Lung function in poorly controlled type 1 North African diabetic patients: A case-control study

Egyptian Journal of Chest Diseases and Tuberculosis (2015) 64, 717–727

H O S T E D BY

The Egyptian Society of Chest Diseases and Tuberculosis

Egyptian Journal of Chest Diseases and Tuberculosis www.elsevier.com/locate/ejcdt www.sciencedirect.com

ORIGINAL ARTICLE

Lung function in poorly controlled type 1 North African diabetic patients: A case-control studyq Ines Slim a,b,1, Ferdaws Khalaf a,b,1, Imed Latiri c, Zouhour Elfkih a,b, Sonia Rouatbi c,d, Ines Khochtali e, Ines Ghannouchi c,d, Abir Zinelabidine f, Leila Ben Othman f, Hedi Miled g, Larbi Chaieb a,b,2, Helmi Ben Saad c,d,h,*,2 a

Department of Endocrinology and Diabetology, Farhat HACHED University Hospital of Sousse, Tunisia Endocrinology and Metabolic Diseases Unit, 02/UR/08-07, Faculty of Medicine of Sousse, University of Sousse, Sousse, Tunisia c Laboratory of Physiology, Faculty of Medicine of Sousse, University of Sousse, Sousse, Tunisia d Department of Physiology and Functional Exploration, Farhat HACHED University Hospital of Sousse, Tunisia e Department of Endocrinology and Diabetology, Fattouma BOURGUIBA University Hospital of Monastir, Tunisia f Laboratory of Biochemistry, Basic Health Group, Sousse, Tunisia g Laboratory of Biochemistry, Farhat HACHED University Hospital of Sousse, Tunisia h Research Laboratory N LR14ES05: Interactions of the Cardiopulmonary System, Faculty of Medicine of Sousse, University of Sousse, Tunisia b

Received 3 December 2014; accepted 15 February 2015 Available online 26 March 2015

KEYWORDS Endocrinology; Diabetes mellitus; Lung function tests; Tunisia;

Abstract Aim: To compare the lung function parameters of poorly controlled type-1-diabetesmellitus (T1DM) patients with age-; height and sex-matched healthy-non-smokers (HNS). Population and methods: Subjects aged 35–60 Yrs who have a poorly controlled T1DM (glycated-Haemoglobin level >7%) with a disease history of more than 10 Yrs (n = 14) and HNS subjects (n = 14) were recruited. Clinical, anthropometric and fasting biological data were

Abbreviations: ATS, American-thoracic-society; BMI, body-mass-index; CLA, chronological-lung-age; DLCO, capacity-to-transfer-carbonmonoxide; DN4, douleur-neuropathique-4-questions; ELA, estimated-lung-age; ERS, European-respiratory-society; FEV1, first-second-forcedexpiratory-volume; FVC, forced-vital-capacity; HbA1c, glycated-haemoglobin; HDL-cholesterol, high-density-lipoprotein-cholesterol; HNS, healthy-non-smokers; LDL-cholesterol, low-density-lipoprotein-cholesterol; LLN, lower-limit-of-normal; MMEF, maximal-mid-expiratory-flow; NY, narghile-years; LAOVD, large-airways-obstructive-ventilatory-defect; PY, pack-years; RV, residual-volume; RVD, restrictive-ventilatorydefect; SD, standard-deviation; SVC, slow-vital-capacity; T1DM, type-1-diabetes-mellitus; T2DM, type-2-diabetes-mellitus; TGV, thoracic-gasvolume; TLC, total-lung-capacity; 95% CI, 95% confidence interval q The French version of the present study abstract was accepted as POSTER DISCUSSION at the annual Congress of the French Society of Pulmonology (SPLF, 31 January, 2 February 2015, Lille, France; http://www.sciencedirect.com/science/article/pii/S0761842514004112). Name and location of the institution where the study was performed: service of Physiology and Functional Exploration and Department of Endocrinology and Diabetology, Farhat HACHED Hospital. Sousse, Tunisia. * Correspondent author at: Laboratory of Physiology, Faculty of Medicine of Sousse, Street Mohamed KAROUI, Sousse, Tunisia. Tel.: +216 98697024; fax: +216 73224899. E-mail address: [email protected] (H. Ben Saad). 1 These authors contributed equally as first authors to this work. 2 These authors contributed equally as senior authors to this work. Peer review under responsibility of The Egyptian Society of Chest Diseases and Tuberculosis. http://dx.doi.org/10.1016/j.ejcdt.2015.02.013 0422-7638 ª 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of The Egyptian Society of Chest Diseases and Tuberculosis. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

718 Aging; Case-control study

I. Slim et al. collected. Plethysmographic data (flows, volumes, estimated-lung-age (ELA), lung-capacity-totransfer-carbon-monoxide (DLCO)) were measured. Large-airway-obstructive-ventilatory-defect (LAOVD) was defined as first–second-forced-expiratory-volume (FEV1)/forced-vital-capacity (FVC) below the lower-limit-of-normal (LLN). Restrictive-ventilatory-defect (RVD) was defined as total-lung-capacity (TLC) < LLN. Lung-hyperinflation was defined as residual-volume (RV) > upper-limit-of-normal. Student t-test and chi-2 test were used to compare plethysmographic data and profiles of the two groups. Results: The two groups were matched in chronological-lung-age (CLA) (respectively 47 ± 7 vs. 50 ± 8 Yrs) and sex (7 males and 7 females in each group) and height. Compared to the HNS group, the T1DM one had significantly lower FEV1, FVC, slow-vital-capacity and maximal-midexpiratory-flow (respectively 99 ± 11% vs. 83 ± 11%, 99 ± 9% vs. 86 ± 11%, 80 ± 8% vs. 67 ± 15% and 98 ± 23% vs. 72 ± 23%), had significantly higher TLC and RV (respectively, 105 ± 20% vs. 123 ± 24% and 108 ± 22% vs. 131 ± 24%) and had significantly higher percentage of subjects with lung-hyperinflation (7.1% vs. 43.0%). Both groups had similar percentages of LAOVD and RVD and similar corrected DLCO values. ELA of the T1DM group (57 ± 10 Yrs) was significantly higher than CLA. Conclusion: Poorly controlled T1DM seems to alter ventilatory mechanics without effect on the alveolo-capillary-membrane. In addition, it accelerates the respiratory ageing. ª 2015 The Authors. Production and hosting by Elsevier B.V. on behalf of The Egyptian Society of Chest Diseases and Tuberculosis. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction Diabetes-mellitus (DM), classified among the top ten leading causes of death worldwide [1], is becoming a major public health emergency [2]. In the Maghreb, a region undergoing an epidemiological transition characterized by a decrease of infectious diseases and an increase in chronic non-infectious ones, the prevalence of DM was as high as 9.5% in 2012 [3]. There are two major types of DM: type 1 (T1DM) and type 2 (T2DM)); differentiated according to the aetiopathogenic factors and the rapidity of pancreatic beta cells of Langerhans islets apoptosis [4]. Since the discovery of insulin, the outcome of T1DM has completely changed. Treatment options and extrarenal-epuration techniques are considered as a turning point in the disease history leading to the improvement of its outcome [5]. Since then, we are no longer worried about the direct mortality of the disease as much about morbidity associated to metabolic emergencies, and to micro- and macroangiopathy [6] which, in turn, may have a negative impact on the function of internal organs [7]. In fact, diabetic microangiopathy specifically affects the eyes (retinopathy), kidney (nephropathy) and peripheral nervous system (neuropathy) [7]. Since it possesses a very wide capillary network and a significant amount of connective tissue, the lung could be a suitable ‘‘target organ’’ [8]. This was earlier suggested by Schuyler et al. [9] more than 40 Yrs ago. These authors [9] investigated lung function in 11 patients with T1DM and age-matched normal control subjects. This pilot study was the first to report measurements of nearly all the available tests of lung function, including lung elasticity, capacity-to-transfer-carbon-monoxide (DLCO), absolute thoracic-gas-volumes (TGV), airflow resistance and maximal forced-vital-capacity (FVC) tests [9]. As their subjects were lifelong non-smokers without allergies or lung disease, the T1DM lung elastic recoil decrease was interpreted to reflect effects of DM on lung elastic proteins [9]. That was hence the first results in the literature suggesting that the lung may be a ‘‘target organ’’ of T1DM. Since then, studies analysing T1DM ventilatory mechanics and/or pulmonary exchanger are more

numerous with controversial conclusions: no effect [9–13], abnormal data [14–27] with conflicting reports on the nature of spirometric alterations: restrictive [14–17] and/or large-airways obstructive [16,17]-ventilatory-defects (respectively, RVD, and LAOVD) and/or decreased DLCO [14,15,18, 23,26], and normal spirometry data but decreased DLCO [27]. In addition, the majority of these studies [9–27] used limited methodology: combination of the two types of DM [14,19– 22]; inclusion of T1DM patients with a high difference in disease duration [17]; inclusion of both controlled and uncontrolled T1DM [21]; lack of control groups such as healthy-subjects [17,19]; measurements of expiratory flows only [14,17,20, 21,24], without measurement of lung volumes and/or lung-hyperinflation; non-application [10,17,21,22,24] of the latest lung function international recommendations [28–30], no report of the applied lung function guidelines [14,19,20,23]; use of unspecified spirometric norms [10,14,17,19,22–24] or of inappropriate spirometric definitions [14,17,19,23] (e.g. application of a fixed thresholds of 70% or 80% as a lower-limit-of-normal (LLN)) and no calculation, in case-control studies [10,14,20,22– 24], of the required sample size which is a statistically crucial point [31]. A recent metaanalysis [32] concluded that when studied in the absence of overt pulmonary comorbidity, both T1DM and T2DM were associated with a modestly impaired pulmonary function in a restrictive pattern. The results were irrespective of body-mass-index (BMI), smoking, DM duration, and glycated-haemoglobin (HbA1c) levels. In subanalyses, the association seemed to be more pronounced in T2DM than in T1DM [32]. Taking into account those conflicting conclusions about the effect of T1DM on lung function and the remaining unanswered question about the reality of diabetic pulmonary alteration that was recently risen [7], the present study aimed to compare the lung function parameters (i.e. plethysmographic and DLCO data measured according to recent lung function test guidelines [28–30]), of a poorly controlled T1DM group (HbA1c >7%) having a history of disease more than 10 Yrs with those of age-, sex- and height matched

Lung Function in Uncontrolled Type 1 Diabetes Mellitus healthy-non-smokers (HNS) . The null hypothesis is that there will be no difference between lung function data mean values in both groups. Population and methods Study design It is a case-control study performed over a period of sixmonths with collaboration between Departments of Physiology and Functional Exploration, of Endocrinology and Diabetes, Laboratories of Hematology and of Biochemistry (University Hospital of Farhat HACHED) and of Biochemistry (Basic Health Group of Sousse), Sousse, Tunisia. The study approval was obtained from the Hospital’s Ethics Committee. Written and informed consent was asked from all study participants. All participants have received a report of their explorations. Sample size The null hypothesis was H0:m1 = m2 and the alternative hypothesis was Ha:m1 = m2 + d where d is the difference between two means and n1 and n2 are the sample size for two groups (T1DM and HNS) such that N = n1 + n2. The total sample size was estimated using the following formula [31]: N = ((r + 1)(Za/2 + Z1 b)2r2)/rd2 where Za is the normal deviate at a level of significance (=2.58 for 1% level of significance); Z1 b is the normal deviate at 1 b% power with b% of type II error (=1.28 at 90% statistical power); r equal to n1/n2, is the ratio of sample size required for two groups (r = 1 gives the sample size distribution as 1:1 for the two groups). r and d are the pooled standard-deviation (SD) and difference of first–second-forced-expiratory-volume (FEV1) means of the two groups. These two values were obtained from a previous study of similar hypothesis [23] where researcher found the mean FEV1 (%) in the two groups were 95.19% and 81.28% respectively and a common SD of 13.48%. The total sample size for the study was 28 subjects (14 T1DM and 14 HNS). Study population Patients’ group Only poorly controlled T1DM patients aged 35–60 Yrs with a history of T1DM of more than 10 Yrs were included. Non-inclusion criteria were: controlled T1DM (HbA1c 67%); T2DM, recent infection or acute metabolic complication (occurred less than one week before the lung function measurements), history of asthma, allergic rhinitis, atopy or chronic obstructive pulmonary disease and oral corticosteroid treatment within four weeks prior to lung function test or bronchodilators use, patient under dialysis. Researchers have verified the folders and files of all patients with T1DM followed in the Department of Endocrinology and Diabetes; and have contacted them. Control group HNS aged 35–60 Yrs were recruited among Hospital workers and the parents of medical school students. Informational

719 letters clarifying the aims of the study were put up at the Hospital and the local Medical School. In addition, an article announcing the need for recruitment of healthy subjects was posted in a social network service (Facebook pages of the persons implicated in the study). Non-inclusion criteria were: history of smoking, DM, confirmed cardiovascular or pulmonary diseases, respiratory symptoms (e.g. chronic cough, wheezing, dyspnoea Pstage two [33]), thoracic surgery, mental disease and chronic medication use. The discovery of a LAOVD and/or a RVD was an exclusion criterion. Medical questionnaire A medical questionnaire [34] was used to assess several subjects’ characteristics (smoking, medical, surgical and gynaecologist-obstetrics histories, medication use, several T1DM characteristics and dyspnoea [33]). Cigarette and narghile use was evaluated, respectively, in pack-Yrs (PY) and narghileYrs (NY) [35]. Two groups of patients who have successfully weaned off smoking (ex-smokers) were defined [0. No, 1. Yes]. T1DM duration (Yr) and specific complications (retinopathy (no/yes), dialysis (no/yes)) were noted. Histories of cardiovascular diseases were searched: angina, heart rhythm disorder and arterial hypertension. Physical examination Sex and age (Yr) were noted. Height (±0.01 m) was measured with a height gauge shoes removed, heels joined, and back straight. Weight (±1 kg) was measured and the BMI (kg/m2) was calculated. Waist circumference (m) was measured [36]. Neurological exam and ‘‘douleur neuropathique 4 questions’’ (DN4) score [37] were conducted in order to assess diabetic neuropathy. Two groups were identified [0. No symptomatic neuropathy (DN4 score <4); 1. Symptomatic neuropathy (DN4 score P4)]. T1DM diagnosis, metabolic data and applied metabolic definitions The positivity of anti-pancreatic antibodies (antiGAD and/or antiIA2) at the diagnosis of DM was the proof of the auto-immune aetiology [38]. The following haematological data were measured/calculated: numbers of blood cells (white (103/mm3), red (106/mm3) and platelets (103/mm3)) and haemoglobin level (g/dl). Anaemia was defined as haemoglobin level <12 g/dl in female and <13 g/dl in male [39]. Two groups were defined [0. No anaemia; 1. Anaemia]. Leukocytosis [40] was defined as white-blood-cell count >11.103/mm3. Two groups were defined [0. No leukocytosis; 1. Leukocytosis]. Fasting-glycaemia (mmol/l), total-cholesterol (mmol/l) and high-density-lipoprotein-cholesterol (HDL-cholesterol, mmol/l) were quantified by spectrophotometry. Triglycerides (mmol/l) were quantified by the enzymatic-colorimetric method. Low-density-lipoprotein-cholesterol (LDL-cholesterol, mmol/l) was calculated [41]. HbA1c (%) was quantified on haemolysed total blood (turbidimetric inhibition immunoassay). Poorly controlled T1DM was retained when HbA1c >7%. Diabetic nephropathy was assessed by searching for increased urinary albumin (mg/24 h) excretion using 24-h collections and serum creatinine (lmol/L)

720 [42]. Two groups were defined [0. No diabetic nephropathy (albuminuria <30 mg/24 h); 1. Diabetic nephropathy (albuminuria P30 mg/24 h)]. Lung function measurements Plethysmographic measurements were performed with a body plethysmograph (ZAN 500 Body II, Mebgrera¨te GmbH, Germany), carefully following international recommendations [29,30]. The plethysmographic technique and the FVC manoeuvre are extensively described in the Supplementary data section. The following plethysmographic parameters were measured/calculated and compared to local predicted values [43]: FEV1 (L), FVC (L), maximal-mid-expiratory-flow (MMEF, L/s), slow-vital-capacity (SVC, L), FEV1/FVC and FEV1/SVC ratios (absolute values), TGV (L), residual-volume (RV, L) and total-lung-capacity (TLC, L). The following definitions were applied: LAOVD: FEV1/SVC and/or FEV1/FVC ratios upper-limit-of-normal [45]. DLCO was measured by the single-breath method performed following international recommendations [28]. Two DLCO measurements in sitting position were performed on each subject. Subjects were asked to assume the appropriate position five minutes before the test with an interval of at least 15 min between each DLCO measurement (the best value as a percentage of predicted values). The following parameters were measured/calculated and compared with European predicted values [46]: DLCO (mmol/min/kPa; %), corrected DLCO (DLCO predicted for haemoglobin [28] (mmol/min/kPa)). A DLCO lower than the LLN was considered as abnormal [28]. In order to evaluate respiratory aging, estimated-lungage (ELA, Yrs) was calculated [47]. Study conduct First visit (day one, between 14 hrs. and 16 hrs.): presentation of the subject to the Department of Endocrinology and Diabetes; signature of consent; medical and DN4 questionnaires and anthropometric data. Second visit (day two, between 9 a.m. and 11 a.m.): fasting blood and urine samples, injection of usual dose of basal and/or prandial insulin (the thighs site of injection was avoided); normal and standard breakfast intake (including 200 cc of milk or equivalent in carbohydrate intake and two 40-carbohydrates portions of bread or equivalent); lung function measurements. Data analysis The Kolmogorov–Smirnov test was used to analyse variable distribution. When the distribution was normal and the variances were equal, the results were expressed by their means ± SD and 95% confidence interval (95%CI). Otherwise, results were expressed by their medians (1st–3rd quartiles). For plethysmographic data, a percentage of change was calculated (=(T1DM value HNS value)/T1DM value) [24]. T-test and chi-2 test were used to compare, respectively, quantitative data and percentages. T-test was used to compare quantitative data of T1DM non-smokers vs. T1DM smokers,

I. Slim et al. of male T1DM vs. male HNS and of female T1DM vs. female HNS. The associations between plethysmographic data expressed in percentage predict and HbA1c or T1DM duration were evaluated by Pearson’s product–moment correlation. Wilcoxon-test was used to compare the ELA with the chronological-lung-age (CLA) of the each group. Analyses were carried out using the Statistica statistical software (Statistica Kernel version 6; Stat Software. France). Significance was set at the 0.05 level. Results Descriptive data Among the 62 examined subjects, only 28 (14 HNS and 14 poorly controlled T1DM) were retained for analysis. The two group characteristics’ are presented in Tables 1 and 2. The main conclusions of these Tables are: (i) both groups were age, sex and height matched; (ii) compared to the HNS group, the poorly controlled T1DM one had significantly lower weight, BMI and waist circumference. It also has significantly higher percentages of subjects having history of cardiovascular diseases or dyslipidemia or who were cigarette smokers. (iii) Compared to the HNS group, the poorly controlled T1DM one had a significantly higher fasting-glycaemia level, a lower haemoglobin level, a higher percentage of anaemic subjects and a lower percentage of subjects with measured dyslipidemia. The DM duration mean ± SD (minimum–maximum) was 21 ± 8 (10–32) Yrs. Four patients have a cardiovascular disease (arterial hypertension (n = 2), angina with arterial hypertension (n = 1) and heart rhythm disorder (n = 1)). Smoking has been stopped since a mean ± SD duration of 9 ± 8 Yrs in the ex-smokers group. The mean ± SD of cigarettes use of the five male smokers included in the study was 30 ± 15 PY. Only one male patient was ex-narghile smoker (20 NY). Analytical data Lung function data and profiles of the two groups are presented in Table 3. Compared to the HNS group, the poorly controlled T1DM one had significantly lower SVC, FVC, FEV1, MMEF, DLCO (percentage changes were, respectively, +16%, +13%, +16%, +27% and +16%) and had significantly higher TLC, RV (percentage changes were, respectively, 17% and 21%). However, there was no statistically significant difference between both groups in corrected DLCO. Compared to the HNS group, the poorly controlled T1DM one had significantly higher percentages of subjects with abnormal SVC, FEV1, MMEF and RV. Table 4 shows the characteristics of the poorly controlled T1DM group divided by smoking status. There was no significant difference between metabolic data of the two groups and compared with the T1DM non-smokers subgroup, the T1DM ex-smokers group has significantly lower FVC (92 ± 8% vs. 79 ± 10%, respectively) and FEV1 (88 ± 8% vs. 75 ± 11%, respectively). Table 5 shows the comparison between T1DM and HNS groups divided by sex. Compared to HNS females, those with T1DM had significant decreases in FEV1 and FVC. Compared

Lung Function in Uncontrolled Type 1 Diabetes Mellitus Table 1 Characteristics of type 1 diabetes-mellitus (T1DM) and healthy-non-smoker (HNS) groups.

Anthropometric data Male/female Age (Yr) Height (m) Weight (kg) Body-mass-index (kg/m 2) Waist circumference (m) Clinical data DN4 Symptomatic neuropathy Diabetic retinopathy Cigarette use Narghile use Dyspnoea P stage 2 History of Cardiovascular diseases Dyslipidemia

Yes No Yes No Yes No Yes No Yes No

Yes No Yes No Dysthyroidy Yes No Abdominal surgery Yes No

HNS (n = 14)

T1DM (n = 14)

7/7 49.84 ± 7.99 1.63 ± 0.10 80 ± 12 30.4 ± 4.0

7/7 47.43 ± 7.19 +5 1.66 ± 0.10 2 68 ± 12 +15* 25.0 ± 5.0 +18*

0.97 ± 0.09

0.88 ± 0.12

Not reported 0 (0.0%) 14 (100.0%) 0 (0.0%) 14 (100.0%) 0 (0.0%) 14 (100.0%) 0 (0.0%) 14 (100.0%) 0 (0.0%) 14 (100.0%)

4±3

0 (0.0%) 14 (100.0%) 0 (0.0%) 14 (100.0%) 0 (0.0%) 14 (100.0%) 2 (14.3%) 12 (85.7%)

721 Table 2 Biological parameters of type 1 diabetes-mellitus (T1DM) and healthy-non-smoker (HNS) groups.

Percentage change

+9*

8 (57.2%)* 6 (42.8%)* 8 (57.2%)* 6 (42.8%)* 5 (35.7%)* 9 (64.2%)* 1 (7.1%) 13 (92.9%) 1 (7.1%) 13 (92.9%) 4 (28.5%)* 10 (71.5%)* 3 (21.5%)* 11 (78.5%)* 2 (14.3%) 12 (85.7%) 6 (42.8%) 8 (57.2%)

DN4: neurological exam and douleur neuropathique 4 questions. Percentage change = (T1DM value HNS value)/HNS value. Data are mean ± SD for anthropometric data and DN4 and number (%) for others. * p<.05 (test-t, or chi-2): HNS vs. T1DM.

to HNS males, those with T1DM had significant decreases in FEV1, FVC and SVC as well as a significant increase in RV. No significant correlation between HbA1c or DM duration and none of the plethysmographic data was found. The ELA of the T1DM group was significantly higher than the CLA, respectively, 56.70 ± 9.89 vs. 47.43 ± 7.19 Yrs (p = 0.012). However, no statistical significant difference was found between the HNS group ELA and CLA, respectively, 52.02 ± 7.18 vs. 47.49 ± 7.99 Yrs (p = 0.454). Discussion The present study shows impaired lung function parameters in 14 patients with poorly controlled T1DM compared to age-, sex- and height matched controls. Therefore, the null hypothesis, that there will be no difference between the two groups’ lung function data mean values’ was rejected. This impairment was associated neither with the T1DM mean duration nor with the HbA1c levels. In addition, T1DM seems accelerating the respiratory ageing.

Biological data Fasting-glycaemia (mmol/L) HbA1c (%) Total-cholesterol (mmol/L) Triglycerides (mmol/L) HDL-C (mmol/L) LDL-C (mmol/L) White-blood-cells (103/mm3) Red-blood-cells (106/mm3) Platelets (103/mm3) Haemoglobin (g/dl) Serum-creatinine (lmol/L) Urinary-albumin (mg/24 h) Biological disorders Anaemia Yes No Leukocytosis Yes No Diabetic Yes nephropathy No

HNS (n = 14)

T1DM (n = 14)

Percentage change

5.39 ± 0.54

10.28 ± 5.60

Not reported 4.70 ± 0.55

10.71 ± 1.45 4.47 ± 0.53

+5

1.11 ± 0.36

0.96 ± 0.53

+14

1.05 ± 0.37 3.15 ± 0.50 6.63 ± 1.16

1.29 ± 0.42 2.75 ± 0.55 7.16 ± 2.16

23 +13 8

4.59 ± 0.32

4.29 ± 0.47

+7

227 ± 44 13.5 ± 1.1 81.79 ± 14.27

249 ± 66 11.6 ± 1.7 91.43 ± 37.01

10 +14* 12

NR

8.91 ± 8.17

2 (14.3%) 12 (85.7%) 0 (0.0%) 14 (100.0%) NR NR

12 (85.7%)* 2 (14.3%)* 1 (7.1%) 13 (92.9%) 2 (14.3%) 12 (85.7%)

91*

NR: not reported. Percentage change = (T1DM value HNS value)/HNS value. Data are mean ± SD for biological data and number (%) for others. * p<.05 (test-t, or chi-2): HNS vs. T1DM.

Discussion of the methodology A Medline/PubMed research was conducted on 8 June 2014 and using the following three Medical Subject Headings ‘‘Diabetes Mellitus, Type 1’’ and ‘‘Respiratory Function Tests’’ and ‘‘Adults’’. Among bibliographies of retrieved articles that were manually reviewed, only international studies published during the last 10 Yrs and all published African studies were retained. Supplementary Tables 1E and 2E describe, respectively, the five international and the four African published studies [10,14,17,19–24]. The above studies have controversial conclusions and used limited methodology:  Combination of the two types of DM [14,19–23]. This could be a source of ‘‘confusion’’ because they have different aetiopathogenies. Unlike T2DM during which the apoptosis occurs slowly and mainly due to glucotoxicity and lipotoxicity [4], T1DM is associated with a very fast and irreversible apoptosis provoked by auto-immune factors that are genetically programmed [48].  Inclusion of T1DM patients with a high difference in disease duration [17]. The DM duration varied from three (range: 0.5–13 Yrs) [17] to 15 ± 7 Yrs [10]. This could be a source

722

I. Slim et al.

Table 3 Lung function data and plethysmographic profiles of type 1 diabetes-mellitus (T1DM) and healthy-non-smoker (HNS) groups. HNS (n = 14) Plethysmographic and SVC (%) FVC (%) FEV1 (%) FEV1/SVC (absolute value) FEV1/FVC (absolute value) MMEF (%) TGV (%) TLC (%) RV (%) DLCO (%) Corrected DLCO (%)

T1DM (n = 14)

Percentage change

DLCO data (data are mean ± SD) 80 ± 8 67 ± 15 +16* 99 ± 9 86 ± 11 +13* 99 ± 11 83 ± 11 +16* 0.8 ± 0.05 0.8 ± 0.10 +0 0.8 ± 0.03

0.8 ± 0.07

+0

98 ± 23 97 ± 7 105 ± 20 108 ± 22 109 ± 21 113 ± 24

72 ± 23 94 ± 12 123 ± 24 131 ± 24 92 ± 22 113 ± 42

+27* +3 17* 21* +16* +0

Lung function profiles (data are number (%)) HNS (n = 14)

T1DM (n = 14)

Probability

Plethysmographic data or DLCO lower than the lower-limit-ofnormal FVC 0 (0.0%) 2 (14.3%) # SVC 4 (28.5%) 12 (85.7%) # FEV1 0 (0.0%) 3 (21.5%) FEV1/FVC 0 (0.0%) 0 (0.0%) FEV1/SVC 0 (0.0%) 0 (0.0%) # MMEF 0 (0.0%) 3 (21.5%) TLC 0 (0.0%) 1 (7.1%) DLCO 1 (7.1%) 3 (21.5%) Corrected DLCO 2 (14.3%) 1 (7.1%) Plethysmographic data higher than the upper-lower-of-normal # RV 1 (7.1%) 6 (43.0%) For abbreviations, see list of abbreviations. %: percent of predicted values. Percentage change = (HNS value T1DM value)/HNS value. * p < 0.05 (test-t): HNS vs. T1DM. # p < 0.05 (chi-2): HNS vs. T1DM.

of misinterpretation, since it was suggested that the longer the duration of T1DM is, the greater the lung function impairment will be [24]. T1DM mean duration of the present study (21 ± 8 Yrs) was higher than what was reported in other studies (Tables 1E and 2E).  Inclusion of both controlled and uncontrolled T1DM (respectively, 40% and 60% in Hickson et al. [21] study). This could be a source of confusion since it was demonstrated that the uncontrolled DM group, when compared to controlled one, has a significant decrease in lung function [20]. Unlike the majority of the published studies [9–27], the present study included only poorly controlled patients.  Lack of the control group such as healthy-subjects [17,19]. Cross-sectional studies [14,24], while practical, do not provide a good basis for establishing causality.  Measurements of expiratory flows only [14,17,20,21,24], without measurement of lung volumes (useful parameters for the diagnosis of RVD [44] and/or lung-hyperinflation

Table 4 Characteristics of type 1 diabetes mellitus group divided by smoking status. Non-smokers (n = 8)

Ex-smokers (n = 6)

Percentage change

6/0 48.96 ± 6.51 59.56 ± 11.95  1.72 ± 0.09 68 ± 10 23.0 ± 3.3 84 ± 7

6 11 7* +1 +13 +8

10.68 ± 4.648

9.75 ± 7.13

+9

11.14 ± 1.458 4.64 ± 0.458

10.13 ± 1.35 4.24 ± 0.56

+9 +9

1.03 ± 0.385

0.87 ± 0.72

+16

1.39 ± 0.359 2.78 ± 0.527 7.23 ± 2.78

1.15 ± 0.48 2.70 ± 0.63 7.06 ± 1.16

+17 +3 +2

4.23 ± 0.41

4.37 ± 0.56

3

268 ± 64

223 ± 64

+17

11.2 ± 0.9

12.2 ± 2.4

9

88.13 ± 40.26

95.83 ± 35.38

9

8.91 ± 9.97

8.91 ± 5.88

+0

0.79 ± 0.07

+8

0.78 ± 0.08

+5

79 ± 10 75 ± 11 64 ± 26 67 ± 6 87 ± 8 119 ± 24 118 ± 16 93 ± 31 136 ± 54

+14* +15* +17 +0 +12 +6 +16 1 42

Anthropometric data Male/female 1/7 CLA (Yr) 46.29 ± 7.90 ELA (Yr) 53.63 ± 7.90  Height (m) 1.61 ± 0.08 Weight (kg) 69 ± 14 BMI (kg.m 2) 26.5 ± 5.72 Waist 91 ± 14 circumference (m) Metabolic data Fasting-glycaemia (mmol/L) HbA1c (%) Total-cholesterol (mmol/L) Triglycerides (mmol/L) HDL-C (mmol/L) LDL-C (mmol/L) White-blood-cells (103/mm3) Red-blood-cells (106/mm3) Platelets (103/mm3) Haemoglobin (g/dl) Serum creatinin (lmol/L) Urinary albumin (mg/24 h)

Plethysmographic and DLCO data FEV1/SVC 0.86 ± 0.11 (absolute value) FEV1/FVC 0.82 ± 0.07 (absolute value) FVC (%) 92 ± 8 FEV1 (%) 88 ± 8 MMEF (%) 77 ± 20 SVC (%) 67 ± 20 TGV (%) 99 ± 12 TLC (%) 126 ± 26 RV (%) 140 ± 26 DLCO (%) 92 ± 13 Corrected DLCO 96 ± 19 (%)

For abbreviations, see list of abbreviations. %: percent of predicted values. Percentage change = (ex-smokers value non-smokers value)/ex-smokers value. Data are mean ± SD for biological data and number (%) for others. * p<.05 (test-t) non-smokers vs. ex-smokers.   p<.05 (Wilcoxon matched pairs test) for the same group.

[45]) or DLCO [10,19,23] (useful for the evaluation of the alveolar-capillary-membrane [28]) or lung-age. One positive point for the present study, as done by others [10,19], was the correction of DLCO according to haemoglobin levels [28]. Lung age, a parameter not previously evaluated, has

Lung Function in Uncontrolled Type 1 Diabetes Mellitus Table 5

723

Comparison between the type 1 diabetes mellitus (T1DM) and healthy non-smoker (HNS) groups divided by sex. Females

Males

T1DM (n = 7)

HNS (n = 7)

T1DM (n = 7)

HNS (n = 7)

Anthropometric data CLA (Yr) Height (m) Weight (kg) BMI (kg.m 2) Waist circumference (m)

46.77 ± 8.40 1.60 ± 0.08 71 ± 14 27.6 ± 5.16 93.86 ± 12.60

49.37 ± 7.94 1.54 ± 0.07 76 ± 12 31.9 ± 2.83 94.7 ± 4.31

48.09 ± 6.37 1.72 ± 0.08 66 ± 10 22.4 ± 3.4 82.29 ± 8.26

50.30 ± 8.64 1.71 ± 0.04 84 ± 12* 28.8 ± 4.5* 99.43 ± 11.30*

Metabolic data Fasting-glycaemia (mmol/L) Total-cholesterol (mmol/L) Triglycerides (mmol/L) HDL-C (mmol/L) LDL-C (mmol/L) White-blood-cells (103/mm3) Red-blood-cells (106/mm3) Platelets (103/mm3) Haemoglobin (g/dl) Serum-creatinin (lmol/L)

11.34 ± 4.59 4.66 ± 0.49 0.96 ± 0.36 1.42 ± 0.38 2.81 ± 0.57 7.30 ± 2.99 4.16 ± 0.39 267 ± 70 11.0 ± 0.8 90.86 ± 42.68

5.40 ± 0.60* 4.86 ± 0.69 1.13 ± 0.36 1.17 ± 0.38 3.18 ± 0.64 6.67 ± 1.35 4.41 ± 0.32 227 ± 46 12.8 ± 1.0* 71.14 ± 6.52

9.21 ± 6.66 4.27 ± 0.52 0.96 ± 0.70 1.15 ± 0.44 2.69 ± 0.57 7.02 ± 1.07 4.42 ± 0.53 231 ± 62 12 ± 2 92.00 ± 33.85

5.39 ± 0.52* 4.54 ± 0.34 1.09 ± 0.40 0.93 ± 0.35 3.13 ± 0.35 6.58 ± 1.03 4.76 ± 0.21 228 ± 47 14 ± 1* 92.43 ± 11.59

Plethysmographic data FEV1 (%) FVC (%) MMEF (%) FEV1/SVC (absolute value) FEV1/FVC (absolute value) SVC (%) TGV (%) RV (%) TLC (%) DLCO (%) Corrected DLCO (%)

87 ± 7 93 ± 8 75 ± 21 0.85 ± 0.11 0.80 ± 0.05 66 ± 21 121 ± 24 139 ± 28 99 ± 12 91 ± 14 91 ± 14

101 ± 9* 103 ± 9* 96 ± 21 0.86 ± 0.05 0.83 ± 0.02 79 ± 8 108 ± 26 114 ± 25 102 ± 7 108 ± 28 108 ± 28

79 ± 13 80 ± 10 68 ± 26 0.81 ± 0.08 0.80 ± 0.10 68 ± 6 125 ± 27 122 ± 18 89 ± 9 94 ± 29 135 ± 49

98 ± 12* 95 ± 9* 99 ± 28 0.83 ± 0.04 0.83 ± 0.04 81 ± 9* 103 ± 13 101 ± 18* 93 ± 3 111 ± 13 118 ± 20

For abbreviations, see list of abbreviations. %: percent of predicted values. Data are mean ± SD. * p<.05 (t-test): T1DM vs. HNS for each sex.

been shown to be an acceptable way to communicate spirometric results for smokers and/or patients with respiratory impairment [47].  Non-application in some studies [10,17,21,22,24] of the latest lung function international recommendations [28–30]. While old recommendations (American-thoracic-society (ATS)1987 [49] or ATS-1995 [50]) were applied in some studies [10,17,21,22,24], the present one predated recent guidelines [28–30]. Surprisingly, the published African studies [14,19,20,23] have not mentioned which spirometry guidelines were used.  Use of unspecified spirometric norms [10,14,17,19,22–24]. This could lead to misinterpretation of spirometry data in a significant proportion of subjects [43] since spirometry results are often expressed as percentage of predicted values derived from local reference equations [43].  Use of inappropriate spirometric definitions (e.g. application of a fixed thresholds of 70% or 80% as a LLN [10,14,17,19– 24]). This approach has been widely criticized and more importantly, clinicians may have to review and revise previous diagnosis [51]. The present study applied the recent international definitions based on a 95%CI [44]. Paradoxically, in some studies [10,20–22,24], the applied spirometric definitions to diagnosis ventilatory-defects were not reported.

 Use of old spirometry equipment [21]. Hickson et al. [21] results were calculated using an equipment (dry rolling-seal spirometers) that gave different results than those currently recommended by the ATS/ERS [29]. However, lung function testing equipment and procedures have been progressively refined over the last 10 Yrs in line with recommendations that have been regularly updated by the ATS/ERS [28–30].  No calculation, in some case-control studies [10,14,20,22–24], of the required sample size. This could be a statistically crucial point [31]. The present study calculated sample size of T1DM (n = 14) which was higher than the sample size of some studies (n = 8 [19]; n = 12 [10]), but was smaller than the sample size of other ones (n = 20 [52]; n = 27 [23,24]; n = 30 [14]; n = 39 [17]). Populations’ characteristics The percentage of T1DM patients with symptomatic diabetic neuropathy (57.2%, Table 1) was similar to that reported in the literature (65.3% [53]). The mean age of the patients group in the present study (47 ± 7 Yrs) was intermediate with those reported in other studies: it varied from 14 ± 6 (range: 7– 28 Yrs) [17] to 54 ± 10 Yrs [22]. It has been proved that

724 T1DM [54] and lung function [55] are sex-dependent. For that reason, and as done in some studies [17,19,21–23], the present one was sex-matched. Limits of the present study One major limitation of the present study was the inclusion of six males T1DM ex-smokers (43%). This inclusion makes the T1DM sample’s profile questionable and the question that may be raised here is: ‘‘should current- or ex-smokers be included in the T1DM patient group?’’ In similar studies (Table 3E) performed among T1DM [26,52], or T2DM [56– 58] or mixed-patients [59,60], 5% [26], 10% [58], 19% [56], 24% [59], 30% [52], 37% [57] and 50% [60] of DM patients were current-smokers and 12.5% [58], 18.4% [59] and 33% [56,60] were ex-smokers. In addition, and due to the high prevalence of smoking in DM patients (22% to 33% of them are smokers [61,62]), the inclusion of subjects with such risk factors make the samples more representative especially of poorly controlled patients. Moreover, comparisons between female or male groups (T1DM vs. HNS) showed the same lung function impairment (Tables 4 and 5). So, restricting the analysis to T1DM females who had never smoked, did not alter the present study findings. At least a recent study [63] have examined pulmonary function in smokers with a >10-PY with and without diabetes. Authors have concluded that participants with DM were observed to have reduced pulmonary function after controlling for known risk factors such as smoking [63]. Finally, in a systematic review and meta-analysis investigating pulmonary function in DM, metaregression analyses showed that between-study heterogeneity was not explained by BMI, smoking, diabetes duration, or HbA1c [32]. Additional discussion about the study design, recruitment method, medical questionnaire and data analysis is detailed in the Supplementary data section. Discussion of the results Effect of T1DM on lung flows and ratios Results about the T1DM effect on peripheral flows (e.g. MMEF) are controversial: unchanged [10,19,24] vs. decreased [14,20,23] values. In the present study, the reduction of MMEF by 27% (Table 3) can be considered as a marker of small airway defect. It is fair to assume that the small airways are a potential target for T1DM and other specified studies are welcome. In a case-control study [10] (12 T1DM non-athletes vs. 12 healthy-non-athletes), no statistical differences between FEV1 and FVC values were found (Table 1E). In another study [19], the FEV1 of the 23 DM patients was qualified as normal (>80%) (Table 2E). However, in line with the majority of studies [14,20–24], the present one confirms that T1DM reduces FEV1 and FVC. These results are a marker of a large airway defect[64]: the FEV1 and/or FVC reduction by almost 0.300 L (13–16%) is higher than the spontaneous variability observed in healthy subjects (of 0.183 L for FEV1 and 0.148 L for FVC [44]) and higher than the 0.200 L (12%) threshold used in reversibility test interpretation [44]. Is T1DM incriminated in the genesis of LAOVD? Data aiming to respond to that question are still controversial: whereas some studies [14,20,23] show a decrease in FEV1/FVC ratio,

I. Slim et al. the present study (Table 3), like others [10,24] (Tables 1E and 2E), does not support this hypothesis and shows unchanged FEV1/FVC ratio between T1DM and HNS groups. In addition, FEV1/SVC was not affected by T1DM, as previously shown by Berriche et al. [19]. Effect of T1DM on lung volumes Data about the effect of T1DM on lung volumes are controversial (Tables 1E and 2E). As observed in the present study, SVC was significantly decreased in some other case-control studies [22,23]. It was, however, qualified as normal (>80%) with Berriche et al. [19]. In practice, SVC is an important parameter used to diagnose LAOVD [64] or ‘‘trapping phenomena’’ [29]. Contrary to Masmoudi et al. [23] study, where TLC was decreased showing a tendency to a RVD, the present study objectifies an increase in TLC values showing a tendency to lung-hyperinflation. Some other studies [10,19] found no significant change in TLC (Tables 1E and 2E). One major result of the present study was the significant increase of RV in T1DM (Table 3). RV is an interesting parameter used in practice to diagnose lung-hyperinflation [45] and the 0.400 L (21%) RV increase is significantly higher than the recommended threshold of 0.300 L (10%) applied in reversibility test interpretation [45]. The present result was in opposition to the findings of Berriche et al. [19] where DM patients RV was lower than 80%. Effect of T1DM on DLCO The present study T1DM group DLCO was significantly decreased (Table 3), which probably indicates a parenchyma dysfunction. However, as DLCO changes with haemoglobin levels and as the T1DM group has a statistically lower haemoglobin level (Table 2) and includes a statistically higher percentage of anaemic subjects, specific adjustments for this parameter (e.g. use of corrected DLCO) should always be made to ensure appropriate interpretation [28]. In the present study, as found by others [10], corrected DLCO values were unchanged (Table 3). On the contrary, a significant decrease in corrected DLCO was noticed in a previous study [23]. However the haemoglobin level was not reported [23] (Table 2E). Plethysmographic and DLCO profiles No patient with T1DM had LAOVD (Table 3). This result, different from the one found by Suresh et al. [17] (7.7% of T1DM present a LAOVD), is explained by the divergence in the applied definitions to retain the LAOVD diagnosis (FEV1 <80% and FVC >80% or FEV1/FVC < 0.70 in Suresh et al. study [17] vs. FEV1/FVC or FEV1/SVC < LLN in the present one). In fact, applied in the present study, Suresh et al. [17] definition gives a percentage of 21.4%. Half of T1DM patients had lung-hyperinflation (Table 3). This result, not previously reported, had a practical interest. It was shown that lung-hyperinflation has important consequences that surpass the framework of respiratory mechanics [45]. Almost seven percent of T1DM patients had RVD (Table 3), which is much lower than the 43.6% reported by Suresh et al. [17]. This difference could be explained by the divergence in the applied definitions to retain the RVD diagnosis (FVC <80% and FEV1/FVC >0.7 [17] vs. TLC < LLN in

Lung Function in Uncontrolled Type 1 Diabetes Mellitus the present study). In fact, the application of Suresh et al. [17] definition, in the present study, gives a percentage of 21.5%. In Abd El-Azeem et al. [14] study, respiratory-defect was predominantly restrictive (no percentage was given). Applied in the present study, the definition used by Abd El-Azeem et al. [14] (normal FEV1/FVC and reduced DLCO) to diagnose RVD, gives a percentage of 21.5%. Almost one fifth of T1DM patients had decreased DLCO values (Table 3). This percentage becomes 7.1%, when DLCO was adjusted for haemoglobin levels (Table 3). To the best of the authors’ knowledge, no previous study has reported a percentage of T1DM with abnormal DLCO.

725 contemplate the lung in the same way as any other complications of T1DM. They should also recognize the size of the problem of pulmonary complications as a consequence of the novel techniques used through the respiratory tract in the treatment of DM such as insulin pump inhaler [68]. Finally, in order to motivate T1DM smokers to stop smoking, it is very useful to convey to them data about their ELA [47,65]. Sources of financial support None.

Correlation between HbA1c or DM duration and plethysmographic data Like some studies [17,19,21] including T1DM patients with a high HbA1c level, the present one, do not find a significant correlation between HbA1c and lung function data. This result confirms that HbA1c, a classical biological data often used to monitor the development of T1DM, does not appear as an indicator of respiratory deficiency. Lung function data, too, does not appear to be influenced by disease duration [14,17,19,21], which is confirmed by the present study, carried out on patients with a mean duration of T1DM of 21 Yrs. However, Meo et al. [24] have concluded that the years of disease showed a dose–response effect on lung function: the longer the duration of disease, the greater the lung function impairment. T1DM and respiratory aging One of the major results of the present study was that T1DM accelerated lung ageing by 8.74 ± 8.63 Yrs. This phenomenon seemed to be accelerated by smoking status. In fact, compared to T1DM non-smokers, those smokers have a significant acceleration of lung ageing (7.35 ± 7.94 vs. 10.61 ± 9.28). This result, not previously described, could be used to encourage smoking cessation [65].

Authors’ contributions IS conceived of the study, and participated in its design, and performed the statistical analysis and coordination and helped to draft the manuscript. FK conceived of the study, participated in its design and performed medical questionnaire and lung function tests, and performed the statistical analysis. IL conceived of the study, participated in its design and performed lung function tests. ZE conceived of the study, participated in its design and performed lung function tests. SR helped to draft the manuscript. IK helped to draft the manuscript. IG performed the lung function tests. AZ realized and interpreted the biological analysis. LBO realized and interpreted the biological analysis. HM realized and interpreted the biological analysis. LC helped to draft the manuscript and approved the final version of the paper. BSH conceived of the study, and participated in its design, and performed the statistical analysis and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.

How T1DM alters lung function? The aetiology of ‘‘diabetic pneumopathy’’ (i.e. a typical histopathological and functional lung involvement) [25] is still a matter of debate. The development of this long-term complication could be explained by the biochemical alteration of connective tissue proteins as well as microangiopathy of pulmonary vessels [8,66], leading to mechanical lung dysfunction and impaired blood gas exchange [67]. The mechanisms by which T1DM could alter lung function are detailed in the Supplementary data section. Practical recommendations It is advisable that patients with T1DM, mainly the poorly controlled ones with duration of disease longer than 10 Yrs, should undergo periodic plethysmographic tests (e.g. one test/year) to assess the severity of lung function impairment [24]. Plethysmography will identify more susceptible patients with T1DM; so that they will be able to take additional preventive measures towards lung damage [24]. These measures will help to prevent lung damage at the initial stage, which often, over in the long run, contributes to morbidity and mortality in diabetic patients [24]. Additionally, physicians should

Conflict of interest The authors declare that they have no conflicts of interest concerning this article.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ejcdt.2015. 02.013. References [1] C.J. Murray, A.D. Lopez, Measuring the global burden of disease, N. Engl. J. Med. 369 (2013) 448–457. [2] R. Bouguerra, H. Alberti, L.B. Salem, C.B. Rayana, J.E. Atti, S. Gaigi, et al, The global diabetes pandemic: the tunisian experience, Eur. J. Clin. Nutr. 61 (2007) 160–165. [3] S. Hammami, S. Mehri, S. Hajem, N. Koubaa, H. Souid, M. Hammami, Prevalence of diabetes mellitus among non

726

[4]

[5]

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

I. Slim et al. institutionalized elderly in Monastir city, BMC Endocr. Dis. 12 (2012) 15. H.E. Thomas, M.D. McKenzie, E. Angstetra, P.D. Campbell, T.W. Kay, Beta cell apoptosis in diabetes, Apoptosis 14 (2009) 1389–1404. R.G. Miller, A.M. Secrest, R.K. Sharma, T.J. Songer, T.J. Orchard, Improvements in the life expectancy of type 1 diabetes: the Pittsburgh epidemiology of diabetes complications study cohort, Diabetes 61 (2012) 2987–2992. D. Daneman, Type 1 diabetes, Lancet 367 (2006) 847–858. K. Kuziemski, K. Specjalski, E. Jassem, Diabetic pulmonary microangiopathy – fact or fiction?, Endokrynol Pol. 62 (2011) 171–176. F. Innocenti, A. Fabbri, R. Anichini, S. Tuci, G. Pettina, F. Vannucci, et al, Indications of reduced pulmonary function in type 1 (insulin-dependent) diabetes mellitus, Diabetes Res. Clin. Pract. 25 (1994) 161–168. M.R. Schuyler, D.E. Niewoehner, S.R. Inkley, R. Kohn, Abnormal lung elasticity in juvenile diabetes mellitus, Am. Rev. Respir. Dis. 113 (1976) 37–41. W. Komatsu, T. Barros Neto, A. Chacra, S. Dib, Aerobic exercise capacity and pulmonary function in athletes with and without type 1 diabetes, Diab. Care 33 (2010) 2555–2557. G. Schernthaner, P. Haber, F. Kummer, H. Ludwig, Lung elasticity in juvenile-onset diabetes mellitus, Am. Rev. Respir. Dis. 116 (1977) 544–546. K. Strojek, D. Ziora, J.W. Sroczynski, K. Oklek, Pulmonary complications of type 1 (insulin-dependent) diabetic patients, Diabetologia 35 (1992) 1173–1176. A. Verrotti, F. Chiarelli, M. Verini, G. Morgese, Pulmonary function in type 1 (insulin-dependent) diabetes mellitus, Diabetologia 36 (1993) 579–580. I. Abd El-Azeem, G. Hamdy, M. Amin, A. Rashad, Pulmonary function changes in diabetic lung, Egypt. J. Chest. Dis. Tuberc. 62 (2013) 513–517. M.S. Boulbou, K.I. Gourgoulianis, V.K. Klisiaris, T.S. Tsikrikas, N.E. Stathakis, P.A. Molyvdas, Diabetes mellitus and lung function, Med. Princ. Pract. 12 (2003) 87–91. P. Makkar, M. Gandhi, R.P. Agrawal, M. Sabir, R.P. Kothari, Ventilatory pulmonary function tests in type 1 diabetes mellitus, J. Assoc. Physicians India 48 (2000) 962–966. V. Suresh, A. Reddy, A. Mohan, G. Rajgopal, P. Satish, C. Harinarayan, et al, High prevalence of spirometric abnormalities in patients with type 1 diabetes mellitus, Pediatr. Endocrinol. Diabetes Metab. 17 (2011) 71–75. D. Bell, A. Collier, D.M. Matthews, E.J. Cooksey, G.J. McHardy, B.F. Clarke, Are reduced lung volumes in IDDM due to defect in connective tissue?, Diabetes 37 (1988) 829– 831 O. Berriche, F. Ben Mami, S. Mhiri, A. Achour, Is the respiratory function altered during diabetes mellitus?, Tunis Med. 87 (2009) 499–504. M.M. El-Habashy, M.A. Agha, H.A. El-Basuni, Impact of diabetes mellitus and its control on pulmonary functions and cardiopulmonary exercise tests, Egypt. J. Chest Dis. Tuberc. 63 (2014) 471–476. D.A. Hickson, C.M. Burchfiel, J. Liu, M.F. Petrini, K. Harrison, W.B. White, et al, Diabetes, impaired glucose tolerance, and metabolic biomarkers in individuals with normal glucose tolerance are inversely associated with lung function: the jackson heart study, Lung 189 (2011) 311–321. M. Irfan, A. Jabbar, A.S. Haque, S. Awan, S.F. Hussain, Pulmonary functions in patients with diabetes mellitus, Lung India 28 (2011) 89–92. K. Masmoudi, F. Choyakh, N. Zouari, Ventilatory mechanics and alveolo-capillary diffusion in diabetes, Tunis. Med. 80 (2002) 524–530.

[24] S.A. Meo, A.M. Al-Drees, S.F. Shah, M. Arif, K. Al-Rubean, Lung function in type 1 Saudi diabetic patients, Saudi Med. J. 26 (2005) 1728–1733. [25] R.A. Primhak, G. Whincup, J.N. Tsanakas, R.D. Milner, Reduced vital capacity in insulin-dependent diabetes, Diabetes 36 (1987) 324–326. [26] C. Schnack, A. Festa, A. Schwarzmaier-D’Assie, P. Haber, G. Schernthaner, Pulmonary dysfunction in type 1 diabetes in relation to metabolic long-term control and to incipient diabetic nephropathy, Nephron 74 (1996) 395–400. [27] M.P. Villa, M. Montesano, M. Barreto, J. Pagani, M. Stegagno, G. Multari, et al, Diffusing capacity for carbon monoxide in children with type 1 diabetes, Diabetologia 47 (2004) 1931–1935. [28] N. Macintyre, R.O. Crapo, G. Viegi, D.C. Johnson, C.P. van der Grinten, V. Brusasco, et al, Standardisation of the singlebreath determination of carbon monoxide uptake in the lung, Eur. Respir. J. 26 (2005) 720–735. [29] M.R. Miller, J. Hankinson, V. Brusasco, F. Burgos, R. Casaburi, A. Coates, et al, Standardisation of spirometry, Eur. Respir. J. 26 (2005) 319–338. [30] J. Wanger, J.L. Clausen, A. Coates, O.F. Pedersen, V. Brusasco, F. Burgos, et al, Standardisation of the measurement of lung volumes, Eur. Respir. J. 26 (2005) 511–522. [31] K. Suresh, S. Chandrashekara, Sample size estimation and power analysis for clinical research studies, J. Hum. Reprod. Sci. 5 (2012) 7–13. [32] B. van den Borst, H.R. Gosker, M.P. Zeegers, A.M. Schols, Pulmonary function in diabetes: a metaanalysis, Chest 138 (2010) 393–406. [33] D.A. Mahler, C.K. Wells, Evaluation of clinical methods for rating dyspnea, Chest 93 (1988) 580–586. [34] B. Ferris, Epidemiology standardization project ii: recommended respiratory disease questionnaires for use with adults and children in epidemiological research, Am. Rev. Respir. Dis. 118 (1978) 7–52. [35] H. Ben Saad, The narghile and its effects on health. Part i: the narghile, general description and properties, Rev. Pneumol. Clin. 65 (2009) 369–375. [36] R. Bouguerra, H. Alberti, H. Smida, L.B. Salem, C.B. Rayana, J. El Atti, et al, Waist circumference cut-off points for identification of abdominal obesity among the tunisian adult population, Diab. Obes. Metab. 9 (2007) 859–868. [37] G. Harifi, I. Ouilki, I. El Bouchti, M.A. Ouazar, A. Belkhou, R. Younsi, et al, Validity and reliability of the arabic adapted version of the dn4 questionnaire (douleur neuropathique 4 questions) for differential diagnosis of pain syndromes with a neuropathic or somatic component, Pain Pract. 11 (2011) 139– 147. [38] American Diabetes Association, Diagnosis and classification of diabetes mellitus, Diabetes Care 36 (Suppl 1) (2013) S67– S74. [39] E. Villar, M. Lievre, M. Kessler, V. Lemaitre, E. Alamartine, M. Rodier, et al, Anemia normalization in patients with type 2 diabetes and chronic kidney disease: results of the nephrodiab2 randomized trial, J. Diabetes Complications 25 (2011) 237– 243. [40] N. Abramson, B. Melton, Leukocytosis: basics of clinical assessment, Am. Fam. Physician 62 (2000) 2053–2060. [41] M. Okada, H. Matsui, Y. Ito, A. Fujiwara, K. Inano, Lowdensity lipoprotein cholesterol can be chemically measured: a new superior method, J. Lab. Clin. Med. 132 (1998) 195–201. [42] H. Kramer, M.E. Molitch, Screening for kidney disease in adults with diabetes, Diabetes Care 28 (2005) 1813–1816. [43] H. Ben Saad, M.N. El Attar, K. Hadj Mabrouk, A.B. Abdelaziz, A. Abdelghani, M. Bousarssar, et al, The recent multi-ethnic global lung initiative 2012 (GLI) reference values don’t reflect contemporary adult’s north african spirometry, Respir. Med. 2013 (107) (2012) 2000–2008.

Lung Function in Uncontrolled Type 1 Diabetes Mellitus [44] R. Pellegrino, G. Viegi, V. Brusasco, R.O. Crapo, F. Burgos, R. Casaburi, et al, Interpretative strategies for lung function tests, Eur. Respir. J. 26 (2005) 948–968. [45] H. Ben Saad, L. Ben Amor, S. Ben Mdalla, I. Ghannouchi, M. Ben Essghair, R. Sfaxi, et al, The importance of lung volumes in the investigation of heavy smokers, Rev. Mal. Respir. 31 (2014) 29–40. [46] J.E. Cotes, D.J. Chinn, P.H. Quanjer, J. Roca, J.C. Yernault, Standardization of the measurement of transfer factor (diffusing capacity). Work group on standardization of respiratory function tests. European community for coal and steel. Official position of the european respiratory society, Rev. Malad. Respir. 11 (Suppl 3) (1994) 41–52. [47] H. Ben Saad, H. Selmi, K. Hadj Mabrouk, I. Gargouri, A. Nouira, H. Said Latiri, et al, Spirometric ‘‘lung age’’ estimation for north african population, Egypt. J. Chest Dis. Tuberc. 63 (2014) 491–493. [48] D. Mathis, L. Vence, C. Benoist, Beta-cell death during progression to diabetes, Nature 414 (2001) 792–798. [49] Standardization of spirometry-1987 update. Statement of the American thoracic society, Am. Rev. Respir. Dis. 136 (1987) 1285–98. [50] American thoracic society, Single-breath carbon monoxide diffusing capacity (transfer factor). Recommendations for a standard technique–1995 update, Am. J. Respir. Crit. Care Med. 152 (1995) 2185–2198. [51] Le groupe Pulmonaria, P.H. Quanjer, P.L. Enright, J. Stocks, G. Ruppel, M.P. Swanney, et al, Open letter to the members of the gold committee, Rev. Mal. Respir. 27 (2010) 1003–1007. [52] L. Fuso, P. Cotroneo, S. Basso, M. De Rosa, A. Manto, G. Ghirlanda, et al, Postural variations of pulmonary diffusing capacity in insulin-dependent diabetes mellitus, Chest 110 (1996) 1009–1013. [53] M. Halawa, A. Karawagh, A. Zeidan, A. Mahmoud, M. Sakr, A. Hegazy, Prevalence of painful diabetic peripheral neuropathy among patients suffering from diabetes mellitus in Saudi Arabia, Curr. Med. Res. Opin. 26 (2010) 337–343. [54] D.M. Maahs, N.A. West, J.M. Lawrence, E.J. Mayer-Davis, Epidemiology of type 1 diabetes, Endocrinol. Metab. Clin. North Am. 2010 (39) (2010) 481–497. [55] M.R. Becklake, F. Kauffmann, Gender differences in airway behaviour over the human life span, Thorax 54 (1999) 1119– 1138. [56] H.C. Yeh, N.M. Punjabi, N.Y. Wang, J.S. Pankow, B.B. Duncan, C.E. Cox, et al, Cross-sectional and prospective study of lung function in adults with type 2 diabetes: the Atherosclerosis Risk in Communities (ARIC) study, Diabetes Care 31 (2008) 741–746.

727 [57] A.R. Ortiz-Aguirre, M.H. Vargas, A. Torres-Cruz, M. QuijanoTorres, Age related spirometric changes in diabetic patients, Rev. Invest Clin. 58 (2006) 109–118. [58] A.C. Lau, M.K. Lo, G.T. Leung, F.P. Choi, L.Y. Yam, K. Wasserman, Altered exercise gas exchange as related to microalbuminuria in type 2 diabetic patients, Chest 125 (2004) 1292–1298. [59] T.M. McKeever, P.J. Weston, R. Hubbard, A. Fogarty, Lung function and glucose metabolism: an analysis of data from the Third National Health and Nutrition Examination Survey, Am. J. Epidemiol. 161 (2005) 546–556. [60] P. Lange, S. Groth, J. Kastrup, J. Mortensen, M. Appleyard, J. Nyboe, et al, Diabetes mellitus, plasma glucose and lung function in a cross-sectional population study, Eur. Respir. J. 2 (1989) 14–19. [61] E.S. Ford, A.M. Malarcher, W.H. Herman, R.E. Aubert, Diabetes mellitus and cigarette smoking. Findings from the National health interview survey, Diabetes Care 1994 (17) (1989) 688–692. [62] R.I. Dierkx, W. van de Hoek, J.B. Hoekstra, D.W. Erkelens, Smoking and diabetes mellitus, Neth. J. Med. 48 (1996) 150– 162. [63] G.L. Kinney, J.L. Black-Shinn, E.S. Wan, B. Make, E. Regan, S. Lutz, et al, Pulmonary function reduction in diabetes with and without chronic obstructive pulmonary disease, Diabetes Care 37 (2014) 389–395. [64] H. Ben Saad, R. Ben Attia Saafi, S. Rouatbi, S. Ben Mdella, A. Garrouche, A. Zbidi, et al, Which definition to use when defining airflow obstruction?, Rev Mal. Respir. 24 (2007) 323– 330. [65] S. Ben Mdalla, H. Ben Saad, N. Ben Mansour, B. Rouatbi, M. Ben Esseghair, S. Mezghani, et al, The announcement of the lung age it is a motivation to quit smoking?, Tunis Med. 91 (2013) 521–526. [66] D.C. Weir, P.E. Jennings, M.S. Hendy, A.H. Barnett, P.S. Burge, Transfer factor for carbon monoxide in patients with diabetes with and without microangiopathy, Thorax 43 (1988) 725–726. [67] M. Sandler, Is the lung a ‘target organ’ in diabetes mellitus?, Arch Intern. Med. 150 (1990) 1385–1388. [68] V.P. Guntur, R. Dhand, Inhaled insulin: extending the horizons of inhalation therapy, Respir. Care 52 (2007) 911–922.