1. Examining the data on human NPH and analogue basal insulin
Update job no and date
Prescribing information is available on the last slide.

2. Insulin therapy in type 2 diabetes
Why is it important?

3. Background to insulin therapy in type 2 diabetes
Type 2 diabetes is the more prevalent form of diabetes, comprising 90% of all cases1
Characterised by insulin deficiency and/or insulin resistance1
The condition is progressive
In the UK Prospective Diabetes Study, approximately 75% of people with type 2 diabetes required multiple therapies to achieve glycaemic targets within 9 years of diagnosis2
Many patients eventually need insulin because of advanced insufficiency in insulin secretion2
Genuth S et al. Diabetes Care 2003; 26(Suppl 1): s28-32.
Turner RC et al. JAMA 1999; 281:
This slide aims to set the scene for the meeting introducing the idea that as pancreatic beta-cell function declines, there is a progression from use of oral medication to insulin.
1. Genuth S et al. Diabetes Care 2003; 26(Suppl 1): s28-32.
2. Turner RC et al. JAMA 1999; 281:

4. The natural history of type 2 diabetes
100 80 60 40
Time (years)
Beta-cell function (%) 20
Time of diagnosis
50% beta-cell function at diagnosis
-10 -8 -6 -4 4 2 -2 6
Extrapolation of decline in beta-cell function suggesting that deterioration in beta-cell function may commence 10–12 years before diabetes is diagnosed1
This slide serves to remind delegates that type 2 diabetes is a progressive condition as a result of beta-cell function declining over time – hence there is a need to gradually intensify the approach to glucose-lowering therapy.
Graph adapted from Holman RR. Diabetes Res Clin Pract 1998; 40(Suppl): S21–25. and UK Prospective Diabetes Study 16, Diabetes 1995: 44;1249–1258
1. Holman RR. Diabetes Res Clin Pract 1998; 40(Suppl): S21–25.

5. Patients (n=339) on a sulphonylurea requiring additional insulin (%)
UKPDS: Many people with type 2 diabetes will eventually need insulin therapy
Patients (n=339) on a sulphonylurea requiring additional insulin (%)
Wright A et al. Diabetes Care :
Six years after diagnosis, 53% of people treated with sulphonylureas required additional treatment to maintain fasting plasma glucose levels <6.0 mmol/L
UKPDS: UK Prospective Diabetes Study.
Graph adapted from Wright A et al. Diabetes Care 2002; 25: 330‒336.

6. Why is it important to intervene early when glycaemic control deteriorates?

7. Lessons from the UKPDS: Better control means fewer complications1
For every 1% (11 mmol/mol) reduction in HbA1c
Relative risk reduction* 1%
Deaths from diabetes
21%
Heart attacks
14%
Microvascular complications
37%
Amputation or death from peripheral vascular disease
Stratton IM et al. BMJ 2000; 321:
It is also known from the UKPDS that addressing glycaemic control in type 2 diabetes reduces the risk of microvascular and macrovascular complications.
43%
*p< UKPDS: UK Prospective Diabetes Study.
1. Stratton IM et al. BMJ 2000; 321: 405‒412.

8. Legacy effect in the UKPDS: Early intensive glycaemic control continues to benefit in the long term
Intensive vs conventional treatment from randomisation (1977–1991) to trial end (1997)1,2
10-year follow-up1 9%
p=0.04
12%
p=0.029
15%
p=0.01
16%
p=0.052
24%
p=0.001
25%
p=0.0099
Holman RR. N Engl J Med 2008; 359:
UKPDS Group. Lancet 1998; 352:
This slide is from the UKPDS study and its 10-year follow-up, and demonstrates that there are long term benefits of early glycaemic control, which is known as the legacy effect.
Any diabetes-related endpoint
Microvascular disease
Myocardial infarction
Risk reductions of 10% and 6% were observed in the intensive group vs the conventional group for diabetes-related death and all-cause mortality, respectively; however this was not statistically significant2
Intensive treatment with sulphonylurea or insulin vs conventional treatment with diet
UKPDS: UK Prospective Diabetes Study
1. Holman RR. N Engl J Med 2008; 359: 1577‒1589; 2. UKPDS Group. Lancet 1998; 352: 837‒853.

9. When should insulin be initiated?

10. However, in the UK, HbA1c is often close to 86 mmol/mol (10%) at the time of insulin initiation1
If mean glucose levels (HbA1c) are rising and unresponsive to current diabetes treatments, consider initiation of insulin especially if:
Concordance with therapy is good2
Lifestyle measures are realistically optimised3
There is an acute medical event causing acute metabolic deterioration or planned surgical intervention3
There are osmotic symptoms2
Situations in which to be cautious when considering insulin initiation:
Rising weight alongside rising HbA1c (likely dietary indiscretion), because weight gain is likely to be exacerbated with insulin, resulting in minimal benefit3
As insulin therapy is a risk factor for hypoglycaemia, patients with impaired cognitive function may have delayed recognition of the symptoms hypoglycaemia
Khunti K et al. Diabetes Care 2013; 36(11): 3411–3417.
Scottish Intercollegiate Guidelines Network (SIGN) 116 Management of diabetes. A national clinical guideline
Home P et al. Diabetes Care 2014; 37: 1499–1508.
Yaffe K et al. JAMA Intern Med 2013; 173: 1300–1306.

11. UK retrospective cohort study (DIN-link database)1
What are the pre-insulin HbA1c levels by the time of insulin initiation in type 2 diabetes?
UK retrospective cohort study (DIN-link database)1
Additionally, in the SOLVE study the majority of people with type 2 diabetes were poorly controlled at the point at which the decision was taken to initiate basal insulin treatment2
Mean pre-insulin HbA1c was highest in the UK (9.8 ± 1.8%) and lowest in China (8.3 ± 1.7%)2
Calvert MJ et al. Br J Gen Pract 2007; 57:
Khunti K et al. Diabetes, Obesity and Metabolism 2012; 14:
This slide emphasises that insulin initiation is often started too late. Both the SOLVE study, published last year and the study by Calvert from 2007 show that people with type 2 diabetes often do not start basal insulin treatment until their HbA1c level is close to 86 mmol/mol or 10%.
SOLVE: Study of Once Daily Levemir
Calvert MJ et al. Br J Gen Pract 2007; 57: 455‒460
Khunti K et al. Diabetes, Obesity and Metabolism 2012; 14: 654‒661.

12. What are the benefits and risks/limitations of insulin therapy?
Achieve tight glycaemic control
Avoid long-term disease complications
Hypoglycaemia
Weight gain
Regular glucose monitoring
Increased patient involvement
This slide shows some of the benefits and risks/limitations for insulin therapy, and that optimal use is based on achieving a balance between these two. ? Ref needed

13. Section summary
Type 2 diabetes is characterised by a progressive decline in beta-cell function1
Many people with type 2 diabetes will require insulin therapy when other treatment becomes ineffective
Early and sustained intervention to improve glycaemic control reduces the risk of microvascular and macrovascular complications in the long term, and is associated with a ‘legacy effect’2,3
However, in the UK, insulin initiation often does not take place until HbA1c is about 86 mmol/mol (10%)3
Optimal insulin use is based on achieving a balance between the benefits and the risks or limitations associated with its use
Holman RR. Diabetes Res Clin Pract (Suppl): S21–25.
Holman RR. N Engl J Med 2008; 359:
Khunti K et al. Diabetes Care 2013; 36(11):
1. Holman RR. Diabetes Res Clin Pract 1998; 40(Suppl): S21–25.
2. Holman RR. N Engl J Med 2008; 359: 1577–1589.
3. Khunti K et al. Diabetes Care 2013; 36(11): 3411–3417.

14. Examining the data on human NPH and analogue basal insulin
Are the differences clinically relevant?
Is this a complete new section? If so, use upper case on all first letters of words in title.

15. This session will discuss:
The insulin options available in clinical practice
The reasons why basal insulin is often used for initiation
The data on human NPH and analogue basal insulin with respect to:
Effects on glycaemic control
Occurrence of hypoglycaemia
Impact on body weight
NPH: Neutral protamine Hagedorn.

16. Better efficacy compared with other insulins?
Reduced risk of hypoglycaemia compared with other insulins?
Less weight gain compared with other insulins?
Patient preference?
Local guideline recommendations?
What is the main reason why you would initiate a basal insulin analogue?
This is a keypad question slide for delegates to respond to and to stimulate discussion at the start of this session.

17. The most appropriate insulin should be selected in a clinical evaluation process
Sub-optimal
control
Clinical evaluation
process
Most appropriate regimen
This slide illustrates that when initiating insulin for a person with type 2 diabetes, a process of clinical evaluation should guide the healthcare professional to the most appropriate treatment regimen for that person.
Once-daily insulin
Twice-daily insulin
Multiple daily injection insulin
Other
solutions

18. Questions to consider in the clinical evaluation process when selecting the appropriate insulin initiation regimen
What is the indication?
HbA1c ‘creep’/symptoms/cardiovascular risk
What is the goal?
HbA1c or plasma glucose targets/risk factor reduction/symptom reduction
What is the ‘metabolic milieu’?
Baseline weight/weight change
Are there relevant co-morbidities?
Cardiac/renal/central nervous system
What are the personal habits of the individual?
Eating pattern, exercise, etc.
Are there specific barriers to insulin?
Employment/driving/needle anxiety/preconceptions of insulin
Has the individual expressed preferences?
Device/regimen
What is the blood glucose experience of the individual?
Target observed hyperglycaemia to achieve best results
This slide shows a selection of questions to consider when selecting the most appropriate insulin initiation regimen for a person with type 2 diabetes.

19. What insulin options are available?

20. Treating-to-target in type 2 diabetes (4-T Study)1
In the 4-T study, in years 1, 2 and 3, if HbA1c levels were >48 mmol/mol (>6.5%), sulphonylurea therapy was stopped and a second type of insulin was routinely added
Year 1
Years 2 and 3
Add biphasic insulin* twice a day
Add prandial insulin at midday
– or –
708 candidates on dual oral agents
Add prandial insulin* three times a day
Add basal insulin before bed R
– or –
Add basal insulin* once (or twice) daily
Add prandial insulin
three times a day
Holman RR et al. N Eng J Med 2009; 361:
The 4-T study evaluated 708 people with type 2 diabetes who had suboptimal glycaemic control while taking metformin and sulphonylurea therapy. Patients were randomly assigned to receive biphasic insulin aspart twice daily, prandial insulin aspart three times daily, or basal insulin detemir once daily (twice if required).
Sulphonylurea therapy was replaced by a second type of insulin if hyperglycaemia became unacceptable during the first year of the study or subsequently if glycated haemoglobin levels were more than 48 mmol/mol or 6.5%.
*Intensify to a combination insulin regimen during year 1 if hyperglycaemia unacceptable.
R: randomisation
1. Holman RR et al. N Eng J Med 2009; 361: 1736‒1747.

21. 4-T study: 3-year efficacy of complex insulin regimens in type 2 diabetes
Initial insulin regimen*
Proportion taking two types of insulin
Median HbA1c (at 3 years)
Proportion achieving ≤48 mmol/mol (6.5%)
Weight gain (kg)
Hypoglycaemic events in all patients**
Biphasic twice daily
(n=235)
67.7%
54 mmol/mol (7.1%)
31.9%
5.7 kg
3.0
Pre-prandial three times daily (n=239)
73.6%
51 mmol/mol (6.8%)
44.8%
6.4 kg
5.7
Basal once daily
(n=234)
81.6%
52 mmol/mol (6.9%)
43.2%
3.6 kg
1.7
Holman RR et al. N Eng J Med 2009; 361: up to here
Median glycated haemoglobin levels were similar for patients receiving biphasic, prandial and basal insulin-based regimens. However, fewer patients had a level of 48 mmol/mol (6.5%) or less in the biphasic group than in the prandial group or in the basal group.
Median rates of hypoglycaemia (grade 2 or 3) per patient per year were lowest in the basal group, higher in the biphasic group, and highest in the prandial group. The mean weight gain was higher in the prandial group than in either the biphasic group or the basal group.
*If glycaemic targets were not met, those allocated basal insulin added rapid acting insulin thrice-daily with meals, those allocated biphasic insulin added rapid acting insulin at lunchtime (midday) and those allocated meal time rapid acting insulin added basal insulin once-daily (twice if needed); **Grade 2 or 3, median number of episodes per patient per year
Table adapted from Holman RR et al. N Eng J Med 2009; 361: 1736‒1747.

22. 4-T study: 3-year efficacy of complex insulin regimens in type 2 diabetes: summary of differences
Outcome
Between group comparisons
Biphasic*
(n=235)
Prandial*
(n=239)
Basal*
(n=234)
Median HbA1c at 3 yrs
No significant differences between groups
HbA1c targets ≤6.5% achieved
Lowest number of patients achieving target
No significant differences
Mean SMBG level achieved
Lower number of patients achieving target compared with prandial
Fewer grade 2 or 3 hypoglycaemic episodes
Intermediate
Greatest number of episodes
Least number of episodes
Less weight gain
Least weight gain
Less increase in waist circumference
Least amount of increase
Median glycated haemoglobin levels were similar for patients receiving biphasic, prandial and basal insulin-based regimens. However, fewer patients had a level of 48 mmol/mol (6.5%) or less in the biphasic group than in the prandial group or in the basal group.
Median rates of hypoglycaemia (grade 2 or 3) per patient per year were lowest in the basal group, higher in the biphasic group, and highest in the prandial group. The mean weight gain was higher in the prandial group than in either the biphasic group or the basal group.
*Initial starting regimen
SMBG: self-monitoring of blood glucose 1. Holman RR et al. N Eng J Med 2009; 361: 1736‒1747.

23. 4-T study: main summary points
Three quarters of patients added a second insulin
Those commencing therapy with a basal or prandial insulin more often achieved glycaemic targets than patients commencing with a biphasic insulin
Patients commencing therapy with basal insulin had fewer hypoglycaemic episodes and less weight gain
These findings provide clear evidence in people with type 2 diabetes to support starting insulin therapy with a once-a-day basal insulin, and then adding a mealtime insulin if glycaemic targets are not met
This slide summarises the findings of the 4-T study.
1. Holman RR et al. N Eng J Med 2009; 361: 1736‒1747.

24. Available human NPH and analogue basal insulins

25. Available human NPH and analogue basal insulins
Disposable pen
Cartridge
Vial
Human NPH insulins
Isophane human insulin ✔
Analogue basal insulins
Insulin glargine
Higher strength insulin glargine
Insulin detemir
Insulin degludec
This slide shows the available human NPH and analogue basal insulins and whether they can be obtained in a disposable pen, cartridge or vial form.
Complete summaries of product characteristics can be found on 
1. Diabetes UK. Meds & Kit Available at: [Accessed August 2015].

26. What is the evidence base regarding the choice of basal insulin at initiation in type 2 diabetes?

27. Initiating human NPH insulin first-line: What do the clinical trials tell us?
Study
Fasting blood glucose
HbA1c
Severe hypoglycaemic episodes
Nocturnal hypoglycaemic episodes
Treat-to-target trial1 ✗ ✔
Comparison of insulin detemir vs human NPH insulin2
✔= significant difference
✗= no significant difference
This slide summarises the differences between human NPH insulin and analogue insulin with reference to two clinical trials which will now be discussed in more detail.
NPH: neutral protamine Hagedorn
1. Riddle MC et al. Diabetes Care 2003; 26: 3080–3086.
2. Philis-Tsimikas A et al. Clinical Therapeutics 2006; 28: 1569–1581.

28. The Treat-to-Target trial: Insulin glargine vs human NPH insulin
Design:
756 overweight people with type 2 diabetes on one or two oral agents
HbA1c >58 mmol/mol (>7.5%)
24-week study
Participants received bed-time glargine or human NPH insulin once daily
Outcome measures:
Fasting plasma glucose
HbA1c
Hypoglycaemia
Percentage of patients reaching HbA1c ≤53 mmol/mol (≤7%) without documented nocturnal hypoglycaemia
The Treat-to-Target trial compared the abilities and associated hypoglycaemia risks of insulin glargine and human NPH insulin added to oral therapy of type 2 diabetes to achieve an HbA1c target of 53 mmol/mol (7%).
This slide summarises the study design and the main outcome measures.
NPH: neutral protamine Hagedorn
Riddle MC et al. Diabetes Care 2003; 26: 3080–3086.

29. Treat-to-Target trial: Both insulins have similar effects on FPG and HbA1c
Fasting plasma glucose
HbA1c 9 75
11.0 8 64 7 53
8.3 6 42
Glargine n=367
NPH n=389
p=NS 5 31
Fasting plasma glucose (mmol/L)
5.5
HbA1c (%)
HbA1c (mmol/mol) 4 20
Glargine n=367
NPH n=389
p=NS 3
2.3 2 1
These figures show that fasting plasma glucose (FPG) decreased smoothly in both groups, reaching a plateau by 12 weeks. Mean FPG at end point was 6.5 mmol/L for glargine and 6.7 mmol/L for NPH (P=not significant).
HbA1c declined at a slower rate, stabilising after 18 weeks. Mean HbA1c at end point was 52.6 mmol/mol (6.96%) with glargine and 52.7 mmol/mol (6.97%) with NPH (P=not significant). 4 8 12 16 20 24 4 8 12 16 20 24
Treatment duration (weeks)
Treatment duration (weeks)
NPH: neutral protamine Hagedorn; FPG: fasting plasma glucose; NS: not significant
Graphs adapted from Riddle MC et al. Diabetes Care 2003; 26: 3080–3086.

30. Treat-to-Target trial: Cumulative incidence of hypoglycaemic events
Significantly fewer hypoglycaemic events with insulin glargine
Events with plasma glucose ≤4.0 mmol/L
Events with plasma glucose ≤3.1 mmol/L
2500
900
800
2000
Glargine
NPH
700
Glargine
NPH
600
1500
Cumulative number of events documented PG ≤72 mg/dL (4 mmol/L)
Cumulative number of events documented PG ≤56 mg/dL (3.1 mmol/L)
500
400
1000
300
500
200
P<0.005
P<0.003
100 24 48 72 96
120
144
168 24 48 72 96
120
144
168
Time (days)
Time (days)
Expressed as hypoglycaemic events per patient year (glargine vs NPH):
9.2 vs 12.9 (P<0.005) for confirmed events with a plasma glucose ≤4.0 mmol/L
3.0 vs 5.1 (P<0.003) for confirmed events with a plasma glucose ≤3.1 mmol/L
These figures show the cumulative incidence of hypoglycaemic events. Fewer events occurred with glargine than NPH, with no tendency for the between treatment difference to decline over time.
NPH: neutral protamine Hagedorn
Graphs adapted from Riddle MC et al. Diabetes Care 2003; 26: 3080–3086.

31. Treat-to-Target trial: Hypoglycaemia by time of day30
1.4 * *
Glargine
NPH *
Glargine
NPH * 25 *
1.2 * *
Insulin injection * 1 * * 20
Insulin injection *
0.8 *
Proportion of patients (%) 15
Events per patent exposure - year *
0.6 10
0.4 5
0.2
2000
2400
0400
0800
1200
1600
2000
2400
0400
0800
1200
1600
Time
Time
*p <0.05 (between-treatment)
*p <0.05 (between-treatment)
Hypoglycaemia with insulin glargine has the same time distribution as human NPH insulin
Insulin glargine does not abolish nocturnal hypoglycaemia
Expressed as events per patient year (glargine vs NPH), rates of nocturnal hypoglycaemia were:
3.1 vs 5.5 (P<0.001) for confirmed events with a plasma glucose level ≤4.0 mmol/L
1.3 vs 2.5 (P<0.002) for confirmed events with a plasma glucose level ≤3.1 mmol/L
Severe hypoglycaemia was similarly uncommon between the two treatments
Significantly more patients experienced hypoglycaemia at night with NPH, but there were no between-treatment differences in the percentage of patients with symptomatic hypoglycaemia confirmed by a measurement of glucose ≤4.0 mmol/L through the day and early evening (see left-hand chart).
Similar patterns were evident for the rates of confirmed hypoglycaemic events per patient-year except for slightly more events at a single daytime time point (11.00–12.00 h) with glargine (see right-hand chart).
n=367 (Glargine); n=389 (NPH)
NPH: neutral protamine Hagedorn
Graphs adapted from Riddle MC et al. Diabetes Care 2003; 26: 3080–3086.

32. Comparison of insulin detemir with human NPH insulin in poorly controlled type 2 diabetes
Objective:
To compare the effectiveness and tolerability of insulin detemir versus human NPH insulin administered once daily together with >1 oral anti-diabetes drug in poorly controlled type 2 diabetes, and to compare different administration times of insulin detemir
Outcome measures:
HbA1c
Fasting plasma glucose
9-point self-measured plasma glucose
Hypoglycaemia
Tolerability
Patient characteristics:
498 people with type 2 diabetes and a BMI ≤40 kg/m2
At least 3 months treatment with ≥1 oral anti-diabetes drug
HbA1c 58–97 mmol/mol (7.5–11%)
20-week study
In another study, insulin detemir was compared with human NPH insulin in people with poorly controlled type 2 diabetes. This slide summarises the characteristics of the study population and the main outcome measures for the study.
NPH: neutral protamine Hagedorn
Philis-Tsimikas A et al. Clinical Therapeutics 2006; 28: 1569–81.

33. Comparison of insulin detemir with human NPH insulin in poorly controlled type 2 diabetes
Morning and evening insulin detemir were associated with reductions in HbA1c similar to those with evening human NPH insulin
FPG was significantly higher at end point in the morning insulin detemir group compared with the two evening insulin group
For the primary end point, mean HbA1c, morning and evening detemir were associated with reductions in HbA1c similar to those with evening NPH. At end of trial, evening detemir was found to be non inferior compared with evening NPH. Morning detemir was also non inferior to evening NPH.
At end of trial, FPG was not significantly different in the evening detemir group compared with that in the evening NPH group. FPG with morning detemir, however, was significantly higher compared with that with evening NPH. FPG with morning detemir was also significantly higher compared with that with evening detemir.
-1.48% (16.17 mmol/mol)
-1.58% (17.27 mmol/mol)
-1.74% (19.02 mmol/mol)
p<0.001
p=0.003
NPH: neutral protamine Hagedorn; FPG: fasting plasma glucose
Philis-Tsimikas A et al. Clinical Therapeutics 2006; 28: 1569–81.

34. Comparison of insulin detemir with human NPH insulin in poorly controlled type 2 diabetes: Hypoglycaemia
Parameter
Morning insulin detemir (n=165)
Evening insulin detemir (n=169)
Evening human NPH insulin (n=164)
Major episodes over 24 hours (requiring third party assistance)
2 events in 2 (1.2%) patients
All confirmed episodes over 24 hours (plasma glucose <3.1 mmol/L and self-managed)
91 events in 32
(19.4%) patients
82 events in 27
(16.0%) patients
153 events in 53
(32.3%) patients
Relative risk of hypoglycaemia over 24 hours vs evening human NPH
0.68
0.47* –
Nocturnal major episodes (occurring between 11pm and 6am)
Nocturnal confirmed episodes (plasma glucose <3.1 mmol/L and self-managed)
6 events in
4 (2.4%) patients
19 events in
8 (4.7%) patients
47 events in
22 (13.4%) patients
Relative risk of nocturnal hypoglycaemia vs evening human NPH
0.13†
0.35‡
*p=0.019; †p<0.001; ‡p=0.031.
NPH: neutral protamine Hagedorn
Table adapted from Philis-Tsimikas A et al. Clinical Therapeutics 2006; 28: 1569–81.

35. Human NPH insulin vs glargine or detemir: Meta-analysis within UK Health Technology Assessment (1)
HbA1c
No significant difference found in HbA1c levels between glargine and human NPH insulin, or between insulin detemir and human NPH insulin
Hypoglycaemia
No differences in the frequency of severe hypoglycaemia between the analogues and human NPH insulin
Overall, hypoglycaemia was less frequent with both glargine (OR* 0.74, 95% CI to 0.89) and detemir (OR* 0.51, 95% CI 0.35 to 0.76)
Many of the hypoglycaemic episodes were nocturnal:
Glargine – OR* 0.47 (95% CI 0.37 to 0.59)
Detemir – OR* 0.48 (95% CI 0.37 to 0.63)
*Absolute risk not calculated in this meta-analysis. CI: confidence interval; NPH: neutral protamine Hagedorn; OR: odds ratio.
Waugh N et al. Health Technology Assessment 2010; 14(36); 1–248.

36. Human NPH insulin vs glargine or detemir: Meta-analysis within UK Health Technology Assessment (2)
Weight
There was no clinically or statistically significant difference observed in terms of weight gain with glargine compared with human NPH insulin (−0.28 kg in favour of insulin glargine; 95% CI –0.72 to 0.15)
For insulin detemir compared with human NPH insulin, the difference was greater, and was statistically significant (1.2 kg in favour of insulin detemir; 95% CI 1.6 to 0.8)
Authors’ summary
“Glargine and detemir are equivalent to NPH (and to each other) in terms of glycaemic control as reflected in HbA1c level, but have modest advantages in terms of hypoglycaemia, especially nocturnal”
CI: confidence interval; NPH: neutral protamine Hagedorn.
Waugh N et al. Health Technology Assessment 2010; 14(36); 1–248.

37. Long-acting insulin analogues: An independent view
“Although this is not secure beyond dispute, the evidence for the long-acting analogues indicates that better fasting glucose control can be achieved with less risk of nocturnal hypoglycaemia in those subjected to aggressive dose titration. In the absence of such titration, they offer little or no benefit. There is, in other words, little point in prescribing these insulins in less intensively managed patients, or in those at low risk of nocturnal hypoglycaemia – which effectively excludes the great majority of patients with type 2 diabetes as currently managed.”
Holleman F & Gale E. Diabetologia 2007; 50: 1783–1790.

38. Summary
The choice of insulin at initiation should be based on a clinical evaluation process
Basal insulin should be considered as one of these options
When comparing long-acting analogues with human NPH insulin for basal insulin treatment, both regimens appear to give comparable reductions in HbA1c with no significant difference in the occurrence of severe hypoglycaemia
There is evidence to suggest a benefit of long-acting insulin analogues over human NPH insulin when considering rates of nocturnal hypoglycaemia when insulin is titrated to meet titration protocols of randomised clinical studies
NPH: neutral protamine Hagedorn.