Upsala J Med Sci 91: 89-98, 1986 Studies of Drug-treated Diabetes in the County of Jamtland, Sweden, Based on Prescription of Insulin and Oral Antidiabetic Drugs Olov W;linder, Ingmar Bergstrom, Goran Boethius and Lars Ugander Department of Internal Medicine and Pulmonary Medicine, Ostersund Hospital, Ostersund, Sweden ABSRACT Epidemiological data on drug-treated diabetes mellitus was obtained by ana- 1.yzing prescriptions of insulin and oral antidiabetics in the county of Jamtland Sweden. During a 12-year period, from 1971 to 1982, the preval.ence of diabetes increased from 1.5% to 2.276, while there was no significant change in the an- nual incidence of around 2.7/1000 inhabitants. Women displayed a considerable higher prevalence and incidence than men in the over 70 age-group, while in the50 - 69 age-group there was a male domi- nance. A significant change in the prescription pattern with an increased use of insulin and decreased use of oral antidiabetics was observed during the periad. More patients were started on oral antidiabetics during the first part o f the year than during the latter. INTRODUCTION By recording drug prescription data valuable information can be obtained concerning the morbidity of a certain disease, provided the drugs are prescri- bed on one well defined indication (2). This is the case for antidiabetic drugs and therefore prescription of insulin and oral antidiabetics should reflect the morbidity of drug-treated diabetes. In the county of Jamtland, Sweden, an individual-hased drug prescription project has Seen in operation since 1970. At the start of this project about 2 % of the county inhahitants were prescribed antidiabetic drugs ( 3 , 4 ) . H e r e we report data on prescription of antidiabetic drugs and diabetes morbidity during the following 12-year period from 1971 to 1982. METHODS Jamtland has a relatively stable population of around 135000 inhabitants. The age distribution differs from that of Sweden as a whole, in so far that the 89 county has relatively more people in the older age groups. In 1 9 8 2 , 1 3 . 3 5 of the population was over 70 years of age compared to 11.52 in the whole of Sweden. Since 1970 all drug prescriptions to inhabitants of the county born on four certain days of the month have been recorded. Thirteen per cent of the in- habitants of the county, i.e. about 17000 inhabitants, have thus been conti- nuously monitored. Drugs dispensed to occasional visitors to the county have not been registered. The recording concerns, inter alia, the patient's identity number, the year and week of dispensation, the prescribing physician, total amount of the drug and the dosage. A detailed presentation of the method was given by Boethius and Wiman (5). Prescriptions of insulin and oral antidiabetics were analyzed retrospecti- vely for the 12-year period, from 1971 to 1982. The prevalence of drug-treated diabetes was obtained by counting the number of individuals diabetic drugs during one particular year. The incidence was calculated from the number of individuals who obtained their first prescription of antidiabetic drugs during that year. Patients who changed to insulin within six months after the start of oral treatment have been regarded as inuslin-treated. A few pat- ients, who started on insulin but changed to tablets within six months, were considered tablet-treated. prescribed anti- The population at risk was the average population of the county at the end of each year according to vital statistics. The figures thus obtained somewhat underestimate the real prevalence and incidence of drug-treated diabetes since the drugs used by diabetics at geriatric clinics are not recorded. Defined daily dose (DDD) is the average dose for a particular drug (2). Recommended doses for insulin are 40 IE and oral antidiabetics, e.g. glibencla- mide 5 mg. Data concerning sale figures were obtained from the National Cor- poration o f Swedish Pharmacies. RESULTS Prevalence and incidence The prevalence of drug-treated diabetes increased gradually from 1.55 in 1971 to 2.2% in 1982 (Table 1 ) . Women displayed a higher prevalence than men throughout the whole period. After correcting the figures according to sex and age of the total Swedish population, the prevalence for 1982 was 1 . 9 % for men and 2 . 2 % for women. 90 Table 1. Prevalence of drug-treated diabetes in the county of Jimtland. Men % Women L Total % 1971 1.3 1.6 1 . 5 -73 1.6 1 . 7 1.6 -75 1.7 1.9 1.8 -77 1.8 2.0 1.9 -79 1.9 2 . 2 2 . 1 -81 1.9 2 . 4 2.1 -82 2.1 2 . 3 2 . 2 During the observation period, the prevalence was fairly constant in the age groups below 60 years, but over this age there was a gradual increase (Fig 1). In the youngest age-group (0-19 years), the average prevalence was 0.11%. In the age group between 50 and 69 years the prevalence was higher for men than for women, whi1.e the reverse was seen in the over 70 age group. In women a rapid in- crease in diabetes prevalence occurred after the age o f 70. At the end of the observation period the prevalence was about 3 % in women aged 60 to 69 years, whereas in women over 70 years the corresponding figure was about 9%. No significant change in the annual incidence was observed during the {I 10 period (Fig 2 ) . The incidence o f diabetes treated with oral drugs was somewhat higher in women than in men ( 2 . 4 and 2.1/1000 respectively). This tendency was also noted for insulin-treated diabetes (0.39 and 0.35/1000). z W V z W J 5 a 3 W a -71 -73 -75 -77 -79 -81-82 YEAR The annual incidence was low in all age groups given insulin (Fig 3). Fig 1. In the age group 0 - 19 years it was Prevalence of diabetes rnellitus, treated 0.16/1000. For those treated with oral antidiabetics, there was a con- with insulin o r oral antidiabetics in various age groups during the period 1971 - 1982 (0-0 men; women). siderable increase with age. The highest incidence (12.5/1000) was found in females over 80 years o f age. During the period no significant change in incidence was observed in any of the age groups. 91 f -71 -73 -75 -77 -79 -81 0 Y E A R B I - 4 B il -71 73 -75 -77 YEAR -79 -81 Fig.2. Annual incidence o f diabetes mellitus treated with insulin ( A ) o r oral anti diabetics ( B ) during the period 1971- 1982. 1 0 0 5 m W t . I z 0 0 - . Ln c z 10 w t a4 U. 0 i 5 5 1 A U d .r I rn b i 0 20 30 LO 50 60 70 00 , , , , , , I , , 9 19 29 39 L9 59 69 79 09 A G E GROUPS ( Y E A R S ) B U d .O 4 14 219 i 9 1 9 49 $9 719 s'9 AGE GROUPS ( Y E A R S ) Fig. 3. Average annual incidence o f diabetes mellitus treated with insulin ( A ) or oral antidiabetics ( 6 ) in various age groups during the period 1971 - 1982. Sales o f insulin and oral antidiabetics (excluding hospitals) Sales o f insulin (excluding hospitals) were higher in the county o f Jamt- land than in Sweden as a whole and increased slowly during the period (Fig.4). Sales o f oral antidiabetic drugs declined in the county in the middle o f the 7 0 ' s and have since then been somewhat. lower than in the rest o f Sweden (Fig.5). I L -75 -76 -77 -78 -79 -80 -81 -82 YEAR Fig.4. -74 -75 -76 -77 -78 -79 -80 -81 - 8 2 YEAR Fig.5. Sales o f insulin (excluding hospitals) in the county o f JWmtland ( 0-0 ) and cluding hospitals) in the county o f i n Sweden as a whole ( 0-0 ) during the Jsmtland ( ) and in Sweden as Sales o f oral antidiabetics ( e x - 92 period 1974 - 1982. Number of defined daily doses (DDD) per 1000 inhabitants 1974 - 1982. Number and day. inhabitants and day as a whole ( c-0 during the period o f DDD per 1000 Change in prescription pattern The prescribing of various types of insulin from 1970 to 1982 is shown in Fig.6. A steady increase was observed during the period for medium-acting in- sulin, while there was a decline in the use o f long-term and regular insulin. Amongst the oral antidiabetic drugs chlorpropamide was largely replaced by glibenclamide (Fig.7). The prescribing o f phenformin diminished after 1973 and it was withdrawn from the Swedish market in 1978. 100 % 50 '10 Medium -. -,.. ong-term \. -.-.- .-____._ 1 1 ~ 1 1 1 ~ 1 1 1 ~ -70 -72 -74 -76 -78 -80 -82 Y E A R Fig.6. Prescription of various types of in- sulin in the county of JSmtLand 1970- 1982. Percentage distribution based on the number of DDD. l o o *I, 50 % -. Metformin -.-. .//.b.-.-.-./ / Phenformin 'l-., ,.' \. ' GLibenctamide Chlor propamide \. -70 -72 -74 -76 -78 -80 -82 YEAR Fig.7. Prescription of various types of oral antidiabetics in the county of Jamt- land 1970 - 1982. Percentage distri- bution based on the number of DDD. The relation between prescription of insulin and oral antidiabetics varied with age and with time (Fig.8). At the beginning of the 70's combined anti- diabetic therapy (insulin + oral drugs, sulfonyl urea + biguanid drugs) was frequently adopted. At the end of the period hardly any patients were prescri- bed insul.in in combination with oral drugs and very few used combinations of oral drugs. In 1982 40% of drug-treated diabetics were prescribed insulin and 605 oral drugs. 93 19 29 39 1 9 59 6 9 7 9 1975 50% 1 1 10 i p (0 5 0 6? 7;O 8; 19 29 39 19 59 69 7 9 1982 50 *I. Fig. 8. Prescription in various age groups; insulin (I), oral antidiabetics (m), insulin,+ oral drugs (m), and sulfonyl urea + biguanid (B). Per- centage distribution o f pat- ients. The change in diabetes treatment in individual pat- ients during the period is illustrated in Table 2 . Several patients changed from oral drugs to insulin while the reverse was l e s s common. A change from one to two o r a l drugs was instituted fairly frequently at the earliest part o f the observation period but then became l e s s common. I I 2 0 3p L p 5,O 60 7,O 8,O i"g 29 3 9 1 9 5; 6'9 ;9 AGE GROUPS ( Y E A R S ) Table 2 . Changes o f treatment in individual patients. Change over -72 -73 -74 -75 -76 -77 -78 -79 -80 -81 -82 Oral drug-> insulin 7 11 8 13 11 12 16 4 15 11 1 3 I nsu 1 i n-> One oral drug--) Newly diagnosed diabetics 52 51 45 60 41 47 4 3 38 5 5 42 49 Diabetics oral drug 3 2 3 1 0 1 2 2 0 1 1 two 4 8 5 i i 5 4 5 5 1 4 3 registered 287 31 8 335 365 378 94 Seasonal variation There was no significant seasonal variation in the total sales of insulin o r oral antidiabetics during the period studied (Fig.9). However, when con- sidering each patient's first purchase of antidiabetic drugs, more patients were found to start therapy with oral antidiabetics during the first half of the year than during the latter (Fig.10). This difference was statistically significant (p