check which method you have selected in your AI Nutrition Targets settings.
Goal Adjustment — The calorie adjustment ranges from -1000 (aggressive weight loss) to +1000 (aggressive weight gain). Check that the goal setting matches the client's actual goal.
Unit System — If your business unit preference is set to "Imperial" but the client's weight is stored in kg (or vice versa), the conversion may produce wrong values. Check Settings for your unit preference.

3. Height or Birth Date Not Synced

Height and birth date come from the client's Trainerize profile metadata. If these fields are empty in Trainerize, the AI can't calculate BMR.

Fix: Have the client update their profile in the Trainerize app (height and birth date), then run a client sync from Settings → Trainerize Integration.

4. Gender Set to "Other"

The Mifflin-St Jeor and Harris-Benedict equations only support male and female. If the client's gender is set to "Other" in Trainerize, the BMR calculation returns 0 and no macros are generated.

5. AI Macros Not Adjusting Weekly

If AI Macros is enabled but not adjusting on schedule:

Check the adjustment day and time — is it the right day? The system checks the hour in the client's configured timezone.
Check that the client has enough progress data for the AI to evaluate (weight trends, nutrition logging)
Make sure the client's status is Active
If the adjustment was overdue (missed its scheduled day), it will be automatically rescheduled to the next correct weekday rather than firing on the wrong day

6. Unexpected Variance Percentage

Trainerize only allows specific variance values: 5%, 10%, 20%, 30%, 40%, 50%. If you set a variance that does not match one of these, the system automatically rounds to the nearest allowed value. This is a Trainerize limitation, not a FitMetrics issue.

AI Nutrition Targets Settings

You can customize the AI's behavior in your AI Nutrition Targets Configuration page.

BMR Calculation Method:

Mifflin-St Jeor — recommended for most clients (default)
Harris-Benedict — alternative standard formula
Katch-McArdle — uses lean body mass, best when body fat percentage is available

Macro Distribution Preferences:

Protein preference — Low (0.8-1.2 g/kg), Moderate (1.2-1.8 g/kg), High (1.8-2.5 g/kg), Very High (2.5-3.0 g/kg)
Carb preference — Very Low (keto), Low, Moderate, High
Fat preference — Low, Moderate, High, Very High (keto)
Protein range — Fine-tune min/max in g/kg or g/lb with the slider

Personalization Toggles: