Mortality and natural progression of type 1 diabetes patients enrolled in the Rwanda LFAC program from 2004 to 2012

  • Sara L. Marshall
  • Deborah V. Edidin
  • Vincent C. Arena
  • Dorothy J. Becker
  • Clareann H. Bunker
  • Crispin Gishoma
  • Francois Gishoma
  • Ronald E. LaPorte
  • Vedaste Kaberuka
  • Graham Ogle
  • Wilson Rubanzana
  • Laurien Sibomana
  • Trevor J. Orchard
Original Article

Abstract

The natural history and mortality of type 1 diabetes in adolescents in Africa is not well characterized. Our aim is, therefore, to describe these characteristics for cases in the Rwanda Life For a Child (LFAC) program. Participants (≤25 years old) were the first 500 children and youth enrolled in the Rwanda LFAC program from 2004 to 2012. Clinical and demographic data were extracted from LFAC forms, and vital status was evaluated as of November 1, 2011. For the first 500 participants, 5-year survival was 93.8% while crude mortality was 13.9/1000 (95% CI 9.0–20.6/1000) person years of diabetes. However, since vital status is unknown for 134 participants, mortality could be as high as 40.2/1000 person years of diabetes if all missing cases died. Mortality was directly associated with age at diagnosis, and inversely to calendar year of first visit, BMI, and monitoring frequency. Hypertension prevalence reached 46% by 2012. Mortality rates associated with type 1 diabetes in Rwanda are similar to those in other African countries, but higher than rates in developed countries. Delayed diagnosis may contribute to excess mortality risk, but recent improvements in survival suggest that advancements are being made. Hypertension and loss to follow-up need to be addressed.

Keywords

Type 1 diabetes Youth Children Natural history Rwanda Africa 

Introduction

Diabetes is a disease of growing concern in the developing world. An estimated 18.7 million people in Africa will be affected with diabetes by 2025 [1], posing a large problem for a population that already has limited access to healthcare and insulin. Help for such countries usually comes in the form of external support. One such program is the Life For a Child (LFAC) program, which is managed by the International Diabetes Federation in conjunction with Australian Diabetes Council and HOPE worldwide. LFAC’s mission is to support the provision of the best possible diabetes healthcare, given local circumstances, by supplying children and adolescents (≤25 years) in developing countries with the necessary insulin and glucose testing supplies and with HbA1c testing capability. The program also offers diabetes education, and advanced training and advice to both patients and local healthcare providers.

The Association Rwandaise des Diabetiques (ARD) is a diabetes association located in Kigali City, Rwanda, that receives support from the LFAC program. The program there was initiated in 2004 with 25 children, and has expanded over the years and now provides support for over 630 children. Participants in this program are provided insulin on a monthly basis and in turn are required to undergo an annual clinic evaluation with suggested quarterly visit follow-up (as of 2010). The short-term effects of this program have been previously reported [2], but little is known concerning the mortality and natural progression of the disease in the members of the Rwanda LFAC program who have been enrolled since its inception.

The objective of this report is, therefore, to describe the status of the first 500 cases registered with the ARD since 2004. We will report on the utilization of the LFAC program (number and frequency of visits), losses to follow-up, clinical measures, and mortality as of November 2011.

Research design and methods

The University of Pittsburgh’s IRB determined that this project was exempt from review under the “Existing Data” category.

Study population

This report focuses on the first 500 children and youth enrolled in the Rwanda LFAC program from 2004 to 2012. To be enrolled, participants must be living in Rwanda, ≤25 years old, and needing assistance with obtaining insulin and other diabetes supplies.

Data collection

Although LFAC registration initiated in 2004, vital status and clinic visit results were first recorded beginning in 2009, and thus, this report comprises clinical data from 2009 to 2012. Clinical and demographic data [sex, province, date of birth, diagnosis date, insulin regimen, glucose monitoring frequency, height, weight, systolic and diastolic blood pressure (BP), tuning fork vibratory sensation and monofilament response tests, HbA1c, and albumin creatinine ratios (A/C ratios)] were extracted from LFAC clinical forms.

Laboratory data

Blood (finger prick) and urine (spot samples) were collected from each patient and processed on the Siemens DCA Vantage™ (which reports DCCT-related values). Data for HbA1c and A/C ratio were collected from these samples. The maximum HbA1c value for this machine is “>14% (>130 mmol/mol),” so for data analysis purposes, these results were reported as “14.1% (131 mmol/mol).” The inter-assay CV range for the HbA1c measures was from 2.1 to 3.8% during the data collection.

Complication assessment

Neuropathy was defined as failure to feel a 10g monofilament (less than seven of ten correct responses) on the dorsum of the great toe and/or failure to feel vibration from a 128-Hz turning fork placed on the dorsum of the great toe for 10 s [3]. Microalbuminuria (MA) was defined as an albumin/creatinine (A/C) ratio of 30–299 mg/g in a spot urine sample and overt nephropathy as an A/C ratio ≥300 mg/g. Hypertension was defined as having either SBP or DBP over the 95th percentile for those <18 years, or systolic BP ≥130 mmHg or diastolic BP ≥80 mmHg or a history of BP medication for those ≥18 years. The percentage of those using BP medication, however, is low due to limited availability.

Vital status assessment

Vital status is reported as of November 1, 2011 through clinic attendance, deaths reported to the ARD, hospital surveys, and additional tracing by a forensic epidemiologist (R.W.). A participant was considered to be alive as of this date if he/she attended clinic after November 1, 2011, or he/she was known to be alive by contact with hospital staff or other participants, or through investigation. Autopsies are not customary in Rwanda and death certificate data is likewise sparse. Thus, cause of death was determined through contact with local hospitals and families, but is limited in both quality and completeness.

Data analysis

Descriptive statistics were calculated for clinical and utilization data. ANOVA, two-sample, and paired t tests were used for comparisons of continuous variables, while χ 2 tests and Fisher’s exact tests were for comparisons of categorical variables. Tukey’s HSD test was used for any post hoc pairwise comparisons. A p value of 0.05 with appropriate Bonferroni corrections was used to assess significance for multiple comparisons. Time between visits was calculated assuming that a year had 365.25 days and a month had 30.44 days.

Person years of diabetes were calculated as of November 1, 2011 (n = 361; total known person years of diabetes = 1792 years). Ninety-five percent confidence intervals were computed assuming a Poisson distribution. Kaplan-Meier curves were constructed to estimate cumulative survival and log-rank tests were used to test for significant differences in survival between sub-groups. Survival curves were censored at 10 years as less than 20% of the original cohort remained after this duration. Cox regression models were used to examine differences in survival for all continuous variables as well as simultaneously adjusting for multiple variables (model 1). The proportional hazards assumptions of the models were confirmed by testing time-dependent interaction variables. All hazard ratios (HRs) are reported per increase of 1 unit of measure.

All statistical analysis was performed using SAS 9.2 (SAS Institute, Cary, NC) statistical software.

Results

Of the first 500 registered LFAC participants, a total of 488 (97.6%) participants have at least one recorded clinic visit, while 12 were registered by name only (received insulin, but have no demographic or clinical data). Seven participants had unrecorded sex. Mean duration of diabetes was 4.9 ± 3.2 years and only 18 (3.8%) participants were diagnosed before age 5 years, while 186 (39.3%) were diagnosed between the ages of 15 and 19 years (Figs. 1 and 2). Seventy-five of the original 500 registered participants (15%) are 26 years or older and are no longer eligible for support from LFAC.
Fig. 1

Histogram of diagnosis age for the first 500 LFAC program participants, by sex

Fig. 2

Flow chart outlining participant follow-up and vital status

Utilization

Seventy-seven (15.4%) participants were registered only with no clinical visit (including 12 with no demographic or clinical data) and 53 (10.6%) had only one visit in 2009–2012. Those who had only one visit from 2009 to 2012 were significantly older than those who had multiple visits and were diagnosed at an older age than participants with no visit (Table 1).
Table 1

Characteristics of participants with multiple clinic visits from 2009 to 2012, those who only had one clinic visit, and those who have no recorded clinic visits

 

Visit status

Multiple visits

One visit

No visitb

N

370

53

65

Age at diagnosis (years)

15.4 ± 4.9

16.8 ± 5.1

12.7 ± 5.8a

Male (column % (n))

43.0 (156)

43.4 (23)

44.6 (29)

Age in 2012 (years)

20.8 ± 4.6

22.3 ± 5.0

20.7 ± 4.9

Province (row % (n))

   

 East

73.1 (38)

7.7 (4)

19.2 (10)

 Kigali City

71.5 (88)

9.7 (12)

18.7 (23)

 North

84.2 (48)

8.8 (5)

7.0 (4)

 West

76.2 (64)

15.5 (13)

8.3 (7)

 South

72.2 (130)

10.6 (19)

17.2 (31)

Year of first visit % (n)

   

 2009

159 (87.4)

23 (12.6)

 2010

126 (87.5)

18 (12.5)

 2011

82 (88.2)

11 (11.8)

 2012

3 (75.0)

1 (25.0)

aSignificantly different from those with multi-visits and one visit

bExcludes an additional 12 with no visit, but with no demographic or clinical data reported

Three hundred seventy participants (74%) had multiple visits from 2009 to 2012 [mean number of visits of 5.1 ± 2.8 (range 1–12 visits)], and their mean HbA1c decreased significantly from 11.1 ± 2.7% to 9.5 ± 2.5% (98 ± 30 mmol/mol to 80 ± 28 mmol/mol) between their first and last visit (Table 2). At the most recent visit, 3.5% (n = 2/57) had neuropathy, 10.1% (n = 10/99) had overt nephropathy, and 18.2% (n = 18/99) had MA. These rates were not significantly different from those seen at the first visit (neuropathy 2.2%, n = 5; overt nephropathy 5.3%, n = 8; MA 18.5%, n = 28; Table 2). The prevalence of hypertension, however, increased significantly over time (35.1 to 46.1%).
Table 2

Clinic data from the first and most recent clinic visit for participants who had multiple visits between 2009 and 2012 (N = 370)

 

First clinic visit

Most recent clinic visit

Time between visits (months)

 

26.6 ± 10.8

Age (years)

18.6 ± 4.5

20.6 ± 4.5*

Male % (n)

43.0 (156)

 

Age at diagnosis (years)

15.4 ± 4.9

 

Duration of diabetes (years)

3.0 ± 2.9

5.0 ± 3.1*

Glucose monitoring (per week)

1.3 ± 3.9

8.0 ± 8.0*

HbA1c (mmol/mol)

98 ± 30

80 ± 28*

HbA1c (%)

11.1 ± 2.7

9.5 ± 2.5*

 <8.0%

17.0 (55)

28.9 (96)*

 8–11.9%

40.9 (132)

51.5 (171)*

 12–14%

14.6 (47)

9.6 (32)*

 >14%

27.6 (89)

9.6 (32)*

Insulin injections (per day)

1.8 ± 0.6

2.0 ± 0.5*

Units of insulin (per day)

32.4 ± 15.7

37.8 ± 15.4*

Units of insulin per kilogram

0.70 ± 0.37

0.75 ± 0.30

Weight (kg)

47.6 ± 12.6

51.6 ± 11.6*

Height (cm)

154.5 ± 14.3

156.7 ± 12.8*

BMI (kg/m2)

19.8 ± 3.5

20.7 ± 3.2*

Systolic BP (mmHg)

113 ± 16

120 ± 17*

Diastolic BP (mmHg)

73 ± 12

79 ± 12*

Neuropathy % (n)

2.2 (5)a

3.5 (2)b

Microalbuminuria % (n)

18.5 (28)c

18.2 (18)d

Nephropathy % (n)

5.3 (8)c

10.1 (10)d

Hypertension % (n)

35.1 (127)f

46.1 (165)g*

*Significantly different from first clinic visit (p < 0.05)

a n tested = 230

b n tested = 57

c n tested = 151

d n tested = 99

f n tested = 362

g n tested = 358

Within the last year (since November 1, 2011), 319 (63.8%) of the first 500 registered participants have been seen. Of the 181 (36.2%) who have not been seen, 33 are known to have died, 5 were found to be duplicate IDs, and 17 are otherwise known to be alive (Fig. 3). Forty of those not seen in the last year have aged out of the LFAC program. The most common reported reasons for non-attendance were being away at boarding school (n = 4), lack of transport (n = 3), sick on day of visit (n = 3), did not hear radio announcement (n = 2), were pregnant (n = 2), or no longer believed they had diabetes (n = 2).
Fig. 3

Survival by sex (a) and year of first clinic visit (2009 vs. all subsequent years) (b)

Vital status

A total of 25 participants [n = 15 (60.0%) females; n = 10 (40.0%) males] were known to have died as of November 1, 2011. Thus, the crude mortality for this cohort was 6.9% (25/361; 95% CI, 4.5–10.2%) or 13.9/1000 person years of diabetes (95% CI 9.0–20.6/1000). [Note: an additional eight participants died in 2012; therefore, a more accurate estimate of mortality is 9.1% (33/361; 95% CI 6.3–12.8) or 18.1/1000 person years of diabetes.] Since vital status was unknown for 134 (26.8%) participants, in a worst-case scenario (assuming all had died), the mortality rate could be as high as 32.1% (159/495; 95% CI 27.0–37.1%) or 40.2/1000 (95% CI 32.0–49.9/1000) person years of diabetes; the most optimistic (assuming all are alive) could be as low as 5.0% (25/495; 95% CI 3.3–7.4) or 12.3/1000 (95% CI 7.9–18.1/1000) person years of diabetes.

For those who died as of November 1, 2011, cause of death was unknown for 16 (64.0%), with hypoglycemia being the most common known cause (n = 4, 16.0%), and renal failure accounting for 2 (8.0%) deaths. Single deaths resulted from gastroenteritis, pneumonia, pulmonary embolism, rectal hemorrhage/hepatitis, and hyperglycemia. Unfortunately, none of the reported deaths had an official autopsy. Mean age at time of death was 19.4 ± 4.0 years (range 12–25 years), mean age at diagnosis for the deceased was 14.1 ± 5.0 years (range 1–23 years), and mean diabetes duration was 4.5 ± 3.5 years (range 0–11 years). Twenty-two of the deceased (88.0%) reported having no glucose meter at home versus 33% (n = 99) of those who were seen in the last year (p < 0.0001). Of those who died, 11 had a recorded HbA1c value (mean 9.8 ± 2.0%; 83 ± 22 mmol/mol), with a mean number of HbA1c measures of 2.1 ± 2.0 per person. This was significantly fewer than those alive (p = 0.002). There were no significant associations between province and mortality, though the Western Province had the highest number of deaths (n = 9).

Five-year (post-diagnosis) survival for this cohort was 93.8% (85.1% worst-case scenario) and 10-year survival was 82.5% (66.2% worst-case scenario). Survival did not differ significantly by sex (Fig. 3a), province, year of first visit, year of diagnosis, year of birth, diagnosis before or after 15 years of age, or complication status. However, those whose first visit was in 2009 had significantly higher mortality than the subsequent years combined (p = 0.03) (Fig. 3b), suggesting that survival is improving over time.

In Cox regression age at diagnosis, BMI, monitoring frequency, number of injections, and weight at last clinic visit were univariately (all negatively except for age at diagnosis) related to survival, as was weight at baseline (Table 3). In a multivariable model, each additional year in age at diabetes diagnosis resulted in a 15% higher mortality risk (HR 1.15, 95% CI 1.03–1.29), each additional unit of BMI decreased the risk of death by 22% (HR 0.78, 95% CI 0.67–0.91), and each additional monitoring per week decreased risk by 7% (HR 0.93, 95% CI 0.86–0.99).
Table 3

Cox regression models for mortality (HRs and their associated 95% CIs are reported)

 

Variables that were univariately significant

Model 1

HR (95% CI)

p

HR (95% CI)

p

Age at diagnosis (years)

1.12 (1.01–1.2)

0.01

1.15 (1.03–1.29)

0.01

BMI at most recent visit

0.80 (0.70–0.92)

0.002

0.78 (0.67–0.91)

0.001

Monitoring frequency at most recent visit (times per week)

0.91 (0.85–0.97)

0.006

0.93 (0.86–0.99)

0.046

Number of injections at most recent visit (per day)

0.50 (0.26–0.98)

0.03

N/S

Weight at baseline (kg)

0.96 (0.93–0.99)

0.02

N/S

Weight at most recent visit (kg)

0.94 (0.91–0.97)

0.002

N/S

Discussion

In this study of the natural history of the first 500 participants registered in the IDF’s LFAC program in Rwanda, we estimated the crude mortality to be 13.9/1000 person years of diabetes and determined that mortality was directly associated with age at diabetes diagnosis and inversely to year of first visit, weight at baseline, and monitoring frequency. However, since vital status is unknown for 134 participants, mortality could be as high as 40.2/1000 person years of diabetes. For the 310 participants with multiple visits since 2009, we saw a significant decrease in HbA1c (11.1 ± 2.7% to 9.5 ± 2.5%; 98 ± 30 mmol/mol to 80 ± 28 mmol/mol) and consistent rates of complications except for hypertension, which rose significantly.

Our estimated mortality and survival rates (5-year survival = 93.8–85.1%; 10-year survival = 82.5–66.2%) are consistent with previous studies in African youth, e.g., Ethiopia (mortality = 15.5/1000 person years) [4] and South Africa (10-year survival = 84%) [5], though 5-year survival in Tanzania was significantly poorer (71–60%) [6]. Mortality in youth with type 1 diabetes from developed countries are much lower rates than Rwanda, ranging from 0.06% in the UK [7] to 6.1/1000 person years of diabetes in Lithuania (Table 4) [8, 9, 10, 11], though it is consistently lower in African Americans than Caucasian Americans [12].
Table 4

Mortality comparisons for Rwanda and other studies

Country

Crude mortality (%) (95% CI)

Mortality per 1000 person years (95% CI)

5-year survival (%)

10-year survival (%)

Rwanda

6.9 (4.5–10.2)

13.9 (9.0–20.6)

93.8

82.5

Rwanda (worst-case scenario)

32.1 (27.0–37.1)

40.2 (32.0–49.9)

85.1

66.2

Rwanda (best-case scenario)

5.0 (3.3–7.4)

12.3 (7.9–18.1)

Ethiopia

15.5

South Africa

84

Tanzania

71–60

United Kingdom

0.6

Lithuania

0.04

6.1

97.3

94.0

Estonia

0.02

3.7

99.0

94.3

Finland

0.006

0.8

99.8

99.1

US African Americans (30 years)

15.8

98

96

The association we found between mortality and diagnosis age is supported by results from Mozambique and Zambia [13], and Estonia, Lithuania, and Finland [8]. Each additional year of age at diagnosis conferred a higher risk of mortality, and we believe this may partially be due to the effects of surviving for several years with undiagnosed diabetes. We hypothesize that by the time some of these participants are formally diagnosed, they are in especially poor condition and therefore at a higher risk of complications and death. This ability to survive with apparent type 1 diabetes for years without insulin is consistent with the presence of a different type of diabetes in this community as discussed later. Our findings that higher BMI and weight at baseline is protective also support this hypothesis as we have previously found that those in worse control have lower weight consistent with lack of insulin and dehydration/hypovolemia [2].

The majority of those who were deceased reported not having a meter at home. Compounding this, adequate emergency services are often distant or unavailable at certain times (e.g., at night), and glucagon was not available due to its cost. Thus, it is not unexpected that a common known cause of death was hypoglycemia. This finding is also consistent with others showing that within the first 10 years, cause of death is primarily due to acute causes [4, 11, 14] and underscores the importance of ensuring that all LFAC members have access to glucose meters and strips and appropriate education, especially in light of the high rate of food insecurity in this population. It is, however, very likely that several of the deaths due to unknown causes were from DKA, but were not properly identified as such. Therefore, we cannot say that hypoglycemia is the most common cause of death, as over half were due to unknown causes.

Although 18.8% of our initial cohort has been lost to follow-up, these results are similar to other mortality studies in Africa [5, 14, 15]. The high turnover rate of the nursing (and medical) staff at the district hospitals has also limited our follow-up of LFAC participants, and many of the new nurses have never seen the patients we are trying to locate. We are currently working on developing awareness programs within schools, working with local nurses to help provide for transport, and developing local support groups of children that provide phone chains and social incentives to attend visits.

Over the course of the LFAC program, several individuals have come back to the clinic years after their last insulin dose (n = 21 after 2 years; n = 2 after 3 years) without apparently maintaining insulin therapy. This phenomenon has also been recorded in children thought to have type 1 diabetes in Ethiopia who had interrupted insulin supplies (for 9 ± 12 weeks; range 1–78 weeks), of whom only 4% developed any DKA [16]. While these individuals were in poor condition when they returned, the fact that they survived so long without any exogenous insulin suggests that there may be a different type of diabetes (with some significant residual beta cell function) present in this population. This would be consistent with results from other studies from Ethiopia, suggesting the previously recognized “Malnutrition-related Diabetes” (MRDM) should be re-considered in Africa [17, 18, 19]. Unfortunately, due to the lack of access to typology testing at this time, we were unable to officially determine diabetes type for participants.

The prevalence of hypertension increased over time and was also elevated in this population in comparison to rates reported in US African Americans (AfAm = 9.8%, Rwanda 46.1%) [20]. We have previously postulated that this is in part due to poorer glucose control in Rwanda, as those whose control worsened over time saw more dramatic increases in BP than those who were in good control or who saw improved control [21]. This increase in hypertension prevalence, and the positive association between hypertension and MA, highlights the need for improved BP control in this cohort for complication and mortality prevention.

The key limitations of this study are the large proportion lost to follow-up and the fact that mortality data was limited by lack of autopsy data and a formal national death index.

The major strengths of this study, however, are that our cohort of 500 participants is the largest cohort for whom mortality of type 1 diabetes in sub-Saharan Africa has been reported, and this is the first follow-up and mortality report for diabetic youth and adolescents in Rwanda.

Conclusion

In summary, these data demonstrate that mortality rates for those with type 1 diabetes in Rwanda are similar to other African type 1 diabetes populations, but higher than those in developed countries. Delayed diagnosis of type 1 diabetes may contribute to the increased risk of mortality, highlighting the importance for increased awareness and timely diagnosis. However, the improvements in survival since 2009 are encouraging and reflect global trends [22]. Utilization of the LFAC program in Rwanda has increased significantly since its inception in 2004, and HbA1c has decreased considerably as a cumulative result of its efforts. The relationship of improved survival with increased frequency of self-monitoring of blood glucose, and more frequent injections provide clear guidelines to improve care. Although several participants have died and many more have been lost to follow-up, we hope that by recognizing and addressing the identified barriers, more extensive and long-term care will be available in the future.

Notes

Acknowledgements

We would like to thank the staff of the ARD in Kigali for all of the help and support they have given to us throughout this project. We also thank the Ministry of Health and Rwanda Biomedical Center for their assistance and guidance. We would also like to acknowledge the LFAC program for providing us with the opportunity to work with them and the ARD. Finally, we thank the wonderful children and youth with diabetes and their caregivers for their willing acceptance of help and the dignity and courage they show in coping with diabetes in the most difficult circumstances.

Compliance with ethical standards

Funding

No funding was received for this study. Insulin, syringes, glucose meters, and strips were received free from the LFAC program. DCA machines, HbA1C, and A/C reagents were provided by the University of Pittsburgh.

Conflict of interest

Sara L. Marshall, Deborah V. Edidin, Vincent C. Arena, Dorothy J. Becker, Clareann H. Bunker, Crispin Gishoma, Francois Gishoma, Ronald E. LaPorte, Vedaste Kaberuka, Graham Ogle, Wilson Rubanzana, and Laurien Sibomana have no conflict-of- interest to declare. Dr. Trevor Orchard serves on Eli Lilly and Company Advisory Panel.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Authors’ contributions

Marshall SL designed the study, performed research and analysis and wrote the paper; Orchard TJ designed the study, reviewed analysis, revised the manuscript; Edidin DV, Gishoma C, Gishoma F, Kaberuka V, Sibomana L, and Rubanzana W performed research, collected data, and revised the manuscript; Arena VC reviewed analysis and revised the manuscript; Becker DJ, Bunker CH, Ogle G, and LaPorte RE provided oversight and revised the manuscript.

References

  1. 1.
    International Diabetes Federation. Diabetes atlas, fourth edition. Brussels, Belgium; 2009.Google Scholar
  2. 2.
    Marshall SL, Edidin D, Sharma V, Ogle G, Arena VC, Orchard T. Current clinical status, glucose control, and complication rates of children and youth with type 1 diabetes in Rwanda. Pediatr Diabetes. 2013;14(3):217–26.CrossRefPubMedGoogle Scholar
  3. 3.
    Walsh MG, Zgibor J, Borch-Johnsen K, Orchard TJ. A multinational assessment of complications in type 1 diabetes: the DiaMond substudy of complications (DiaComp) level 1. Diabetes & vascular disease research : official journal of the International Society of Diabetes and Vascular Disease. 2006;3(2):80–3.CrossRefGoogle Scholar
  4. 4.
    Lester FT. Clinical features, complications and mortality in type 1 (insulin-dependent) diabetic patients in Addis Ababa, Ethiopia, 1976–1990. The Quarterly journal of medicine. 1992;83(301):389–99.PubMedGoogle Scholar
  5. 5.
    Gill GV, Huddle KRL, Monkoe G. Long-term (20 years) outcome and mortality of Type 1 diabetic patients in Soweto, South Africa. Diabetic Medicine. 2005;22(12):1642–6.CrossRefPubMedGoogle Scholar
  6. 6.
    McLarty DG, Athaide I, Bottazzo GF, Swai AM, Alberti KG. Islet cell antibodies are not specifically associated with insulin-dependent diabetes in Tanzanian Africans. Diabetes Res Clin Pract. 1990;9(3):219–24.CrossRefPubMedGoogle Scholar
  7. 7.
    Edge JA, Ford-Adams ME, Dunger DB. Causes of death in children with insulin dependent diabetes 1990–96. Arch Dis Child. 1999;81(4):318–23.CrossRefPubMedPubMedCentralGoogle Scholar
  8. 8.
    Podar T, Solntsev A, Reunanen A, Urbonaite B, Zalinkevicius R, Karvonen M, LaPorte RE, Tuomilehto J. Mortality in patients with childhood-onset type 1 diabetes in Finland, Estonia, and Lithuania: follow-up of nationwide cohorts. Diabetes Care. 2000;23(3):290–4.CrossRefPubMedGoogle Scholar
  9. 9.
    O’Grady MJ, Delaney J, Jones TW, Davis EA. Standardised mortality is increased three-fold in a population-based sample of children and adolescents with type 1 diabetes. Pediatr Diabetes. 2013;14(1):13–7.CrossRefPubMedGoogle Scholar
  10. 10.
    Warner DP, McKinney PA, Law GR, Bodansky HJ. Mortality and diabetes from a population based register in Yorkshire 1978-93. Arch Dis Child. 1998;78(5):435–8.CrossRefPubMedPubMedCentralGoogle Scholar
  11. 11.
    Secrest AM, Becker DJ, Kelsey SF, Laporte RE, Orchard TJ. Cause-specific mortality trends in a large population-based cohort with long-standing childhood-onset type 1 diabetes. Diabetes. 2010;59(12):3216–22.CrossRefPubMedPubMedCentralGoogle Scholar
  12. 12.
    Secrest AM, Becker DJ, Kelsey SF, LaPorte RE, Orchard TJ. All-cause mortality trends in a large population-based cohort with long standing childhood-onset type 1 diabetes: the Allegheny County type 1 diabetes registry. Diabetes Care. 2010;33(12):2573–9.CrossRefPubMedPubMedCentralGoogle Scholar
  13. 13.
    Beran D, Yudkin JS, De Courten M. Access to care for patients with insulin-requiring diabetes in developing countries: case studies of Mozambique and Zambia. Diabetes Care. 2005;28(9):2136–40.CrossRefPubMedGoogle Scholar
  14. 14.
    McLarty DG, Kinabo L, Swai AB. Diabetes in tropical Africa: a prospective study, 1981–7. II. Course and prognosis. BMJ (Clinical research ed). 1990;300(6732):1107–10.CrossRefGoogle Scholar
  15. 15.
    Gill GV, Huddle KR, Rolfe M. Mortality and outcome of insulin-dependent diabetes in Soweto, South Africa. Diabetic Medicine. 1995;12(6):546–50.CrossRefPubMedGoogle Scholar
  16. 16.
    Habtu E, Gill G, Tesfaye S. Characteristics of insulin requiring diabetes in rural northern Ethiopia—a possible link with malnutrition? Ethiop Med J. 1999;37(4):263–7.PubMedGoogle Scholar
  17. 17.
    Siraj ES, Reddy SSK, Scherbaum WA, Abdulkadir J, Hammel JP, Faiman C. Basal and postglucagon C-peptide levels in Ethiopians with diabetes. Diabetes Care. 2002;25(3):453–7.CrossRefPubMedGoogle Scholar
  18. 18.
    Alemu S, Dessie A, Seid E, Bard E, Lee PT, Trimble ER, Phillips DIW, Parry EHO. Insulin-requiring diabetes in rural Ethiopia: should we reopen the case for malnutrition-related diabetes? Diabetologia. 2009;52(9):1842–5.CrossRefPubMedGoogle Scholar
  19. 19.
    Gill GV, Tekle A, Reja A, Wile D, English PJ, Diver M, Williams AJK, Tesfaye S. Immunological and C-peptide studies of patients with diabetes in northern Ethiopia: existence of an unusual subgroup possibly related to malnutrition. Diabetologia. 2011;54(1):51–7.CrossRefPubMedGoogle Scholar
  20. 20.
    Mayer-Davis EJ, Beyer J, Bell RA, Dabelea D, D’Agostino R, Imperatore G, Lawrence JM, Liese AD, Liu L, Marcovina S, Rodriguez B. Diabetes in African American youth: prevalence, incidence, and clinical characteristics: the SEARCH for Diabetes in Youth Study. Diabetes Care. 2009;32(Suppl 2):S112–22.CrossRefPubMedPubMedCentralGoogle Scholar
  21. 21.
    Marshall SL, Edidin DV, Arena VC, Becker DJ, Bunker CH, Gishoma C, Gishoma F, LaPorte RE, Kaberuka V, Sibomana L, Orchard TJ. Glucose control in Rwandan youth with type 1 diabetes following establishment of systematic, HbA1c based, care and education. Diabetes Res Clin Practice. 2015;107(1):113–22.CrossRefGoogle Scholar
  22. 22.
    Miller RG, Secrest AM, Sharma RK, Songer TJ, Orchard TJ. Improvements in the life expectancy of type 1 diabetes: the Pittsburgh Epidemiology of Diabetes Complications study cohort. Diabetes. 2012;61(11):2987–92.CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Research Society for Study of Diabetes in India 2016

Authors and Affiliations

  • Sara L. Marshall
    • 1
  • Deborah V. Edidin
    • 2
  • Vincent C. Arena
    • 1
  • Dorothy J. Becker
    • 3
  • Clareann H. Bunker
    • 1
  • Crispin Gishoma
    • 4
  • Francois Gishoma
    • 4
  • Ronald E. LaPorte
    • 1
  • Vedaste Kaberuka
    • 4
  • Graham Ogle
    • 5
  • Wilson Rubanzana
    • 6
  • Laurien Sibomana
    • 1
  • Trevor J. Orchard
    • 1
  1. 1.University of Pittsburgh Graduate School of Public HealthPittsburghUSA
  2. 2.Feinberg School of MedicineNorthwestern UniversityChicagoUSA
  3. 3.University of Pittsburgh School of MedicinePittsburghUSA
  4. 4.Association Rwandaise des DiabetiquesKigaliRwanda
  5. 5.International Diabetes Federation Life for a Child Program and Australian Diabetes CouncilSydneyAustralia
  6. 6.National University of RwandaKigaliRwanda

Personalised recommendations