Recruitment to this project began on 1 February 2006 and finished on 1 February 2008. We reviewed the cases of all women with PGDM who received their pregnancy care and delivered a live baby at our site (Norfolk and Norwich University Hospital NHS Trust), whose offspring would be 16–23 years old (inclusive) at the time of the study. We excluded any women who had a baby born with a major neurological or cardiac malformation. Between 1977 and 2007, all pregnant women with PGDM in our total population of 500,000 have been delivered at this site. Between 1984 and 1991 (excluding 1989 because of missing data), 117 women with PGDM gave birth at our site and 92 met the offspring age criteria. These 92 mothers were invited to participate in the study and asked to inform their 16–23-year-old offspring about the study and invite them to participate.
Of this population, 18 (19.6%) did not respond to recruitment contact or were untraceable, and 13 (14.1%) declined to contact their offspring. In addition, a further 12 mother–offspring pairs (13.0%) were excluded for the following reasons : offspring major cardiac malformation (n = 1), cerebral palsy (n = 1), twin pregnancy (n = 1), classical type 1 diabetes diagnosed in childhood (n = 3), Addison’s disease and type 1 diabetes (n = 1), severe rheumatoid arthritis (n = 1), current offspring pregnancy (n = 2), missed appointments or outside age range (n = 2). Of the remaining 49 mother–offspring pairs, 28 (57.1%) of the offspring aged 16 to 23 years declined to participate, leaving a study population of 21 offspring of PGDM, or 22.8% of the total possible. Controls were selected by media advertisement inviting individuals who were born at the Norfolk and Norwich University Hospital NHS Trust between 1984 and 1991 whose mother did not have diabetes at the time of conception or during pregnancy. No selection criteria other than age and location of birth were applied to the recruitment of the control group; no attempt was made to weight-match controls (n = 23). All participants, except one control, were white. All individuals participated after providing written informed consent, and after review and approval of the study by the local research ethics committee.
All participants underwent a standard early morning OGTT, using 75 g of anhydrous glucose, after 10 h of fasting. Peripheral venous samples were collected from individuals in a recumbent position at −10, −5, 0, 60 and 120 min following the oral consumption of glucose. All plasma analyses were undertaken on fasting blood samples, other than plasma insulin, which was measured at time 0 and 120 min of the OGTT. Anthropometric measurements were made after the OGTT. Participants were also asked if they were prepared to undergo a transdermal skin biopsy for skin fibroblast studies.
Anthropometric and clinical measurements
We recorded height (cm; without shoes), weight (kg; unclothed), waist circumference (cm), waist:hip ratio, BMI (kg/m2), skinfold thickness (in mm; triceps, abdominal and subscapular). Body fat mass and fat-free mass were measured by bioimpedance (segmental Body Composition Analyser BC-418MA Tanita UK, West Drayton, UK). Peripheral vascular disease was defined by determination of the ankle–brachial pressure index using Doppler pressure measurement . The blood pressure of the participants was measured by an Accoson sphygmomanometer (BS 2744; Accoson Works, Harlow, UK) after they had rested for 5 min, with three readings taken at 1 min intervals and the mean of the last two readings recorded. At the end of the visit each participant was given a food diary to complete and return. Participants were asked to record a 3 day dietary history over 1 week, including two weekdays and one weekend day. The diaries were analysed by the Department of Nutrition and Dietetics (Norfolk and Norwich University Hospital NHS Trust) using CompEat software . Each adult participant was assigned a social deprivation score based on current address, derived by the UK government for all small areas in England from a range of economic, social and housing variables . In addition, for each adult volunteer, an estimate of relative birthweight ratio, derived from actual recorded birthweight and normative data for gestational age- and sex-related 50th percentile for birthweight was calculated .
Insulin, glucose and the homeostatic assessment model of insulin resistance
Fasting and 2 hour plasma insulin levels were measured using a commercially available human insulin specific ELISA (Bio-stat Diagnostic Healthcare, Stockport, UK). Plasma glucose was analysed by the glucose oxidase method (Abbott Laboratories, Maidenhead, UK). Homeostasis model assessment of insulin resistance (HOMA-IR) was used to estimate peripheral insulin sensitivity . Fasting glucose and insulin values are shown as the mean of samples collected at −10, −5 and 0 min.
Mouse anti-human CD3:AF647, CD19:AF647, CD14:AF647 and CD45RA:PE antibodies were purchased from AbD Serotec (Oxford, UK), and mouse anti-human CD3:PE, CD19:PE-Cy7, CD14:PE-TR and CD45RA:PE-Cy5 antibodies were purchased from Caltag Medsystems (Buckingham, UK).
Isolation of peripheral blood mononuclear cells
Peripheral blood mononuclear cells (PBMCs) were isolated using BD Citrate Vacutainer CPT tubes (Southern Syringe Services, Leicester, UK) according to the manufacturer’s instructions. Briefly, the cells were centrifuged at 1,500 g for 20 min, washed twice in PBS and then stored in liquid nitrogen until analysis. On average, the total number of viable cells was 95% when counted under a microscope with a haemocytometer using trypan blue staining.
Isolation and culture of skin fibroblasts
Human skin fibroblasts were derived from skin biopsy samples taken from the inside upper arm of 15 individuals, as previously described . The biopsy samples were transferred to cell culture dishes (60 mm diameter) and incubated at 37°C, in an atmosphere of 5% CO2 and 95% air, in Eagle’s minimum essential medium (MEM; Sigma-Aldrich, Gillingham, UK) supplemented with 10% (vol./vol.) fetal bovine serum to allow fibroblast outgrowth. After two passages the cells were grown to confluence, washed and either used in the senescence assay or stored in liquid nitrogen until needed for telomere analysis.
Telomere length analysis
Telomere length in individual PBMCs was measured using a Telomere Peptide Nucleic Acid (PNA) kit/fluorescein isothiocyanate (FITC) for flow cytometry (Bio-stat Diagnostic Healthcare, Stockport, UK). Telomere length was measured in different leucocyte subsets as previously described , with minor modifications. Briefly, 5 × 105 cells were incubated for 30 min at 4°C with either CD3:AF647, CD19:AF647 or CD14:AF647. The labelled cells were resuspended in either hybridisation solution containing the telomere probe or hybridisation solution alone as a negative control. The cells were then incubated for 10 min at 80°C in a water bath, followed by overnight hybridisation in the dark at room temperature. The cells were incubated twice for 10 min at 40°C with wash buffer, centrifuged and then resuspended in 50 μl of PBS. The cells labelled with CD3:AF647 and CD19:AF647 were further incubated with CD45RA:PE for 30 min at 4°C, washed once and then resuspended in DNA staining solution to identify cells in G0/G1 of the cell cycle. After 2–3 h 20,000 events were acquired on a Cytomics FC500MPL flow cytometer (Beckman Coulter, High Wycombe, UK) and CXP software (Beckman Coulter) was used for analysis. The different leucocyte subsets were identified by their forward and side scatter properties and staining with the appropriate antibody. Naive T cells were identified as CD3+/CD45RA+, memory T cells as CD3+/CD45RA−, B cells as CD19+ and monocytes as CD14+. The mean telomere fluorescence of each cell type was calculated by subtracting the mean background fluorescence of unlabelled control cells from the mean fluorescence of cells hybridised in the presence of the FITC-PNA probe. Telomere fluorescence was converted into molecular equivalents of soluble fluorochrome (MESF) using Quantum FITC premixed low-level MESF beads (Bangs Laboratories, Fishers, IN, USA), which were run in each experiment. Telomere length in cultured fibroblasts was measured using the Telomere PNA kit/FITC, as described above. In addition, one tube of cells per sample was incubated with mouse anti-human CD90/Thy 1 (FITC), a fibroblast marker, for 30 min at 4°C, fixed with 1% (vol./vol.) paraformaldehyde and stored at 4°C until analysed by flow cytometry.
Oxidative DNA damage
An OxyDNA kit (Biotrin, Dublin, Republic of Ireland) was used to evaluate oxidative DNA damage in different leucocyte subsets as previously described , with some modifications. Briefly, 5 × 105 cells were washed in PBS and incubated for 15 min at room temperature with the antibodies CD3:PE, CD14:PE-TR, CD19:PE-Cy7 and CD45RA:PE-Cy5 (Caltag Medsystems, Buckingham, UK) to identify the different cell types. After washing, the cells were fixed and permeabilised with a Fix and Perm cell permeabilisation kit (Invitrogen, Paisley, UK), followed by a further wash. The cells were then incubated for 1 h at 37°C with 50 μl of Biotrin blocking solution, washed and incubated for 1 h in the dark at room temperature with 100 μl of FITC-conjugate probe specific for 8-oxoguanine. Following incubation, cells were washed twice, analysed by flow cytometry with the acquisition of 20,000 events and the mean fluorescence intensity recorded. A control sample was run in each experiment to account for inter-assay variation.
Fibroblast senescence assay
Fibroblast senescence was measured by senescence-associated β-galactosidase activity using the fluorogenic substrate 5-dodecanoylamino fluorescein di-β-d-galactopyranoside (C12FDG; Invitrogen, Paisley, UK), as previously described  with some minor modifications. Briefly, cells were incubated with either C12FDG (33 μmol/l), or DMSO as a control, for 2.5 hours. Following a wash with PBS the cells were trypsinised, analysed by flow cytometry; 10,000 events were acquired and the mean fluorescent intensity recorded.
HbA1c, lipid and lipoprotein profiles
HbA1c was measured by HPLC on a Biomen HA-8160 analyser (Menarini Diagnostics, Wokingham, UK). Plasma lipoprotein analysis was performed by Lipomed Biosciences (Liposcience, Raleigh, NC, USA) using nuclear magnetic resonance (NMR) spectroscopy .
Commercially available ELISA kits were used to analyse plasma levels of TNFα, IL-6, soluble cell adhesion molecules (soluble intercellular adhesion molecule 1 [sICAM-1] and soluble vascular cell adhesion molecule 1 [sVCAM-1]) (all R&D Systems Europe, Abingdon, UK) high-sensitivity C-reactive protein (hs-CRP) (Kalon Biological, Guildford, UK).
Data are expressed as mean (SD) or as median (interquartile range) as appropriate. All data were normally distributed, except plasma IL-6 and hs-CRP concentrations. Differences between groups were analysed by the unpaired Student’s t test or Mann–Whitney U test as appropriate. A p value of <0.05 was considered significant. In our recent studies , monocyte telomere length was 5. 74 (0.88) MESF in controls, which suggested an initial sample size of 30 in each group was necessary to provide 89% power at the 5% level to detect a 0.75 SD difference between groups. In the present study, based on actual mean and SD data for telomere length in PBMCs in both groups (Table 4), the study had 90.1% power at the 5% level to detect a 0.75 SD in mean telomere length compared with the control group without a maternal history of diabetes.