510794 2013
SJP42210.1177/1403494813510794F. Nilson et al.Short Title
Scandinavian Journal of Public Health, 2014; 42: 201–206
Original Article
A comparison of hip fracture incidence rates among elderly in Sweden by latitude and sunlight exposure
Finn Nilson, Syed Moniruzzaman & Ragnar Andersson Division of Risk Management, Department of Environmental and Life Sciences, Karlstad University, Karlstad, Sweden
Abstract Background: Research has shown that hip fracture risk increases with latitude; hypothetically due to reduced sunlight exposure and its effect on bone quality. Sweden, with large differences in latitude and UV radiation, is ideal to study in order to analyse the association between latitude and UV radiation on age- and sex-specific hip fracture rates among elderly. Method: Aggregated (2006–2008) age- and sex-specific hip fracture data was obtained for each Swedish municipality as well as the municipality’s latitudinal coordinates and aggregated (2006–2008) UV radiation levels. Pearson correlations were calculated between hip fracture incidence rates, latitude and UV radiation. Independent t tests were calculated on tertile-categorized latitudinal data in order to investigate the difference in hip fracture risk between these categories. Results: Statistically significant correlations were seen in all groups between hip fracture incidence rates and latitude as well as UV radiation. The independent t tests showed that this correlation was mainly due to high incidence rates in high latitude municipalities. Conclusions: Statistically significant correlations are seen between hip fracture incidence rates and latitude as well as UV radiation in Sweden and the northern parts of Sweden have an increased risk of hip fractures compared to the middle and southern parts. To our knowledge this is the first study using a national discharge register that shows this relationship and provides a starting point for further research to investigate why populations in northern Sweden have a higher risk of hip fractures compared to other Swedish regions. Key Words: Elderly, epidemiology, falls, hip fractures, latitude, UV radiation
Background Hip fractures among elderly are recognized as a major public health problem in the Scandinavian countries. Despite recent studies showing declines in hip fracture incidence rates in Scandinavia since the 1990s [1–4], Scandinavian countries, and Austria, have the highest hip-fracture incidence rates in the world and hip-fracture rates are generally higher in Northern countries [5]. Some hypothetical causes have been suggested to this; the effect of sunlight exposure on bone quality [6], increased risk of hip fracture due to ice and snow [7] and impaired physical capacity during winter [8]. The direct effect of ice and snow on hip fracture risk is questionable in Sweden as a majority of hip fractures occur indoors, rather than outdoors [9]. However, the effect of ice
and snow on the individual’s physical capacity in terms of the ability and opportunity to exercise should not be underestimated, although this is difficult to measure without following a defined population longitudinally. Sunlight exposure in Swedish municipalities, on the other hand, has been measured and registered for a long period of time [10] and it is therefore possible to investigate potential correlations to hip fracture rates. Sunlight exposure is thought to have an effect on hip fracture risk due to the well-established importance of vitamin D in osteoporosis prevention and the subsequent increased risk of hip fractures due to osteoporosis. Within sunlight exposure, the important factor is UVB radiation that synthesizes vitamin
Correspondence: Finn Nilson, Division of Risk Management, Department of Environmental and Life Sciences, Karlstad University, SE-651 88 Karlstad, Sweden. E-mail: [email protected] (Accepted 7 October 2013) © 2013 the Nordic Societies of Public Health DOI: 10.1177/1403494813510794
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
202 F. Nilson et al. Table I. Pearson correlations between municipalities age- and sex-specific hip fracture incidence rates and the municipality’s latitude and UV radiation. Latitude
UV Radiation
Pearson correlation
Sig. (2-tailed)
Pearson correlation
Sig. (2-tailed)
Women aged 65–79 Women aged 80+ Men aged 65–79 Men aged 80+ Latitude UV Radiation
0.307 0.206 0.207 0.337 1 −0.922
0.000 0.000 0.000 0.000
−0.305 −0.124 −0.227 −0.316 −0.922 1
0.000 0.034 0.000 0.000 0.000
0.000
D in the body, leading to an increase in the uptake of calcium [11]. A number of studies have shown increases in fracture probability with changes in latitude and have explained this in part due to sunlight exposure [12,13]. Using national hospital discharge registers, the latitudinal differences in hip fracture incidence rates have yet to be studied in Sweden, though have been studied in Germany and France [14,15]. These studies have observed some correlations to socioeconomic factors, but no evidence of latitudinal differences. However, both Germany and France are central-European countries with small latitudinal differences (Germany: 47° to 55° N; France: 41° to 51° N). Sweden on the other hand is located in northern Europe between latitudes 55.3° and 69.0° N, leading to considerable differences in sunlight exposure, and is therefore ideal in studying latitudinal differences in hip fracture incidence. The current study therefore, aims to analyse the association between latitude and UV radiation on age- and sex-specific hip fracture rates among elderly in Sweden. Methods Data source and selection Data for this study was obtained from the Swedish National Patient Register (NPR) regarding hip fractures among those 65 years and above, and information on gender, age and residential municipality, on a yearly basis, from 2006 to 2008. Hip fractures have in this study been defined as S720-7229 according to the International Classification of Diseases (ICD10). The NPR includes all hospital admitted patients and their diagnoses and is considered valid and reliable [16]. Yearly population data for those 65 years and above for each municipality were obtained from Statistics Sweden. Due to the Swedish Personal Identification Number (PIN), the data is highly reliable with 99.9 % of the Swedish population having a PIN [17]. Triangulated GPS coordinates for each
municipality were obtained from Statistics Sweden in order to test for latitude. These coordinates were then used in the database STRÅNG [18] to obtain a 3-year aggregate (2006–2008) UV radiation value for each municipality. Statistical analysis The statistical analysis was performed in four steps. Firstly, 3-year interval (2006–2008) hip fracture incidence-rates per 100,000 in age- and sex-specific groups at municipality level as well as 3-year interval (2006–2008) UV radiation values per municipality were calculated. Secondly, Pearson correlations were calculated between the age- and sex-specific hip fracture incidence rates and the municipality’s latitude and UV radiation. Thirdly, municipalities’ age- and sex-specific hip fracture incidence rates were plotted against the municipality’s latitude. Based on these scatterplots, the latitudinal data were categorized, based on tertiles, into a low (up to 57.58 degrees), mid (57.59 – 60.11 degrees) or high (above 60.12 degrees) latitude category. Independent t tests were calculated in order to investigate the difference in hip fracture risk between these categories. Due to some municipalities having small elderly populations, all analysis was also performed after having removed the first fifth percentile in terms of sex-specific elderly (over 65 years) population size. P values ≤ 0.05 were considered to be statistically significant. Results As is seen in Table I, statistically significant correlations were seen between all age- and sex-specific groups with latitude, showing an increased hip fracture risk with increases in latitude. Statistically significant correlations were also seen between all groups and UV radiation with an increased hip fracture risk with decreasing UV radiation. A high correlation between latitude and UV radiation (correlation coefficient: −0.922) was also observed
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
A comparison of hip fracture incidence 203
Figure 1. Scatterplot with linear trend line for municipalities’ age- and sex-specific hip fracture incidence rates and the municipality’s latitude. Table II. Pearson correlations between municipalities’ age- and sex-specific hip fracture incidence rates and the municipality’s latitude and UV radiation using restricted data. Latitude
UV Radiation
Pearson correlation
Sig. (2-tailed)
Pearson correlation
Sig. (2-tailed)
Women aged 65–79 Women aged 80+ Men aged 65–79 Men aged 80+ Latitude UV Radiation
0.308 0.229 0.191 0.327 1 −0.912
0.000 0.000 0.001 0.000
−0.306 −0.131 −0.221 −0.300 −0.912 1
0.000 0.029 0.000 0.000 0.000
0.000
(Table I). As is also clear, the distribution of plots seems to differ depending upon the latitudinal level and indicate a shift towards higher incidence rates especially at higher latitudes (Figure 1). Similar results are seen to the original data when the restricted data was used. Statistically significant correlations were still seen with all age- and sexspecific groups and latitude as well as UV radiation. Very close correlations were still seen between latitude and UV radiation (Table II). Due to the high correlation between latitude and UV radiation, analysis was focused on latitude. Also, as no great effects were observed when the smallest populations were removed, the original data was
used for the continued analysis. In order to test the effect of latitude on a group level, latitudinal values were categorized into high, mid or low categories and independent t tests were computed. As is seen in Table III, no significant differences were observed between the mid and low latitudes whereas significant differences were seen between high and mid as well as high and low latitude for all age- and sex-specific groups. This is also seen in the percentage difference in means. While only small differences are seen between low and mid latitudes (3.03%–5.79%), considerably larger differences are seen between mid and high latitudes (8.85%–21.51%) and low and high latitudes (12.15%–28.54%).
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
24.30
22.67
12.15
0.000 5.195 18.97
0.001 3.491 17.31
0.000 3.762 8.85
0.000 4.242 4.48
0.004 2.901 4.57
0.005 2.849 3.03 1.257 0.210
0.896 0.371
1.339 0.182
Women aged 80+
Men aged 65–79
Men aged 80+
0.118
27.28 (−6.99–61.55) 77.94 (−44.37–200.25) 13.35 (−16.06–42.76) 71.79 (−33.97–177.56)
5.79
4.836
0.000
107.29 (63.42–151.15) 234.29 (71.55–397.04) 52.9 (16.83–88.96) 317.94 (169.54–466.33)
21.51
6.198
0.000
134.57 (91.58–177.55) 312.23 (147.91–476.55) 66.25 (28.70–103.79) 389.73 (241.13–538.33)
28.54
Discussion
Women aged 65–79 1.57
t value Sig. (2-tailed) Mean Difference Difference in (95% CI) mean (%) t value Sig. (2-tailed) Mean Difference Difference in t value Sig. Mean Difference Difference in (95% CI) mean (%) (2-tailed) (95% CI) mean (%)
High–middle Middle–low
Table III. Independent t tests for latitudinal categories and hip fracture incidence rates.
High–low
204 F. Nilson et al. This study shows that strong statistically significant correlations are seen between hip fracture incidence rates and latitude as well as UV radiation with an increased hip fracture risk in municipalities at high latitude or with low UV radiation. This study also shows that with regards to hip fracture incidence, the large statistical effects of latitude on hip fracture incidence are when comparing the southern and middle part of Sweden with the northern part. Between latitudes of 55.3° to 60.11° N, no statistically significant differences are seen on a group level while statistically significant differences are seen in comparison with the incidence risk among those living at latitudes of 60.11° to 69.0° N. The previous national studies on latitude and hip fracture incidence rates from France and Germany showed no effect of latitude on hip fracture rates, hypothetically due to small differences in latitude within the country [14,15]. The dose-dependency that seems to be seen in this study, with the northernmost part of Sweden having considerably higher risk, may support this hypothesis and explain why no differences were seen in France and Germany. However, while a closely correlated relationship between latitude and UV radiation is seen in this study, as well as in previous studies [19,20], this merely indicates an association between the factors, not a causal relationship. The populations in northern Sweden seem to be at a higher risk for hip fractures but whether this is caused by less UV radiation or because of other variables, either solely or in combinations, such as climate, snow/ice, socio-economy and population density, needs to be studied further. The importance of osteoporosis, vitamin D and sunlight on hip fracture risk has been shown in a number of previous studies [21,22]. However, sunlight is not the only source of vitamin D. Studies in northern Norway and Sweden indicate that an intake of vitamin D, either through diet or supplements, greatly reduces the effect of latitudinal and seasonal variations [23,24], indicating instead that the low availability of sunlight-produced vitamin D can be compensated by dietary intake. As no data is available for hip fracture patients’ vitamin D status, this cannot be controlled for. Vitamin D is of importance for bone strength but is also a crucial ingredient in the preservation of muscle strength and reduced risk of sarcopenia [25]. Bone and muscle strength is also closely related to physical activity, and low physical activity is a known risk factor for hip fractures [21]. In the north of Sweden (latitudes of between 60.11° to 69.0° N), snow covers the ground for between 150 and 225 days per year [26], meaning that the ability
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
A comparison of hip fracture incidence 205 to pursue outdoor physical activity for elderly is likely to be limited during a considerable part of the year. A number of other possible confounders exist that could explain the increased risk in the north of Sweden. For example, potential differences in diet between the north and south of Sweden or BMI differences among elderly may affect the hip fracture risk though data has not been available in order to control for these aspects. Also the impact of rural and urban areas, as well as socio-economic factors, may affect the results. The north of Sweden is considerably less densely populated compared to the middle or southern parts, although a study from Norway has shown higher hip fracture rates in urban areas compared to rural areas, hypothetically due to a more physically active lifestyle in rural areas [27]. Socio-economic factors, such as educational level, income and occupation, have previously shown to affect hip fracture risk [28] and could therefore confound the results. These are well known risk factors for hip fractures and in future studies these types of factors should be investigated further. There are some limitations in this study. Firstly, while the Swedish NPR data is deemed reliable, the authors of this study have not had control over the injury registration process and therefore mistakes and misses cannot be ruled out. Secondly, although the data is controlled to eliminate double registrations due to readmissions within a single year, it has not been possible to connect hip fracture patients between different years. As a readmission can occur between 2 years, this is a limitation though we have judged this as a small risk that doesn’t substantially affect the results. Lastly, UV radiation is known to be difficult to measure accurately and therefore there may be errors in the STRÅNG database. Also, although the effects of UV radiation on vitamin D levels occurs quickly [11], the long-term effects of low vitamin D levels have a more profound effect with regards to osteoporosis. Therefore, it could be argued that the levels of UV radiation for 2006–2008 are not directly linked to the hip fractures occurring during the same time period. However, although increases are seen in UV radiation in Sweden since 1983, there is no evidence suggesting that the latitudinal differences in UV radiation within Sweden has changed substantially over time, or that large fluctuations have occurred between years [10]. To overcome this limitation, data analysis was focused on latitudinal data that was highly correlated to UV radiation data. Conclusions This study can show that statistically significant correlations are seen between hip fracture incidence rates and latitude as well as UV radiation in Sweden,
and that the northern parts of Sweden have an increased risk of hip fractures compared to the middle and southern parts. While this study cannot show why latitude affects hip fracture risk among elderly, it is to our knowledge the first study using a national discharge register that shows this relationship and provides a starting point for further research to investigate why populations in northern Sweden have a higher risk of hip fractures compared to other Swedish regions. Conflicts of interest None declared. Funding This research received no specific grant from any funding agency in the public, commercial, or not-forprofit sectors. References [1] Kannus P, Niemi S, Parkkari J, et al. Nationwide decline in incidence of hip fracture. J Bone Miner Res 2006;21: 1836–8. [2] Nilson F, Moniruzzaman S, Gustavsson J, et al. Trends in hip fracture incidence rates among the elderly in Sweden 1987–2009. J Publ Health (Oxf) 2013;35:125–31. [3] Abrahamsen B and Vestergaard P. Declining incidence of hip fractures and the extent of use of anti-osteoporotic therapy in Denmark 1997–2006. Osteoporos Int 2010;21:373–80. [4] Omsland TK, Holvik K, Meyer HE, et al. Hip fractures in Norway 1999–2008: time trends in total incidence and second hip fracture rates. A NOREPOS study. Eur J Epidemiol 2012;27(10):807–814. [5] Kanis JA, Oden A, McCloskey EV, et al. A systematic review of hip fracture incidence and probability of fracture worldwide. Osteoporos Int 2012;23(9):2239–2256. [6] Rosen CJ, Morrison A, Zhou H, et al. Elderly women in northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones. Bone Miner 1994;25:83–92. [7] Ralis ZA. Epidemics of fractures during periods of snow and ice. Br Med J (Clin Res Ed) 1986;293:484. [8] Bastow MD, Rawlings J and Allison SP. Undernutrition, hypothermia, and injury in elderly women with fractured femur: an injury response to altered metabolism? The Lancet 1983;1: 143–6. [9] Schyllander J and Rosenberg T. Skador bland äldre personer i Sverige [Injuries among elderly people in Sweden]. Karlstad: Enheten för lärande från olyckor och kriser, Myndigheten för samhällsskydd och beredskap, 2009. [10] Josefsson W. UV-radiation 1983–2003 measured at Norrköping, Sweden. Theor Appl Climatol 2006;83:59–76. [11] Engelsen O. The relationship between ultraviolet radiation exposure and vitamin D status. Nutrients 2010;2:482–95. [12] Ismail AA, Pye SR, Cockerill WC, et al. Incidence of limb fracture across Europe: results from the European Prospective Osteoporosis Study (EPOS). Osteoporos Int 2002;13:565–71. [13] Johnell O, Borgstrom F, Jonsson B, et al. Latitude, socioeconomic prosperity, mobile phones and hip fracture risk. Osteoporos Int 2007;18:333–7. [14] Barbier S, Ecochard R, Schott AM, et al. Geographical variations in hip fracture risk for women: strong effects
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
206 F. Nilson et al. hidden in standardised ratios. Osteoporos Int 2009;20: 371–7. [15] Icks A, Haastert B, Wildner M, et al. Hip fractures and area level socioeconomic conditions: a population-based study. BMC Publ Health 2009;9:114. [16] Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Publ Health 2011;11:450. [17] Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, et al. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol 2009;24:659–67. [18] SMHI. STRÅNG - a mesoscale model for solar radiation, http://strang.smhi.se (2013, accessed 30 May 2013). [19] Webb AR, Kline L and Holick MF. Influence of season and latitude on the cutaneous synthesis of vitamin D3: exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J Clin Endocrinol Metab 1988;67:373–8. [20] Engelsen O, Brustad M, Aksnes L, et al. Daily duration of vitamin D synthesis in human skin with relation to latitude, total ozone, altitude, ground cover, aerosols and cloud thickness. Photochem Photobiol 2005;81:1287–90.
[21] Marks R. Hip fracture epidemiological trends, outcomes, and risk factors, 1970–2009. Int J Gen Med 2010;3: 1–17. [22] Rubenstein LZ. Falls in older people: epidemiology, risk faktors and strategies for prevention. Age Ageing 2006;35(Suppl 2): ii37–ii41. [23] Burgaz A, Akesson A, Oster A, et al. Associations of diet, supplement use, and ultraviolet B radiation exposure with vitamin D status in Swedish women during winter. Am J Clin Nutr 2007;86:1399–1404. [24] Brustad M, Edvardsen K, Wilsgaard T, et al. Seasonality of UV-radiation and vitamin D status at 69 degrees north. Photochem Photobiol Sci 2007;6:903–8. [25] Holick MF and Chen TC. Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr 2008;87:1080S–6S. [26] SMHI. Climate data, http://www.smhi.se/klimatdata/meteorologi/sno (2009, accessed 1 May 2013). [27] Finsen V and Benum P. Changing incidence of hip fractures in rural and urban areas of central Norway. Clin Orthop Relat Res 1987;May:104–10. [28] Benetos IS, Babis GC, Zoubos AB, et al. Factors affecting the risk of hip fractures. Injury 2007;38:735–44.
Information for Authors Why do we ask you to transfer copyright to the Nordic Societies of Public Health? 1. Ownership of copyright by the Nordic Societies of Public Health ensures maximum protection against infringement of copyright protected material through breach of copyright or piracy anywhere in the world. 2. It also ensures that requests by third parties to reprint or reproduce a contribution, or part of it in any format, are handled efficiently in accordance with our general policy which encourages dissemination of knowledge inside the framework of copyright. SAGE administers the copyright and permissions for SJPH on behalf of the Nordic Societies of Public Health. For more information please go to the Author gateway: http://www.sagepub.co.uk/ journalEditors.nav
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
SJP42210.1177/1403494813510794F. Nilson et al.Short Title
Scandinavian Journal of Public Health, 2014; 42: 201–206
Original Article
A comparison of hip fracture incidence rates among elderly in Sweden by latitude and sunlight exposure
Finn Nilson, Syed Moniruzzaman & Ragnar Andersson Division of Risk Management, Department of Environmental and Life Sciences, Karlstad University, Karlstad, Sweden
Abstract Background: Research has shown that hip fracture risk increases with latitude; hypothetically due to reduced sunlight exposure and its effect on bone quality. Sweden, with large differences in latitude and UV radiation, is ideal to study in order to analyse the association between latitude and UV radiation on age- and sex-specific hip fracture rates among elderly. Method: Aggregated (2006–2008) age- and sex-specific hip fracture data was obtained for each Swedish municipality as well as the municipality’s latitudinal coordinates and aggregated (2006–2008) UV radiation levels. Pearson correlations were calculated between hip fracture incidence rates, latitude and UV radiation. Independent t tests were calculated on tertile-categorized latitudinal data in order to investigate the difference in hip fracture risk between these categories. Results: Statistically significant correlations were seen in all groups between hip fracture incidence rates and latitude as well as UV radiation. The independent t tests showed that this correlation was mainly due to high incidence rates in high latitude municipalities. Conclusions: Statistically significant correlations are seen between hip fracture incidence rates and latitude as well as UV radiation in Sweden and the northern parts of Sweden have an increased risk of hip fractures compared to the middle and southern parts. To our knowledge this is the first study using a national discharge register that shows this relationship and provides a starting point for further research to investigate why populations in northern Sweden have a higher risk of hip fractures compared to other Swedish regions. Key Words: Elderly, epidemiology, falls, hip fractures, latitude, UV radiation
Background Hip fractures among elderly are recognized as a major public health problem in the Scandinavian countries. Despite recent studies showing declines in hip fracture incidence rates in Scandinavia since the 1990s [1–4], Scandinavian countries, and Austria, have the highest hip-fracture incidence rates in the world and hip-fracture rates are generally higher in Northern countries [5]. Some hypothetical causes have been suggested to this; the effect of sunlight exposure on bone quality [6], increased risk of hip fracture due to ice and snow [7] and impaired physical capacity during winter [8]. The direct effect of ice and snow on hip fracture risk is questionable in Sweden as a majority of hip fractures occur indoors, rather than outdoors [9]. However, the effect of ice
and snow on the individual’s physical capacity in terms of the ability and opportunity to exercise should not be underestimated, although this is difficult to measure without following a defined population longitudinally. Sunlight exposure in Swedish municipalities, on the other hand, has been measured and registered for a long period of time [10] and it is therefore possible to investigate potential correlations to hip fracture rates. Sunlight exposure is thought to have an effect on hip fracture risk due to the well-established importance of vitamin D in osteoporosis prevention and the subsequent increased risk of hip fractures due to osteoporosis. Within sunlight exposure, the important factor is UVB radiation that synthesizes vitamin
Correspondence: Finn Nilson, Division of Risk Management, Department of Environmental and Life Sciences, Karlstad University, SE-651 88 Karlstad, Sweden. E-mail: [email protected] (Accepted 7 October 2013) © 2013 the Nordic Societies of Public Health DOI: 10.1177/1403494813510794
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
202 F. Nilson et al. Table I. Pearson correlations between municipalities age- and sex-specific hip fracture incidence rates and the municipality’s latitude and UV radiation. Latitude
UV Radiation
Pearson correlation
Sig. (2-tailed)
Pearson correlation
Sig. (2-tailed)
Women aged 65–79 Women aged 80+ Men aged 65–79 Men aged 80+ Latitude UV Radiation
0.307 0.206 0.207 0.337 1 −0.922
0.000 0.000 0.000 0.000
−0.305 −0.124 −0.227 −0.316 −0.922 1
0.000 0.034 0.000 0.000 0.000
0.000
D in the body, leading to an increase in the uptake of calcium [11]. A number of studies have shown increases in fracture probability with changes in latitude and have explained this in part due to sunlight exposure [12,13]. Using national hospital discharge registers, the latitudinal differences in hip fracture incidence rates have yet to be studied in Sweden, though have been studied in Germany and France [14,15]. These studies have observed some correlations to socioeconomic factors, but no evidence of latitudinal differences. However, both Germany and France are central-European countries with small latitudinal differences (Germany: 47° to 55° N; France: 41° to 51° N). Sweden on the other hand is located in northern Europe between latitudes 55.3° and 69.0° N, leading to considerable differences in sunlight exposure, and is therefore ideal in studying latitudinal differences in hip fracture incidence. The current study therefore, aims to analyse the association between latitude and UV radiation on age- and sex-specific hip fracture rates among elderly in Sweden. Methods Data source and selection Data for this study was obtained from the Swedish National Patient Register (NPR) regarding hip fractures among those 65 years and above, and information on gender, age and residential municipality, on a yearly basis, from 2006 to 2008. Hip fractures have in this study been defined as S720-7229 according to the International Classification of Diseases (ICD10). The NPR includes all hospital admitted patients and their diagnoses and is considered valid and reliable [16]. Yearly population data for those 65 years and above for each municipality were obtained from Statistics Sweden. Due to the Swedish Personal Identification Number (PIN), the data is highly reliable with 99.9 % of the Swedish population having a PIN [17]. Triangulated GPS coordinates for each
municipality were obtained from Statistics Sweden in order to test for latitude. These coordinates were then used in the database STRÅNG [18] to obtain a 3-year aggregate (2006–2008) UV radiation value for each municipality. Statistical analysis The statistical analysis was performed in four steps. Firstly, 3-year interval (2006–2008) hip fracture incidence-rates per 100,000 in age- and sex-specific groups at municipality level as well as 3-year interval (2006–2008) UV radiation values per municipality were calculated. Secondly, Pearson correlations were calculated between the age- and sex-specific hip fracture incidence rates and the municipality’s latitude and UV radiation. Thirdly, municipalities’ age- and sex-specific hip fracture incidence rates were plotted against the municipality’s latitude. Based on these scatterplots, the latitudinal data were categorized, based on tertiles, into a low (up to 57.58 degrees), mid (57.59 – 60.11 degrees) or high (above 60.12 degrees) latitude category. Independent t tests were calculated in order to investigate the difference in hip fracture risk between these categories. Due to some municipalities having small elderly populations, all analysis was also performed after having removed the first fifth percentile in terms of sex-specific elderly (over 65 years) population size. P values ≤ 0.05 were considered to be statistically significant. Results As is seen in Table I, statistically significant correlations were seen between all age- and sex-specific groups with latitude, showing an increased hip fracture risk with increases in latitude. Statistically significant correlations were also seen between all groups and UV radiation with an increased hip fracture risk with decreasing UV radiation. A high correlation between latitude and UV radiation (correlation coefficient: −0.922) was also observed
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
A comparison of hip fracture incidence 203
Figure 1. Scatterplot with linear trend line for municipalities’ age- and sex-specific hip fracture incidence rates and the municipality’s latitude. Table II. Pearson correlations between municipalities’ age- and sex-specific hip fracture incidence rates and the municipality’s latitude and UV radiation using restricted data. Latitude
UV Radiation
Pearson correlation
Sig. (2-tailed)
Pearson correlation
Sig. (2-tailed)
Women aged 65–79 Women aged 80+ Men aged 65–79 Men aged 80+ Latitude UV Radiation
0.308 0.229 0.191 0.327 1 −0.912
0.000 0.000 0.001 0.000
−0.306 −0.131 −0.221 −0.300 −0.912 1
0.000 0.029 0.000 0.000 0.000
0.000
(Table I). As is also clear, the distribution of plots seems to differ depending upon the latitudinal level and indicate a shift towards higher incidence rates especially at higher latitudes (Figure 1). Similar results are seen to the original data when the restricted data was used. Statistically significant correlations were still seen with all age- and sexspecific groups and latitude as well as UV radiation. Very close correlations were still seen between latitude and UV radiation (Table II). Due to the high correlation between latitude and UV radiation, analysis was focused on latitude. Also, as no great effects were observed when the smallest populations were removed, the original data was
used for the continued analysis. In order to test the effect of latitude on a group level, latitudinal values were categorized into high, mid or low categories and independent t tests were computed. As is seen in Table III, no significant differences were observed between the mid and low latitudes whereas significant differences were seen between high and mid as well as high and low latitude for all age- and sex-specific groups. This is also seen in the percentage difference in means. While only small differences are seen between low and mid latitudes (3.03%–5.79%), considerably larger differences are seen between mid and high latitudes (8.85%–21.51%) and low and high latitudes (12.15%–28.54%).
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
24.30
22.67
12.15
0.000 5.195 18.97
0.001 3.491 17.31
0.000 3.762 8.85
0.000 4.242 4.48
0.004 2.901 4.57
0.005 2.849 3.03 1.257 0.210
0.896 0.371
1.339 0.182
Women aged 80+
Men aged 65–79
Men aged 80+
0.118
27.28 (−6.99–61.55) 77.94 (−44.37–200.25) 13.35 (−16.06–42.76) 71.79 (−33.97–177.56)
5.79
4.836
0.000
107.29 (63.42–151.15) 234.29 (71.55–397.04) 52.9 (16.83–88.96) 317.94 (169.54–466.33)
21.51
6.198
0.000
134.57 (91.58–177.55) 312.23 (147.91–476.55) 66.25 (28.70–103.79) 389.73 (241.13–538.33)
28.54
Discussion
Women aged 65–79 1.57
t value Sig. (2-tailed) Mean Difference Difference in (95% CI) mean (%) t value Sig. (2-tailed) Mean Difference Difference in t value Sig. Mean Difference Difference in (95% CI) mean (%) (2-tailed) (95% CI) mean (%)
High–middle Middle–low
Table III. Independent t tests for latitudinal categories and hip fracture incidence rates.
High–low
204 F. Nilson et al. This study shows that strong statistically significant correlations are seen between hip fracture incidence rates and latitude as well as UV radiation with an increased hip fracture risk in municipalities at high latitude or with low UV radiation. This study also shows that with regards to hip fracture incidence, the large statistical effects of latitude on hip fracture incidence are when comparing the southern and middle part of Sweden with the northern part. Between latitudes of 55.3° to 60.11° N, no statistically significant differences are seen on a group level while statistically significant differences are seen in comparison with the incidence risk among those living at latitudes of 60.11° to 69.0° N. The previous national studies on latitude and hip fracture incidence rates from France and Germany showed no effect of latitude on hip fracture rates, hypothetically due to small differences in latitude within the country [14,15]. The dose-dependency that seems to be seen in this study, with the northernmost part of Sweden having considerably higher risk, may support this hypothesis and explain why no differences were seen in France and Germany. However, while a closely correlated relationship between latitude and UV radiation is seen in this study, as well as in previous studies [19,20], this merely indicates an association between the factors, not a causal relationship. The populations in northern Sweden seem to be at a higher risk for hip fractures but whether this is caused by less UV radiation or because of other variables, either solely or in combinations, such as climate, snow/ice, socio-economy and population density, needs to be studied further. The importance of osteoporosis, vitamin D and sunlight on hip fracture risk has been shown in a number of previous studies [21,22]. However, sunlight is not the only source of vitamin D. Studies in northern Norway and Sweden indicate that an intake of vitamin D, either through diet or supplements, greatly reduces the effect of latitudinal and seasonal variations [23,24], indicating instead that the low availability of sunlight-produced vitamin D can be compensated by dietary intake. As no data is available for hip fracture patients’ vitamin D status, this cannot be controlled for. Vitamin D is of importance for bone strength but is also a crucial ingredient in the preservation of muscle strength and reduced risk of sarcopenia [25]. Bone and muscle strength is also closely related to physical activity, and low physical activity is a known risk factor for hip fractures [21]. In the north of Sweden (latitudes of between 60.11° to 69.0° N), snow covers the ground for between 150 and 225 days per year [26], meaning that the ability
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
A comparison of hip fracture incidence 205 to pursue outdoor physical activity for elderly is likely to be limited during a considerable part of the year. A number of other possible confounders exist that could explain the increased risk in the north of Sweden. For example, potential differences in diet between the north and south of Sweden or BMI differences among elderly may affect the hip fracture risk though data has not been available in order to control for these aspects. Also the impact of rural and urban areas, as well as socio-economic factors, may affect the results. The north of Sweden is considerably less densely populated compared to the middle or southern parts, although a study from Norway has shown higher hip fracture rates in urban areas compared to rural areas, hypothetically due to a more physically active lifestyle in rural areas [27]. Socio-economic factors, such as educational level, income and occupation, have previously shown to affect hip fracture risk [28] and could therefore confound the results. These are well known risk factors for hip fractures and in future studies these types of factors should be investigated further. There are some limitations in this study. Firstly, while the Swedish NPR data is deemed reliable, the authors of this study have not had control over the injury registration process and therefore mistakes and misses cannot be ruled out. Secondly, although the data is controlled to eliminate double registrations due to readmissions within a single year, it has not been possible to connect hip fracture patients between different years. As a readmission can occur between 2 years, this is a limitation though we have judged this as a small risk that doesn’t substantially affect the results. Lastly, UV radiation is known to be difficult to measure accurately and therefore there may be errors in the STRÅNG database. Also, although the effects of UV radiation on vitamin D levels occurs quickly [11], the long-term effects of low vitamin D levels have a more profound effect with regards to osteoporosis. Therefore, it could be argued that the levels of UV radiation for 2006–2008 are not directly linked to the hip fractures occurring during the same time period. However, although increases are seen in UV radiation in Sweden since 1983, there is no evidence suggesting that the latitudinal differences in UV radiation within Sweden has changed substantially over time, or that large fluctuations have occurred between years [10]. To overcome this limitation, data analysis was focused on latitudinal data that was highly correlated to UV radiation data. Conclusions This study can show that statistically significant correlations are seen between hip fracture incidence rates and latitude as well as UV radiation in Sweden,
and that the northern parts of Sweden have an increased risk of hip fractures compared to the middle and southern parts. While this study cannot show why latitude affects hip fracture risk among elderly, it is to our knowledge the first study using a national discharge register that shows this relationship and provides a starting point for further research to investigate why populations in northern Sweden have a higher risk of hip fractures compared to other Swedish regions. Conflicts of interest None declared. Funding This research received no specific grant from any funding agency in the public, commercial, or not-forprofit sectors. References [1] Kannus P, Niemi S, Parkkari J, et al. Nationwide decline in incidence of hip fracture. J Bone Miner Res 2006;21: 1836–8. [2] Nilson F, Moniruzzaman S, Gustavsson J, et al. Trends in hip fracture incidence rates among the elderly in Sweden 1987–2009. J Publ Health (Oxf) 2013;35:125–31. [3] Abrahamsen B and Vestergaard P. Declining incidence of hip fractures and the extent of use of anti-osteoporotic therapy in Denmark 1997–2006. Osteoporos Int 2010;21:373–80. [4] Omsland TK, Holvik K, Meyer HE, et al. Hip fractures in Norway 1999–2008: time trends in total incidence and second hip fracture rates. A NOREPOS study. Eur J Epidemiol 2012;27(10):807–814. [5] Kanis JA, Oden A, McCloskey EV, et al. A systematic review of hip fracture incidence and probability of fracture worldwide. Osteoporos Int 2012;23(9):2239–2256. [6] Rosen CJ, Morrison A, Zhou H, et al. Elderly women in northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones. Bone Miner 1994;25:83–92. [7] Ralis ZA. Epidemics of fractures during periods of snow and ice. Br Med J (Clin Res Ed) 1986;293:484. [8] Bastow MD, Rawlings J and Allison SP. Undernutrition, hypothermia, and injury in elderly women with fractured femur: an injury response to altered metabolism? The Lancet 1983;1: 143–6. [9] Schyllander J and Rosenberg T. Skador bland äldre personer i Sverige [Injuries among elderly people in Sweden]. Karlstad: Enheten för lärande från olyckor och kriser, Myndigheten för samhällsskydd och beredskap, 2009. [10] Josefsson W. UV-radiation 1983–2003 measured at Norrköping, Sweden. Theor Appl Climatol 2006;83:59–76. [11] Engelsen O. The relationship between ultraviolet radiation exposure and vitamin D status. Nutrients 2010;2:482–95. [12] Ismail AA, Pye SR, Cockerill WC, et al. Incidence of limb fracture across Europe: results from the European Prospective Osteoporosis Study (EPOS). Osteoporos Int 2002;13:565–71. [13] Johnell O, Borgstrom F, Jonsson B, et al. Latitude, socioeconomic prosperity, mobile phones and hip fracture risk. Osteoporos Int 2007;18:333–7. [14] Barbier S, Ecochard R, Schott AM, et al. Geographical variations in hip fracture risk for women: strong effects
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
206 F. Nilson et al. hidden in standardised ratios. Osteoporos Int 2009;20: 371–7. [15] Icks A, Haastert B, Wildner M, et al. Hip fractures and area level socioeconomic conditions: a population-based study. BMC Publ Health 2009;9:114. [16] Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Publ Health 2011;11:450. [17] Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, et al. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol 2009;24:659–67. [18] SMHI. STRÅNG - a mesoscale model for solar radiation, http://strang.smhi.se (2013, accessed 30 May 2013). [19] Webb AR, Kline L and Holick MF. Influence of season and latitude on the cutaneous synthesis of vitamin D3: exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J Clin Endocrinol Metab 1988;67:373–8. [20] Engelsen O, Brustad M, Aksnes L, et al. Daily duration of vitamin D synthesis in human skin with relation to latitude, total ozone, altitude, ground cover, aerosols and cloud thickness. Photochem Photobiol 2005;81:1287–90.
[21] Marks R. Hip fracture epidemiological trends, outcomes, and risk factors, 1970–2009. Int J Gen Med 2010;3: 1–17. [22] Rubenstein LZ. Falls in older people: epidemiology, risk faktors and strategies for prevention. Age Ageing 2006;35(Suppl 2): ii37–ii41. [23] Burgaz A, Akesson A, Oster A, et al. Associations of diet, supplement use, and ultraviolet B radiation exposure with vitamin D status in Swedish women during winter. Am J Clin Nutr 2007;86:1399–1404. [24] Brustad M, Edvardsen K, Wilsgaard T, et al. Seasonality of UV-radiation and vitamin D status at 69 degrees north. Photochem Photobiol Sci 2007;6:903–8. [25] Holick MF and Chen TC. Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr 2008;87:1080S–6S. [26] SMHI. Climate data, http://www.smhi.se/klimatdata/meteorologi/sno (2009, accessed 1 May 2013). [27] Finsen V and Benum P. Changing incidence of hip fractures in rural and urban areas of central Norway. Clin Orthop Relat Res 1987;May:104–10. [28] Benetos IS, Babis GC, Zoubos AB, et al. Factors affecting the risk of hip fractures. Injury 2007;38:735–44.
Information for Authors Why do we ask you to transfer copyright to the Nordic Societies of Public Health? 1. Ownership of copyright by the Nordic Societies of Public Health ensures maximum protection against infringement of copyright protected material through breach of copyright or piracy anywhere in the world. 2. It also ensures that requests by third parties to reprint or reproduce a contribution, or part of it in any format, are handled efficiently in accordance with our general policy which encourages dissemination of knowledge inside the framework of copyright. SAGE administers the copyright and permissions for SJPH on behalf of the Nordic Societies of Public Health. For more information please go to the Author gateway: http://www.sagepub.co.uk/ journalEditors.nav
Downloaded from sjp.sagepub.com at Uniwersytet Warszawski on May 15, 2015
