CHAPTER PART 3 OF 9
The reasoning behind the numbers, so staff can defend a prescription and adapt it rather than reciting it.
The reasoning behind the numbers, so staff can defend a prescription and adapt it rather than reciting it.
Protein restriction is the oldest idea in renal nutrition and the one whose standing has fallen furthest. In non-diabetic CKD, Cochrane found that a very low protein diet of 0.3 to 0.4 g/kg with ketoanalogues reduced end-stage kidney disease (RR 0.65, moderate certainty) but that a conventional low protein diet of 0.5 to 0.6 versus a normal 0.8 or more made little or no difference (RR 1.05, low certainty). In diabetic kidney disease the evidence is thinner still.
What remains firm: avoid high protein intake above 1.3 g/kg/day in anyone at risk of progression, and never restrict in a metabolically unstable patient. Both are KDIGO practice points and both are more actionable than the argument over 0.6 versus 0.8.
Prioritise protein quality and distribution over the last 0.1 g/kg. Spread it across three meals rather than loading dinner, because renal gluconeogenesis is lost as cortical mass falls and long fasting intervals stop being tolerable.
25 to 35 kcal/kg/day. The guideline says only "body weight" and defers the choice of weight to clinical judgement, so state your basis explicitly on the chart. In practice: pre-illness or usual weight for a normal or overweight patient, and the weight corresponding to BMI 25 in obesity. An under-fed patient on a low protein diet will burn their own muscle for gluconeogenesis and the urea will rise anyway, which is the classic reason a "compliant" low protein diet fails.
Under 2 g of sodium per day, which is under 5 g of salt, which is a little under one level teaspoon. One level teaspoon of salt is about 1,940 mg of sodium, so the whole day's allowance is roughly a teaspoon of salt across all cooking, all pickle, all papad and all packet food.
| Item | Sodium | Note |
|---|---|---|
| Table salt | 388 mg per g | 1 level tsp (5 g) = 1,940 mg. Rock salt and sendha namak are the same. They are not low sodium. |
| Baking soda in cooking | 274 mg per g | Half a teaspoon in dhokla, pakoda batter or to soften rajma and chana is about 630 mg. |
| Pickle / achar | 1,759 mg per 100 g | One 10 g teaspoon is about 176 mg, roughly 9 percent of the day. |
| Papad | 56–244 mg per piece | Brand variation is enormous. Two papads can be a fifth of the allowance. |
| Masala, sambar and chutney powders | 1,468 mg per 100 g | Average of Indian packaged spice blends. Home-ground without salt solves this. |
| Instant noodles | ~1,000 mg per 100 g | One cake plus tastemaker is roughly half a day's sodium. |
| Processed cheese slice | 1,074–1,671 mg per 100 g | Also carries additive phosphate. Worst single item on this list. |
| Sodium bicarbonate tablets | 177 mg per 650 mg tab | 650 mg three times daily is 530 mg of sodium prescribed by us. |
This is where practice has changed most and where our handouts are most out of date. KDOQI 2020 and KDIGO 2024 both abandoned blanket potassium restriction. Neither gives a milligram ceiling. KDIGO's practice point asks us to limit foods rich in bioavailable potassium, naming processed foods, and to do so in people with a history of hyperkalaemia or during periods when the risk is raised. The Indian Journal of Nephrology 2025 practice review puts it bluntly: dietary potassium restriction is for treating hyperkalaemia, not for preventing it, and leaching is only required in patients who are hyperkalaemic.
"Potassium 2,000 to 3,000 mg per day" is legacy KDOQI 2000 and 2003 dialysis teaching. "1 mmol/kg/day" was never guideline text at all, only renal-dietetic convention. Both survive in handouts and hospital charts. The average adult eats about 2,000 mg anyway, so the "restriction" is often the population mean dressed up as a prescription.
Where potassium restriction is indicated, the order of attack is: potassium-based salt substitutes first, then processed and packaged food, then tender coconut water and fruit juice, then tamarind-heavy gravies, then portion control of high-potassium produce, and only then leaching. Attacking produce first is backwards.
Three sources, three absorptions, one serum value. Teach the source, not the milligram.
Dals, atta, millets, nuts. Humans have no phytase. IFCT 2017 measures phytate, so you can show the patient why dal is different.
Milk, curd, paneer, khoa, egg yolk, meat, fish. No phytate at all.
Cola, processed cheese, instant noodles, bakery, tumbled meat. No matrix, no phytate, no binder can compete.
The useful metric at the table is the phosphorus-to-protein ratio, target under about 10 mg per gram (some renal-nutrition texts use under 12, or under 16). Note that this ratio target is renal-dietetic convention from the review literature and appears in no guideline, so present it as a tool, not a rule.
Binders are a dietetic responsibility. They must be matched to the phosphate content of each meal and snack, not taken on a fixed clock. Calcium carbonate is chewed just before food. Calcium acetate, sevelamer and sucroferric oxyhydroxide go in after the first two or three mouthfuls. Lanthanum is chewed at the end of the meal. A phosphate-free snack does not need a dose. A phosphate-heavy snack does.
There is no fluid statement in KDOQI 2020 or KDIGO 2024. Everything in circulation is convention. For non-dialysis CKD without oedema, 1.5 to 2 L is a reasonable working figure, individualised to urine output, weight and blood pressure. Do not coach patients to drink more to "flush" the kidneys: the CKD-WIT randomised trial showed that coached increased water intake over 12 months did not slow eGFR decline.
On haemodialysis, the number that matters is interdialytic weight gain, target under 4 to 4.5 percent of body weight. High sodium intake, not thirst personality, is what drives it, so fluid counselling that does not start with salt is wasted breath.
Diagnosed when at least three of four categories are met, confirmed on two occasions: serum chemistry (albumin under 3.8 g/dL, cholesterol under 100 mg/dL), body mass (BMI under 23, unintentional loss of 5 percent in 3 months or 10 percent in 6), muscle mass (equivalent losses, reduced mid-arm muscle circumference), and intake (unintentional protein under 0.8 g/kg on dialysis or under 0.6 g/kg in CKD 2 to 5, or energy under 25 kcal/kg, for two months or more).
The moment PEW is present the patient stops being metabolically stable, and every low protein statement is void. Escalate instead: a minimum three-month trial of oral nutritional supplements, then enteral feeding, then intradialytic parenteral nutrition on haemodialysis. Screen at least twice a year, and more often on a GLP-1 receptor agonist.