For Health Professionals

  Some people with type 1 diabetes  may struggle to achieve optimal   glycaemic control despite  

 accurate carb counting   

Carbohydrate counting is the gold standard for adjusting mealtime insulin doses in type 1 diabetes.  It is based on the premise that carbohydrate is the predominant macronutrient affecting postprandial blood glucose levels. However the method fails to take into account that carbohydrates varying in their glycaemic index (GI) and fat and protein are also known to affect normal insulin secretion.

Our recent meta-analysis shows that carbohydrate counting alone doesn't significantly improve glycaemic control in type 1 diabetes (Bell, 2013).  This doesn't mean we should throw the proverbial baby out with the bathwater. But it does suggest that perhaps there is room for improvement...

Beyond Carb Counting

The Guidelines

Following on from our research, in 2015 the American Diabetes Association introduced a new recommendation that:

"People with type 1 diabetes who have            mastered carbohydrate counting, should            receive education on protein and fat".

ADA Standards of Medical Care, 2015

This recommendation has remained in the guidelines since then, with slight adjustments to the wording.  The 2017 guideline now says "selected individuals who have mastered carbohydrate counting should be educated on fat and protein gram estimation".  While it is great to see our papers being cited and the importance of this area recognised by such a prestigious association, it is not clear whether people with type 1 diabetes really need to start counting grams of fat and protein - and if they were to, how this information should be used to adjust insulin doses.  Indeed, as the summary below will show, there isn't enough evidence to support exactly how insulin should be adjusted for a 'high' fat and/or protein meal, let alone how to adjust insulin for incremental differences in grams of fat and protein.

The 2017 American Diabetes Association Guidelines can be accessed by clicking on the image.

Understanding the Evidence

Recently, the major researchers in this field from around the world, including ourselves, Joslin Diabetes Center & Harvard Medical School in the USA and John Hunter Childrens Hospital in Newcastle, Australia, collaborated on the first comprehensive systematic literature review examining the relative impact of nutrients on postprandial glycaemia in people with T1D and prandial insulin dosing strategies (Bell, 2015).  

Nutritional Factors Affecting Postprandial Glycaemia

Every study investigating the impact of dietary fat (7 studies, total 103 patients) or dietary protein (7 studies, total 125 patients) showed these nutrients resulted in significant differences in postprandial glycaemia in people with type 1 diabetes (Bell, 2015).

Dietary fat tended to reduce blood glucose levels 

in the early postprandial period (0-3hrs) and delay 

the peak blood glucose level, potentially due to

delayed gastric emptying.  In the late postprandial

period (>3hrs), dietary fat tended to cause sustained

hyperglycaemia.  This can be clearly seen in the

graph to the right adapted from our collaborative

study with Joslin Diabetes Center & Harvard Medical

School, USA (Bell, 2015).

Dietary protein has a delayed affect on glycaemia compared with carbohydrate and doesn't start to increase blood glucose levels substantially until approximately 1-1.5hrs after the start of the meal.  In the late postprandial period, dietary protein can also cause sustained hyperglycaemia.   It is important to note that dietary protein has different affects when consumed with or without carbohydrate (e.g. a steak and green salad versus beef spaghetti bolognese).  Our friends and collaborators in Newcastle, Australia have shown that when protein is eaten with carbohydrate, approximately 30g of protein appears sufficient to influence glycaemia. However if carbohydrate is not included in the meal, then there needs to be approximately 75g of protein before a significant affect is observed (Paterson, 2016).

Insulin Adjustments for Nutrition

We identified ten studies examining the impact of different insulin bolus algorithms for high-fat and/or high-protein meals.  There were a number of different study designs used which meant it was difficult to compare studies.  However, overall the studies consistently indicate that additional insulin is needed for high fat and/or high protein meals, despite carbohydrate content being consistent.  

The two main insulin adjustment algorithms in the literature were the Fat/Protein Unit method (also known as the Warsaw or Pankowska Method) and our Food Insulin Index.

Fat/Protein Unit Method was developed by Prof Pankowska at the Warsaw Insulin Pump School in Poland and has gained popularity in Paediatric centres in Europe.  The method is based on their Fat/Protein Unit (FPU), which is equal to 100 calories from either fat or protein.  For every FPU, additional insulin is given in the same ratio as the individuals' insulin: carbohydrate exchanges ratio.  The additional insulin is given as a dual-wave and the duration is extended as the number of FPU increases.  There are a few studies trialling the method in practice and while there are improvements in postprandial glycaemia, there is also an increased risk of hypoglycaemia.

Food Insulin Index was developed our team at the University of Sydney.  The Food Insulin Index (FII)  takes an alternative approach to thinking about insulin dosing in type 1 diabetes.  Rather than focusing on how much exogenous insulin is needed to keep blood glucose levels in target, this approach looks at how much insulin the pancreas would normally produce in response to a meal in people without diabetes.  The FII can then be used by people with type 1 diabetes to determine how much insulin they need for a meal.  Our studies showed that postprandial glycaemia is significantly improved when insulin is dosed according to the FII but there are non-significant trends regarding hypoglycaemia - the risk increases in acute studies but decreased in our 3 month RCT.  More work is needed here to find out what's really happening and to determine the ideal dosing pattern (i.e. dual-wave split and duration).

Ref: Bell et al. Impact of Fat, Protein, and Glycemic Index on Postprandial Glucose Control in Type 1 Diabetes - Implications for Intensive Diabetes Management in the Continuous Glucose Monitoring Era. Diabetes Care, 2015;38(6):1008-15.

While there is no conclusive evidence on how prandial insulin doses need to be adjusted for different meals, the key experts in this field from University of Sydney, John Hunter Children's Hospital in Newcastle, Australia and Joslin Diabetes Center and Boston Children's Hospital in Boston, USA, have collaborated to come up with a flowchart for the adjustment of prandial insulin (see left).

1. Healthy Eating & Carb Counting is Important

This flowchart emphasises the importance of healthy eating and traditional insulin therapy as these two factors are likely to have the biggest impact on blood glucose levels.  Focusing on regular mealtime routines, healthy food choices, remembering insulin doses, and brushing up on carbohydrate counting skills will go a long way towards improving glycaemic control.

2. Identify problematic meals
Once these factors have been optimised, it should be much clearer to see the effects of protein and fat, remembering that everyone will respond slightly differently.  Rather than introducing insulin adjustments for every meal containing protein or fat, it may be easier to start out identifying problematic meals, i.e. those high fat/protein meals that present the biggest blood glucose challenges.  This might be family pizza night, a pub meal with the boys or a cafe big breakfast.  

3. Consider alternative meals or adjust insulin dose

One option is to consider alternative, healthier food choices as part of an overall healthy eating approach in line with the individuals' goals.  The second option is to adjust the insulin dose to better match the physiological need. The evidence suggests that insulin doses may need to be substantially increased (e.g. 65% increase over usual insulin: carb ratio for high fat, high protein meal) but as there is a wide variation in sensitivities to different macronutrients, we suggest starting much more conservatively.  Remember, it is much safer to start small and increase the dose on review.  Based on the evidence, we suggest starting with a 15-20% increase for high protein meals and a 30-35% increase for high fat meals (not additive).  The insulin will need to be delivered either using a dual wave with an insulin pump or as a second dose  approximately an hour after the meal commences, rather than all upfront to reduce the risk of hypo and ensure the insulin is available to cover the late postprandial period.

Remember, as with all insulin titration, it is vitally important to monitor closely and review frequently.

Ref: Bell et al. Impact of Fat, Protein, and Glycemic Index on Postprandial Glucose Control in Type 1 Diabetes - Implications for Intensive Diabetes Management in the Continuous Glucose Monitoring Era. Diabetes Care, 2015;38(6):1008-15.

Clinical Practice Translation