Improvements in the incidence and survival of cancer and 1 cardiovascular but not infectious disease have driven recent 2 mortality improvements in Scotland: nationwide cohort study 3 of linked hospital admission and death records 2001–2016

Objectives To identify the causes and future trends underpinning improvements in life quantify relative Main outcome measures Hospital admission rates and survival in the five following admission for 28 diseases, stratified by sex and socioeconomic status. Results The five hospital admission diagnoses associated with the greatest burden of death 28 subsequent to admission were “Influenza and pneumonia”, “Symptoms and signs involving 29 the circulatory and respiratory systems”, “Malignant neoplasm of respiratory and 30 intrathoracic organs”, “Symptoms and signs involving the digestive system and abdomen”, 31 and “General symptoms and signs”. Using disease trends, we modelled a mean mortality 32 hazard ratio of 0.737 (95% CI 0.730–0.745) across decades of birth, equivalent to a life 33 extension of ~3 years per decade. This improvement was 61% (30%–93%) accounted for by 34 improvements in disease survival after hospitalisation (principally cancer) with the remainder 35 accounted for by a fall in hospitalisation incidence (principally heart disease and cancer). In 36 contrast, deteriorations in the incidence and survival of infectious diseases reduced mortality 37 improvements by 9% (~3.3 months per decade). Overall, health-driven mortality 38 improvements were slightly greater for men than women (due to greater falls in disease 39 incidence), and generally similar across socioeconomic deciles. We project mortality 40 improvements will continue over the next decade but will slow down by 21% because much 41 of the progress in disease survival has already been achieved. making


Summary box
What is already known on this topic 54 55 • Long term improvements in Scottish mortality have slowed down recently, while life 56 expectancy inequalities between socioeconomic classes are increasing. 57 58 • Deaths attributed to ischaemic heart disease and stroke in Scotland have declined in 59 the last two decades. 60 What this study adds 61 62 • Gains in life expectancy can largely be attributed to improvements in cancer survival 63 and falls in incidence of cancer and cardiovascular disease. 64 65 • The hospitalisation rate and survival of several infectious diseases have deteriorated, 66 and for urinary infections, this decline has been more rapid in more 67 socioeconomically deprived classes. 68 69 • Improvements in morbidity are projected to slow down, with much progress in 70 survival of heart disease and cancer already achieved, and align with the recently 71 observed slow-down in mortality improvements. 72

73
In recent decades, there has been a substantial improvement in life expectancies at birth in 74 the UK(1). More recently, several studies have suggested that there has been slowdown in 75 improvements in the USA, UK, France, Germany, Sweden, the Netherlands and other 76 Organisation for Economic Co-operation and Development countries; however, the causes 77 are less clear, with speculation that they may arise from slowing improvements in 78 cardiovascular disease, increased influenza mortality and/or pressure on health and social 79 care services(1-8). Understanding trends in disease incidence and subsequent survival 80 could illuminate such trends in mortality, and disentangling how and how much different 81 diseases contribute has the potential to reveal whether investment in healthcare and 82 research is directed at the most urgent diseases and most affected individuals. 83 84 Through its electronic Data Research and Innovation Service (eDRIS), Scotland has linkage 85 of historical individual death and electronic health records in a controlled environment, with 86 specific study approvals by the Public Benefit and Privacy Panel. This allows direct 87 modelling at an individual level of the incidence of disease and subsequent death or survival 88 of subjects. Furthermore, because historic records are available and the whole population is 89 covered, a retrospective cohort study can be constructed (with inherent representativeness 90 of the initial sample, with very complete levels of follow-up, and without survivor bias). 91 92 Here, we use population-wide data between 2001 and 2016 on residents of Scotland born 93 before 1966 to explore how trends in longevity were driven by different trends in broad 94 classes of disease incidence or survival, and highlight diseases which have shown more or 95 less improvement in their contribution to overall mortality. We partition overall mortality by 96 sex and socioeconomic status and, assuming past disease improvements continue to the 97 same extent in the future, use these results to project future improvements in mortality and 98 their changing sources. 99 Design 142 Mortality trends were modelled using morbidity trends: we first determined the major disease 143 categories (ICD10 blocks) associated with the most lives lost by taking into account the 144 frequency of the disease (as measured by hospitalisation) and its effect on survival (as 145 measured by the subsequent all-cause mortality of patients admitted for the disease 146 compared to the mortality of everyone else). We combined these two measures to calculate 147 a burden of death weighting for each block. We then looked at how the age-adjusted trends 148 in hospitalisation rates (as a proxy for incidence) changed for each disease, by decade of 149 birth, projecting that if incidence of a disease fell by a given percentage, its contribution to 150 mortality would fall similarly. We then calculated a weighted average change in 151 hospitalisation rates, reflecting the expected effect of all measured disease incidence 152 changes on mortality rates, driven by decade of birth. Similarly, we looked at how the (age-153 adjusted) 5-year survival rates following first hospitalisation changed by year of 154 hospitalisation. For each block this again gives a contribution towards reduced mortality, and 155 the weighted average, the expected effect of changes in survival of the combined diseases 156 on overall mortality. Adding these effects (and noting we assessed changes in survival from 157 incidences over one decade), gives the expected effect on overall mortality from decade of 158 birth to subsequent decade of birth from the effect of changes in disease incidence and 159 survival, under the (necessarily simplified) model that incidence is a function of birth cohort 160 and survival post incidence is a function of year of incidence. 161

Statistical analysis 162
Mortality 163 A Cox proportional hazards model using NRS mortality data -fitting sex, decade of birth, 164 and deprivation -was used to quantify mortality in the Scottish population during the study 165 period. The same analysis was run stratified by sex and deprivation. Unless otherwise 166 stated, (for example median age differences in Kaplan-Meier curves), years of life of a 167 hazard effect have been calculated by multiplying the lnHR by 10 (ref. 13). Only individuals 168 with complete records were included in the analysis. 169 Morbidity 170 We grouped the main diagnoses of each NSS hospital admission into categories, as laid out 171 by the ICD10 Chapters, and included only the first instance of admission for a category per 172 individual (discarding subsequent repeat visits to hospital for the same disease). Analysis 173 was restricted to more common disease blocks. Visual inspection suggested a pragmatic 174 threshold of at least 15,000 first-time admissions (see Table 2 Where h 0 is the baseline hazard, x the patient age, and X 1 -X 4 the covariates sex, deprivation, 193 and Where N firstadmission is the total number of first hospital admissions of the disease during the 204 study period, N total is the total number of individuals in the study, and h (0,5) is the mortality of 205 individuals in the first five years following hospitalisation compared to individuals who were 206 never hospitalised for the disease, measured in lnHR. The resulting value was then scaled to 207 [0-1] and provides a relative measure of the number of lives lost due to the disease, with 208 higher values indicating a disease which is more common or associated with higher 209 subsequent mortality, and lower values indicating a disease which is rare or associated with 210 lower subsequent mortality. Whilst this measure may in principle be affected by differing age 211 patterns on incidence, it was judged sufficient for our purpose -to establish broad relative 212 weightings of the importance of each disease. 213 214 To maintain a feasible computational burden within the national safe haven, subsequent 215 analysis was restricted to the 25 blocks with the highest burden of death on the population 216 (Table 2). We added C50-C50 malignant neoplasm of the breast, C60-C63, malignant 217 neoplasms of male genital organs, and G30-G32 other diseases of the nervous system to 218 this list, out of specific interest: in the sex-specific effects and awareness of the limitations of 219 our method for Alzheimer's disease (see discussion). All further analyses were performed on 220 these top 28 blocks (T-28). The use of (first) hospitalisation for a disease as our definition of 221 incidence is imperfect (e.g. for Alzheimer's disease where hospitalisation following incidence 222 is rare or delayed, and even first diagnosis in the community will often be preceded by a long 223 latent period)(15

281
Multiplying total number of hospitalisations during the study period (as a proxy for disease 282 prevalence) by 5-year mortality after hospital admission (as a proxy for disease severity) 283 provided a weight for the death burden of hospitalisation of each ICD10 block. We restricted 284 our analyses to 28 of the top disease blocks for burden of death (T-28, see methods). 285 Among the T-28, total cases of disease incidence (i.e. first-time admissions) during the study 286 period ranged from 33,613 (A30-A49, Other bacterial disease) to 225,504 (R00-R09, 287 Symptoms and signs involving the circulatory and respiratory systems) (

315
Apart from sex-specific cancers, we observe significant differences in burden of death 316 between men and women for injuries to the hip and thigh (S70-S79) and head (S00-S09), 317 with the former having a higher burden in women due to more female cases and the latter 318 having a higher burden in men due to more male cases. For both disease blocks, the effect 319 of hospitalisation on subsequent mortality is greater in men than women (S70-S79 HR men: 320 3.19, women 2.44; S00-S09 HR men: 2.32, women 1.88 HR). Strikingly, 5-year mortality 321 after hospital admission for IHD is higher for women (HR 2.01/1.70 women/men), but this is 322 offset by the lower prevalence of hospitalisation in women (Table 2-Source Data 1). 323 324 To understand changes in disease survival rates, we next modelled the effects of a disease 325 on all-cause mortality by year of hospital admission for admissions between 2001 and 2011 326 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019. ; https://doi.org/10.1101/19003012 doi: medRxiv preprint 1 and 5-year survival subsequent to admission. We find an overall improvement over time in 327 patient survival following hospitalisation, with a median decline between 2001 and 2011 in 328 the 5-year HR of 16.8% for admitted cases across the T-28 diseases. The biggest 329

388
The shape of disease modelled mortality improvements by decade of birth broadly tracks the 389 observed changes. Furthermore, when stratified by sex a similar effect is seen: apparent 390 distinctions in projected morbidity driven mortality are associated with distinctions with 391 observed mortality (correlation 0.89); i.e. mortality trends in the study can largely be 392 explained by trends in disease incidence and survival (S Table 1; S Figure 6). Across sex 393 and deprivation strata, taking into account disease survival improvements between 2001 and 394 2011 and all improvements in disease incidence between decades of birth, we find the 395 largest reductions in death are due to improvements in ischaemic heart diseases (I20-I25), 396 malignant neoplasms of digestive organs (C15-C26), and malignant neoplasm of respiratory 397 and intrathoracic organs (C30-C39), while the largest increases in death are due to other 398 bacterial diseases (A30-A49) and influenza and pneumonia (J09-J18). In addition, the 399 deterioration in other diseases of the urinary system (N30-N39) morbidity shows a consistent 400 increase with deprivation, while other diseases of the digestive system (K90-K93) shows 401 consistently larger improvements in more deprived classes (  1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955  Decade of birth commencing Improvements in morbidity . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019. ; https://doi.org/10.1101/19003012 doi: medRxiv preprint improvement attributable to survival rather than incidence), C50 (85%) and C76-80 (68%), 408 whereas for cardiovascular diseases, the balance was more even I80-89 (52%), I60-I69 409 (46%), I30-I32 (43%), I20-I25 (38%). 410 411 As previously noted, disease severity was defined as the log hazard ratio for subsequent all-412 cause mortality among those with a previous admission for an index group of conditions 413 compared with those with no such admission. We regarded the rate of improvement in 414 disease severity over time as being constant if there was the same relative fall in log hazard 415 rate over successive time periods (so for example we regarded a fall in lnHR from 0.6 to 0.3 416 as equivalent to a fall from 0.3 to 0.15). Assuming the improvements in survival following 417 hospitalisation continue for the coming decade, and differences between incidence in birth 418 cohorts remains the same, we project a 21% slowing of improvements in mortality (-0.242 419 lnHR c.f. -0.305 lnHR;

435
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) pneumonia' (more common and with higher subsequent mortality than the average T-28 446 disease), 'Symptoms and signs involving the circulatory and respiratory systems' (common), 447 'Malignant neoplasm of respiratory and intrathoracic organs' (higher mortality) ,'Symptoms 448 and signs involving the digestive system and abdomen' (common), and 'General symptoms 449 and signs' (common and higher mortality). Whilst the latter might appear a benign diagnosis, 450 our results suggest it is a fairly strong and frequent marker of subsequent all-cause mortality. 451 452 Across decades of birth, we modelled a reduction in mortality hazard of 0.737 (95% CI 453 0.730-0.745) due to improvements in morbidity, which broadly tracked improvements in 454 observed mortality. The modelled improvement was 61% accounted for by reduction in 455 excess mortality subsequent to admission and 39% accounted for by a fall in incidences of 456 disease (as measured by hospital admission rates). The important (i.e. burden-of-death 457 weighted) improvements in incidence were driven by cancers and heart disease, whilst 458 improvement in outcomes following admission were mostly driven by cancer, particularly 459 breast and prostate cancer. In contrast, we found deteriorations in the incidence of bacterial 460 disease and in mortality following admission for respiratory and urinary infections. Levels of 461 morbidity and mortality varied strongly across socioeconomic groups, but patterns in 462 changes of such were generally less apparent. Men showed greater rates of improvement in 463 mortality and morbidity than women, with lung and throat cancers contributing most to male 464 improvements and IHD contributing most to female improvements. 465 466 In conclusion, we find trends in morbidity appear to partly explain trends in mortality. The 467 progress in prevention and cure within oncology and prevention of heart disease account for 468 the greatest parts of mortality improvement in 2001-2016, and our model suggests mortality 469 improvements may slow, simply because the absolute effect of progress in treatment of 470 these diseases will be difficult to repeat. However, there is scope for further improvements in 471 life expectancy, especially if new progress is made in the treatment of other diseases 472 associated with death, or if prevention initiatives accelerate. 473 474 This study has avoided some of the known issues with cause of death recording(16) since it 475

Strengths and weaknesses of the study
does not use cause-specific mortality and tracks wider disease effects and subsequent 476 mortality (such as frailty) beyond direct causes of death, by combining hospitalisation and 477 death records. Implicit tracking of underlying causes through an associated effect (admission 478 to hospital for a disease) may improve estimates of trends in mortality, even if the underlying 479 cause is obscure. We are also able to partition trends in deaths due to a disease based on 480 trends in prevalence and incidence, which has been done for IHD(17), but not 481 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019. capture chronic conditions which are managed in general practice or the community but 502 eventually lead to death, such as dementia and multiple sclerosis. Also, association between 503 hospital admission for a condition and the latter occurrence of death does not necessarily 504 imply causation, and may be confounded by other health risks, such as socioeconomic 505 status, a second disease, or behavioural factors, although our analysis stratified by 506 socioeconomic status may partly mitigate these effects. Modelling comorbidities 507 simultaneously or modelling the causal effects of one morbidity on another and consequently 508 on death has been beyond the scope of this study. In addition, our model assumed that 509 (age-adjusted) incidence is a function of year of birth, but that mortality post-incidence is a 510 function of the year of incidence. This is clearly an oversimplification, but necessary given 511 that year of birth and year of incidence are completely confounded for a given age at 512 incidence. 513 514 By relying on routine data, we have been able to create a very large dataset, at the expense 515 of being unable to externally check individual fields, such as primary diagnosis, where there 516 are known to be inaccuracies(19). However, our conclusions will be less affected if these 517 inaccuracies are reasonably stable over time. Furthermore, we have used a simple 518 definition of burden of disease: the total number of cases multiplied by the excess hazard 519 across all ages. More complex definitions might have looked at Years of Life Lost (YLL) and 520 taken into account any age-related pattern in incidence. However, our measure is simple 521 and is predominantly used to weight the diseases and especially their improvements to 522 calculate an overall improvement in all-disease impact, rather than as an absolute measure 523 of burden. 524 525 Lastly, our model assumed disease incidence and survival hazards were proportionate. 526 Incidence of hospital admission and death used "age" as the baseline hazard determinant, 527 while mortality following admission used "time since admission". In practice we do not expect 528 the risks analysed to be truly proportionate, nonetheless we consider the analysis of trends 529 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019. ; https://doi.org/10.1101/19003012 doi: medRxiv preprint to be revealing. Conversely, recently doubt was cast on the effect of using decade of birth 530 groupings in a standardised mortality analysis(20) due to changing patterns of births within 531 the decade: that concern does not apply here as the Cox baseline hazard used full (not 532 rounded) age information, with decade of birth fitted as a covariate. 533 534 Changes in hospital admission policy have likely affected changes in observed 535 hospitalisation rate. Conversely, survival rates could be influenced if for example, only more 536 serious cases were admitted. One example from our study is the observed increase in 537 incidence of influenza and pneumonia, which was offset by increased survival and may 538 therefore reflect changes in coding practices and/or frequency of referral(21). Some degree 539 of caution is needed in observing the changes in morbidity, although such biases will have 540 tended to offset each other in terms of the overall change in the mortality arising from 541 morbidity for such a disease block. However, clearly the partition of effect between incidence 542 and survival can be affected. 543 544 There was a degree of correspondence in the principal burdens assessed here and a recent 545 study by the Scottish Burden of Disease study (SBD)(22). This study used the same 546 population and the same study period but assessed YLL (weighting young deaths more as 547 opposed to our method which counted all deaths equally), included individuals younger than 548 35 years old, and used different disease groupings. Their principal burdens were IHD 549 (ranked 13 th in our list of burdens), tracheal, bronchus and lung cancers (3 rd ), chronic 550 obstructive pulmonary disease (12 th ), stroke (10 th ), and Alzheimer's disease (-). Aside from 551

Strengths and weaknesses in relation to other studies
Alzheimer's disease, discussed below, much of the distinction appears to arise from our 552 observation of an association between death and admissions with indistinct diagnosis (not 553 considered a valid specific cause of death by SBD). In the case of influenza and pneumonia, 554 differences arise due to our study identifying a marker of frailty as well as a direct cause of 555 death, combined with SBD grouping influenza and pneumonia under lower respiratory 556 infections. A relative strength of our study stems from usage of incident morbidity (as marked 557 by hospitalisation) in advance of death, based on recorded diagnosis at the time of hospital 558 visit, thus tracking remote effects such as long term frailty rather than cause of death (which 559 has known limited accuracy, particularly at older ages(16)). However, the principal strength 560 arises from the ability to distinguish trends in incidence of morbidity from trends in 561 subsequent survival. On the other hand, a relative weakness is that we are reliant on 562 hospital admission as a marker of incidence; therefore, diagnosed or latent (presumably 563 milder) cases in the absence of admission are not visible to us, leading for example to 564 significant discrepancy with SBD in the apparent relative burden of Alzheimer's disease, 565 likely due to an understatement of its importance in our results. 566 567 The closing gap in mortality between the sexes and its widening across social classes 568 observed in our study is consistent with recent findings from the Office of National Statistics, 569 summarised by Torjesen(23), which looked at socioeconomic deprivation in England and 570 Wales. Similarly, a recent study of health inequality in England found rising levels of lifespan 571 inequality across socioeconomic groupings arising from increasing inequalities across a 572 broad span of causes of death(24). These studies had the advantage of a larger sample size 573 (~7.5 million deaths cf. 600,000 in our study) and could therefore track trends in mortality 574 and cause of death between stratified groups more accurately. However, Scotland's unique 575 linkage of death records and electronic health records through eDRIS allowed us to directly 576 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019. ; https://doi.org/10.1101/19003012 doi: medRxiv preprint model changes disease mortality at an individual level (avoiding issues with cause of death 577 recordings and shifts in population demographics). Our study has the advantage of partly 578 explaining these trends in mortality inequality through changes in disease incidence and 579 survival: men experienced greater improvements in incidence of lung cancer and survival 580 following heart disease hospitalisation compared to women, while more socially deprived 581 individuals (men and women combined) suffered worse deteriorations in infectious disease, 582 especially for the incidence and survival of hospitalisation for urinary tract infections. 583 However, in contrast to Bennet et al. (24), we do not find a clear pattern in overall morbidity 584 improvements across socioeconomic deciles in Scotland, and we do not observe a widening 585 inequality in cancer, respiratory and Alzheimer's disease morbidity within our study 586 population, although we are underpowered to detect the latter and our disease groupings 587 were not identical. 588 589 Lastly, a recent study of coronary heart disease mortality in Scotland, using a sophisticated 590 model to apportion improvement between prevention and treatment, found improvements for 591 coronary heart disease between 2001-2010 were similar across social classes, and reported 592 33%-61% of these improvements could be attributed to advances in treatment(17). Given 593 the very different methods, albeit studying the same population, there is reasonable 594 concordance with our own study: we find roughly equal improvements in heart disease 595 across social classes and estimate 38% (95% CI 28%-48%) of these improvements stem 596 from increased survival after hospitalisation for ischaemic heart disease. Hotchkiss et al. (17)  597 are able to further partition improvements by uptake of primary and secondary prevention 598 drugs and treatments. Such detailed analysis of specific diseases has been beyond the 599 scope of our study. 600

601
Much of the improvements in mortality observed in Scotland between 2001-2016 can be 602 attributed to reductions in morbidity, as captured by hospital admissions. While this study 603 examined mortality and morbidity in the Scottish population only, there is a substantial 604 concordance in mortality trends across high-income countries(7), as well as similarities in 605 disease-related mortality trends between Scotland and the rest of the UK(6), warranting 606 similar studies to be performed in other high-income countries. It is a testament to healthcare 607 services that the majority of mortality improvements appear to stem from advances in 608 disease survival post-admission. Observed improvements in cancer incidence and survival -609 especially breast and prostate -coincide with a continued effort within Scotland(25), the 610 UK(26), and other high-income nations(27) to improve prevention and care of these 611 diseases. However, the rapid advances in survival of both heart disease and cancer 612 modelled by our study between 2001 and 2011 will be hard to continue to the same extent, 613 as so much progress has already been made. At the same time, the observed deteriorations 614 in infectious disease coincide with global increases in antimicrobial resistance(28) and 615 emphasise the need to prioritise research in this area: infectious disease will become a 616 larger contributor to mortality and may contribute to a widening of health inequalities 617 between socioeconomic classes. If these current trends in morbidity continue, we expect 618 morbidity-driven improvements in mortality to slow down. However, the life expectancy gap 619 between Scotland and other high-income countries(29) suggests further mortality 620 improvements are possible. The rate of this improvement will hinge upon whether advances 621 in all major diseases categories -especially infectious disease -can catch up with the 622 progress we have recently seen on heart disease and cancer, and whether preventable 623 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019. ; https://doi.org/10.1101/19003012 doi: medRxiv preprint deaths from external causes (such as suicide and drug-related deaths), which cannot be 624 accurately tracked using hospital admissions, decrease rather than rise. 625

626
The study was funded by the Lloyds Banking Group, for the creation, curation and 627 dissemination of knowledge in the public interest, in particular to improve estimates of future 628 population size and morbidity and mortality rates to facilitate healthcare and other 629 government planning. All analyses stratified by socioeconomic deprivation (Table 2, Table 3,  630  Supplementary Table 1, and Supplementary Data Files 1-5) were confidentially re-analysed 631 using 10 socioeconomic groups specified by Lloyds, rather than the Scottish Index of 632 Multiple Deprivation. No Lloyds employees were granted access to individual patient data. 633 Data access was granted to University of Edinburgh researchers only and only through the The study and its design were conceived by the last author. The funder reviewed the design 638 and said that adding stratification by socioeconomic status was a key requirement for 639 meaningful analysis and their funding was conditional upon this, a request to which the 640 authors readily agreed. The funder was kept informed of interim analyses and reviewed the 641 draft manuscript, occasionally suggesting additional analyses or requesting clarifications. 642 The funders were not permitted to and did not request the removal of any results. 643 644 Authors affiliated with eDRIS and NHS Scotland were compensated by PKJ using the grant 645 from Lloyds Banking Group, in line with normal pricing for the work undertaken. PRHJT

801
Contains sheets for 1) total incident cases, 2) mortality in the first 5 years following disease

875
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. not certified by peer review) (which was The copyright holder for this preprint this version posted August 5, 2019.  q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q q 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955 1915 1925 1935 1945 1955  R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 R 2 = 0.6405 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04 P < 3e−04