MEDICAL SURVEILLANCE MONTHLY REPORT

Transcription

1 OCTOBER 2014 Volume 21 Number 10 msmr MEDICAL SURVEILLANCE MONTHLY REPORT PAGE 2 Suicides nd suicide ttempts mong ctive component members of the U.S. Armed Forces, : methods of self-hrm vry by mjor geogrphic region of ssignment Willim P. Corr, III, MD, MPH PAGE 6 Risk of type II dibetes nd hypertension ssocited with chronic insomni mong ctive component, U.S. Armed Forces, Pul E. Lewis, MD, MPH; Oseizme V. Emselu, MD, MPH; Ptrici Rohrbeck, DrPH, MPH; Zheng Hu, MS PAGE 14 Updte: cold wether injuries, ctive nd reserve components, U.S. Armed Forces, July 2009 June 2014 Ricrdford R. Connor, MPH PAGE 20 Surveillnce snpshot: influenz immuniztion mong U.S. Armed Forces helthcre workers, August 2008 April 2014 PAGE 21 Surveillnce snpshot: mnner nd cuse of deth, ctive component, U.S. Armed Forces, SUMMARY TABLES AND FIGURES PAGE 22 Deployment-relted conditions of specil surveillnce interest A publiction of the Armed Forces Helth Surveillnce Center

2 Suicides nd Suicide Attempts Among Active Component Members of the U.S. Armed Forces, : Methods of Self-Hrm Vry by Mjor Geogrphic Region of Assignment Willim P. Corr, III, MD, MPH (COL, USA) METHODS This report nlyzed dt from the Deprtment of Defense Suicide Event Report progrm bout suicide events (suicide ttempts nd suicides) mong ctive component service members during Most ttempts (85.2%) nd suicides (83.5%) occurred mong service members sttioned in the U.S. Drugs were identified s the method of self-hrm in 54.8% of ttempts but in only 3.6% of suicides. Firerms were the leding method of suicide in both the U.S. nd combt zones (61.1% nd 97.2%, respectively) but ccounted for only 5.4% of suicides in those sttioned in Europe/Asi. Hnging/sphyxition (22.9% overll) ws the second most common method in suicides. For suicides using firerms, the rtes of suicide nd the types of firerm used vried ccording to service members geogrphiclly relted ccess to firerms. Chllenges to reducing the frequency of service member suicides by firerms re discussed. suicide ws the third leding mnner of deth fter ccidents nd illness mong U.S. ctive duty service members from 1998 until 2003, when deths due to wr surpssed the number of deths from illness nd suicide. 1 3 From 2006 through 2011, deths from suicide exceeded those ssocited with illness, nd suicide ws gin the third leding mnner of deth. After the strt of the drwdown in forces in Afghnistn in 2011, wr-relted deths declined so much tht suicide becme the second leding mnner of deth ctegory fter ccidents mong ctive component service members in 2012 nd In 2008, the Depr tment of Defense Suicide Event Report (DoDSER) progrm ws lunched in support of the DoD suicide prevention inititives. 5 The DoDSER progrm is collbortive undertking of the DoD s Suicide Prevention nd Risk Reduction Committee, the Services DoDSER progrm mngers, nd the Ntionl Center for Telehelth nd Technology (T2). The progrm stndrdizes suicide surveillnce efforts for the four Services to inform their suicide prevention progrms. The stndrdized reports (DoDSERs) of suicide behviors (i.e., suicides, suicide ttempts, nd other types of self-hrm) re submitted vi secure web-bsed DoDSER ppliction. Ech DoDSER form contins pproximtely 250 dt fields (e.g., demogrphics, militry history) nd must be completed for ll suspected suicides nd suicide ttempts tht result in hospitliztion or evcution. The T2 nd the Defense Centers of Excellence for Psychologicl Helth nd Trumtic Brin Injury publish nnul reports bsed on n nlysis of the ggregted informtion from ech yer s DoDSERs. 5 This study uses DoDSER dt to nlyze the rtes of suicide events in three mjor geogrphic regions of duty ssignment nd ssesses the differences in the methods of self-hrm between those regions. The surveillnce period ws 1 Jnury 2010 through 31 December The study popultion included ll ctive component service members of the U.S. Army, Air Force, Mrine Corps, nd Nvy. Suicide events (both suicides nd non-ftl suicide ttempts) were scertined from T2 s centrlized records of the DoDSER progrm. No personlly identifible informtion or demogrphic dt were provided by T2 nd the dt were ggregted for ll 3 yers of the period. For the clcultion of rtes, denomintor dt cme from the Armed Forces Helth Surveillnce Center s Defense Medicl Surveillnce System, which utilizes personnel dt from the Defense Mnpower Dt Center, including informtion on dtes nd loctions of service members entire periods of militry service. The principl dt elements were suicides nd non-ftl suicide ttempts; ech group ws ctegorized ccording to the geogrphic loctions where they occurred nd the methods of self-hrm used. The geogrphic informtion ws used to crete three mutully exclusive groups of suicide-relted events: those tht occurred in combt zone (Irq, Afghnistn, Kuwit, nd Djibouti), those in the U.S., or those in Europe nd Asi (exclusive of the combt zones). For the purposes of this nlysis, the three suicide ttempts in the Western Hemisphere outside the U.S. were grouped with those inside the U.S. Although ll methods of self-hrm were scertined, the nlysis focused on the more commonly used methods. Tble 1 lists the 15 methods of self-hrm into which the DoDSER progrm ctegorizes suicide events. 5 The ctegory drugs included illicit drugs, inhlnts, prescription drugs, nd over-the-counter medictions. For the clcultion of rtes, numbers of suicides nd suicide ttempts were divided by the person-time ssocited with either the entire ctive component or the ctive component ssigned to the geogrphic region of interest. Pge 2 MSMR Vol. 21 No. 10 October 2014

3 TABLE 1. Deprtment of Defense Suicide Event Report (DoDSER) ctegories of methods of suicide ttempts nd suicides Ctegories of methods of suicide ttempts nd suicides Drugs Alcohol Gs, vehicle exhust Gs, utility (or other) Chemicls Hnging/sphyxition Drowning Firerm, militry issued Firerm, not militry issued Fire/stem Shrp/blunt object Jumping from high plce Lying in front of moving object Crshing motor vehicle Other Dt unvilble RESULTS The DoDSER progrm received reports of 2,553 suicide ttempts nd 812 suicides mong ctive component service members during the 3-yer surveillnce period. Bsed on these reports, the overll rtes were 53.6 suicide ttempts per 100,000 person-yers (p-yrs) nd 17.1 suicides per 100,000 p-yrs. The numbers nd rtes of these events differed mong the three geogrphic regions (Tble 2). The overwhelming mjority of reported suicide ttempts (85.2%) nd suicides (83.5%) occurred in the U.S. group. Although the rte of reported suicide ttempts in the combt zones (17.8 per 100,000 p-yrs) ws less thn one-third of the rte in the U.S. group (61.7 per 100,000 p-yrs), the rte of suicides in combt zones ws only 9.9% lower thn the rte in the U.S. group (17.3 vs per 100,000 p-yrs). The rte of suicides in Europe nd Asi (4.5 per 100,000 p-yrs) ws only bout one-fourth the rtes in the other two geogrphic regions (Tble 2). DoDSERs lcked sufficient informtion for 2.5% of ttempts nd 3.2% of suicides to ssocite them with ny of the three geogrphic regions. The most commonly reported method of self-hrm ws drugs, which were ssocited with 54.8% of suicide ttempts but only 3.6% of suicides (Tble 3). The most common TABLE 2. Number, rtes, nd percent distribution of suicide ttempts nd suicides ccording to the geogrphic loction of the event, ctive component service members, U.S. Armed Forces, Suicide ttempts Suicides No. Rte % of totl No. Rte % of totl 2, Totl Region 2, U.S Europe/Asi Combt zones Other/missing Rte per 100,000 person-yers methods ssocited with suicide ttempts were drugs, use of shrp or blunt object, nd hnging/sphyxition (Tble 3). Suicides were most often ssocited with the use of non-militry firerm, hnging/sphyxition, nd the use of militry-issued or duty firerm. The seven most frequently reported methods were ssocited with 90.6% of ll suicide ttempts nd 89.9% of ll suicides. The methods used for suicide ttempts nd suicides vried cross geogrphic regions. Tble 4 presents the rtes (per 100,000 p-yrs) for the seven most common methods of self-hrm for ech of the three geogrphic regions. Among reported suicide ttempts, drugs were the most common method employed in ll three regions. However, in the U.S. nd Europe/Asi groups, shrp/blunt objects nd hnging/sphyxition were the next most common methods, while in combt zones, militry-issued firerms were the second most used method of ttempted suicides. The methods used during suicides nd ttempted suicides vried mrkedly (Tble 4). Notbly, drugs were used in more thn hlf of suicide ttempts but were the method of only 3.5% of suicides. Firerms were the leding method of suicide in both the U.S. nd combt zones (61.1% nd 97.2%, respectively) but ccounted for only 5.4% of suicides in Europe/Asi. Figure 1 depicts the rtes of suicides by firerms, hnging/ sphyxition, nd ll other methods in the three regions. Hnging/sphyxition ws the most frequent method of suicide in the Europe/Asi group (67.6%) nd the second most frequent method in the U.S. (22.1%) (Tble 4). Of ll firerm-relted suicides in the U.S. group (n=414), reltively few (n=26, 6.3%) were ssocited with militry-issued wepons (Tble 4). In contrst, 94.2% (n=65) of ll firerm-relted suicides in combt zones were ttributed to militry-issued wepons. The two firerm-ssocited suicides in the Europe/Asi group were ttributble to militry-issued wepons. Although lcohol ws reported s the method of self-hrm for 3.3% of ll suicide ttempts, there were no suicides ttributed to lcohol s the method. Moreover, there were no reports from combt zones of suicide ttempts where lcohol ws reported s the method. The only methods of suicide reported from combt zones were firerms (n=69), hnging (n=1), nd other (n=1). EDITORIAL COMMENT This nlysis of DoDSER dt from found n overll rte of suicides of 17.1 per 100,000 p-yrs mong ll ctive component service members. The suicide rte ws much lower mong those serving in Europe nd Asi thn in the U.S. nd combt zones. The striking difference in rtes of suicides overll in Europe/Asi compred to other regions reflects, to lrge extent, mrked differences in firerm-relted suicides in Europe/Asi (n=2, 0.2 per 100,000 p-yrs) compred to the U.S. (n=414, 11.7 per 100,000 p-yrs) nd combt zones (n=69, 16.9 per 100,000 p-yrs). In generl, U.S. militry members hve more limited ccess to firerms when serving in Europe/Asi thn in other regions. Nerly ll of the countries in Europe nd Asi where U.S. service members re sttioned hve very restrictive wepons lws; 6 in turn, reltively October 2014 Vol. 21 No. 10 MSMR Pge 3

4 TABLE 3. Seven most common methods of suicide ttempts nd suicides, ctive component service members, U.S. Armed Forces, Suicide ttempts Suicides No. % of totl No. % of totl Firerm/gun, other thn militry issue Hnging/sphyxition Firerm/gun, militry issued or duty wepon , Drugs Other Shrp or blunt object Alcohol All other methods (fter top seven) , Totl TABLE 4. Numbers nd rtes of suicide ttempts nd suicides by geogrphic region nd method, ctive component, U.S. Armed Forces, All ctive component U.S. Europe/Asi Combt zones Suicide ttempts No. Rte b No. Rte b No. Rte b No. Rte b Firerm, not militry issue Hnging/sphyxition Firerm, militry issue Drugs 1, , Other Shrp/blunt object Alcohol All other methods Totl of ll ttempts 2, , All ctive component U.S. Europe/Asi Combt zones Suicides No. Rte b No. Rte b No. Rte b No. Rte b Firerm, not militry issued Hnging/sphyxition Firerm, militry issue Drugs Other Shrp/blunt object Alcohol All other methods Totl of ll suicides Totl for ll ctive component does not equl sum of numbers for the three geogrphic regions becuse some Deprtment of Defense Suicide Event Report dt did not indicte the geogrphic loction of the suicide event. b Rte per 100,000 person-yers few service members hve opportunities to cquire privtely owned firerms leglly while ssigned in Europe/Asi. In ddition, militry regultions generlly prohibit keeping privtely owned firerm (POF) in one s residence during overses ssignments; ny POF brought to n overses re must be stored in secure militry rms room or rmory. 7 Given the restrictions on possession of firerms by U.S. militry members while serving in Europe/Asi, it is not surprising tht the only firerm-relted suicide events in Europe nd Asi during the 3-yer period of interest were ssocited with militry-issued wepons. Of note in this regrd, most service members in Europe nd Asi do not routinely hve ccess to militry-issued firerms. At the other extreme of the ccess to firerms spectrum, nerly ll service members in combt zones re issued firerms nd mmunition which must remin under the individul s control t ll times. 8 The suicide rte from firerms in combt zones ws higher thn in the other geogrphic regions, nd only two of the 71 suicides in combt zones were not ttributed to firerms. In the U.S. group, most suicides (61.1%) were ssocited with firerms, nd 93.7% of ll firerm-relted deths were linked to non-militry-issued wepons. As is the cse in Europe nd Asi, most service members duties in the U.S. do not require them to crry militry-issued firerms on regulr or routine bsis. Although lws nd regultions governing gun ownership vry from stte to stte, most citizens in the U.S. my leglly own POFs nd mny service members do. 9 Almost ll deths of ctive component service members re subject to close scrutiny by the Armed Forces Medicl Exminer; documenttion of the mnner nd underlying cuse of deth is thorough. When deth is determined to be the result of suicide, tht ctegoriztion is required to be recorded in DoDSER. Accordingly, the identifiction nd chrcteriztion of service member suicides re firly complete nd the corresponding dt from the DoDSER records re useful in documenting trends in suicides nd ssocited fctors. A limittion of this nlysis is tht suicide ttempts re not investigted or reported s completely s re deths. In this report, suicide ttempts were scertined vi DoD- SERs only. The DoDSER defines suicide ttempt s non-ftl, self-directed, potentilly injurious behvior with ny intent to die Pge 4 MSMR Vol. 21 No. 10 October 2014

5 FIGURE 1. Incidence rte of suicide ccording to principl methods nd geogrphic loction of event, ctive component, U.S. Armed Forces, Incidence rte per 100,000 person-yers Firerms Hnging/sphyxition All other Other s result of the behvior. A suicide ttempt my or my not result in injury. For exmple, if someone seizes gun from service member who ws holding it to his hed, the event should be clssified s suicide ttempt even though there were no injuries. In ddition, DoDSERs re required only when the event results in hospitliztion or n evcution from theter. Other types of ttempts my result in physicl injuries tht were reltively minor, were not detected by others, or were ttributed to ccidents. Unless such events result in the submission of DoD- SER, they were not included in the dt provided for this report. Becuse some suicide ttempts likely were not reported, the suicide ttempts enumerted in this nlysis must be regrded s underestimtes of the incidence of such events. 10 From the perspectives of both public helth nd clinicl cre, the prevention of suicide is formidble chllenge. The DoD s efforts to reduce the incidence of suicides nd suicide ttempts hve included inititives to educte militry leders, service members, nd fmily members bout the risk fctors for self-hrm; to increse the vilbility of, nd ccess to, mentl helth resources to help individuls del with suicidl idetion nd other psychologicl distress; nd to destigmtize seeking help for mentl helth problems. In recent yers, the DoD hs gretly incresed the level of mentl helth resources to ssist with chieving such objectives, but the toll of suicide remins high. It remins to be seen whether or not the recently reduced level of combt opertions will be followed by reduction in suicide incidence. The mjority of suicides in both the U.S. civilin popultion nd in the U.S. Armed Forces hve been due to self-hrm with firerms. A RAND study commissioned by the DoD recommended considertion U.S. Europe/Asi Combt zones of restrictions on ccess to lethl mens of self-hrm. 10 The results of the nlysis in this report document tht, mong the seven most common methods of lethl self-hrm, the only mens of suicide tht re not commonplce in every service member s dily life re drugs nd firerms. Restricting ccess to these risk fctors presents complex chllenges. Medictions to tret illnesses re vluble mesures to enhnce the qulity nd durtion of life. Similrly, in the Armed Forces, firerms re importnt, often essentil, nd commonly employed tools for security, protection of life, nd chievement of the mission. DoDSER nnul nlyses hve noted tht firerms were present in the homes or immedite environments of bout hlf of suicide decedents. 11,12 Meeting the chllenge involved in restricting ccess to these potentil instruments of self-hrm depends on methods of identifying those distressed individuls whose ccess should be constrined nd then ctully controlling ccess. Impediments to ccomplishing such tsk include the relities tht mny individuls with suicidl thoughts choose not to mke known to others their distress, tht the Services cn control ccess to only militryissued firerms, nd tht ccess to non-militry-issued firerms nd drugs is not redily menble to limittion by officil militry controls, prticulrly in the U.S. It is importnt to note tht this nlysis suggests n ssocition between ccess to firerms nd suicide rtes, prticulrly those suicides due to firerms, but it does not demonstrte cler cuse-nd-effect reltionship. In prticulr, the results of this nlysis should not be interpreted s proof tht restricting ccess to firerms would reduce suicide rtes. It is possible tht other fctors plyed n importnt role in the differences in suicide rtes mong the geogrphic regions studied. Author ffilition: Armed Forces Helth Surveillnce Center, Silver Spring, MD. Acknowledgement: Specil thnks to the DoD Suicide Event Report tem t the Ntionl Center for Telehelth nd Technology (T2) for their ssistnce, feedbck, nd insights. REFERENCES 1. Army Medicl Surveillnce Activity. Mortlity trends mong ctive duty militry service members, MSMR. 1999; 5(2): Armed Forces Helth Surveillnce Center. Deths while on ctive duty in the U.S. Armed Forces, MSMR. 2009; 16(5): Arme d Forces Helth Surveillnce Center. Deths while on ctive duty in the U.S. Armed Forces, MSMR. 2012; 19(5): Armed Forces Helth Surveillnce Center. Surveillnce snpshot: mnner nd cuse of deth, ctive component, U.S. Armed Forces, MSMR. 2014; 21(10): Smolenski DJ, Reger MA, Alexnder CL, et l. Deprtment of Defense Suicide Event Report, Clendr Yer 2012 Annul Report. Ntionl Center for Telehelth nd Technology, Defense Centers of Excellence for Psychologicl Helth nd Trumtic Brin Injury. Avilble t: sites/defult/files/dodser_r2012_ _0.pdf. Accessed on 10 October Wikipedi. Overview of gun lws by ntion. en.wikipedi.org/wiki/overview_of_gun_lws_by_ ntion. Accessed on 20 October U.S. Forces Kore Reg Registrtion nd Control of Privtely Owned Firerms. Avilble t: Policy/Publiction_Records_Reg_USFK.htm. Accessed on 30 My Hedqurters, Interntionl Security Assistnce Force, U.S. Ntionl Support Element, Kbul, Afghnistn. Memorndum for Record, Subject: U.S. Ntionl Stndrds, HQ ISAF. 18 July Avilble t: Accessed on 29 My Hepburn L, Miller M, Azrel D, Hemenwy D. The U.S. gun stock: results from the 2004 ntionl firerms survey. Inj Prev. 2007; 13: Rmchnd R, Acost J, Burns R, Jycox L, Pernin C. The Wr Within: Preventing Suicide in the U.S. Militry. RAND Kinn JT, Luxton DD, Reger MA, et l. Deprtment of Defense Suicide Event Report, Clendr Yer 2010 Annul Report. Ntionl Center for Telehelth nd Technology, Defense Centers of Excellence for Psychologicl Helth nd Trumtic Brin Injury. Avilble t: files/dodser/dodser_2010_annul_report.pdf. Accessed on 10 October Luxton DD, Osenbch JE, Reger MA, et l. Deprtment of Defense Suicide Event Report, Clendr Yer 2011 Annul Report. Ntionl Center for Telehelth nd Technology, Defense Centers of Excellence for Psychologicl Helth nd Trumtic Brin Injury. Avilble t: sites/defult/fi les/dodser/dodser_2011_annul_ Report.pdf. Accessed on 10 October October 2014 Vol. 21 No. 10 MSMR Pge 5

6 Risk of Type II Dibetes nd Hypertension Associted with Chronic Insomni Among Active Component, U.S. Armed Forces, Pul E. Lewis, MD, MPH (LtCol, USAF); Oseizme V. Emselu, MD, MPH; Ptrici Rohrbeck, DrPH, MPH (Mj, USAF); Zheng Hu, MS Chronic insomni is common clinicl complint nd its incidence in both U.S. militry nd civilin popultions hs incresed. Severl studies hve evluted the ssocition between chronic insomni nd the development of other chronic diseses. This study estimtes the incidence of chronic insomni. In ddition, this report exmines the ssocition between both hypertension nd type II dibetes nd chronic insomni in ctive component militry members. The Defense Medicl Surveillnce System ws used to identify cohort of individuls with chronic insomni between 1998 nd 2013 nd to mtch them by ge nd gender with cohort without insomni. During , there were 205,740 incident cses of chronic insomni mong ctive component service members with n overll rte of 90.3 per 10,000 person-yers. Individuls in the chronic insomni cohort were t higher risk for type II dibetes (djusted hzrd rtio [HR], 2.17 [95% CI, ]) nd hypertension (djusted HR, 2.00 [95% CI, ]). Sleep hygiene eduction long with evlution nd tretment of persistent symptoms re of public helth importnce in ctive duty service members. insomni is chrcterized by impired dytime function s result of difficulty inititing or mintining sleep or wking up erly in the morning without the bility to return to sleep despite dequte opportunity nd circumstnces for sleep. Although mny individuls experience intermittent or short-term insomni, chronic insomni is dignosed when symptoms occur t lest three times per week for 3 months or more. 1 Insomni is clinicl dignosis, with sleep history s the only dignostic evlution required. Tretment consists of sleep hygiene eduction progressing to mediction mngement when necessry. Prevlence estimtes of chronic insomni re s high s 19% in the U.S. popultion, nd survey of primry cre ptients found tht 69% hd occsionl or chronic insomni. 2 Insomni increses with ge, with only 12% of elderly individuls reporting norml sleep. 3 Other demogrphic fctors ssocited with the development of insomni include femle gender, employment sttus, mritl sttus, nd socioeconomic sttus. 4 Insomni is common complint in ctive duty service members. Of those returning from deployment in Irq nd Afghnistn, 41% reported problems sleeping. 5 A 10-yer study of ctive component service members found n overll incidence rte of 48 cses of chronic insomni per 10,000 person-yers (p-yrs). Insomni ws dignosed t rte 57% higher mong femles (70.0 cses per 10,000 p-yrs) thn mles (44.7 cses per 10,000 p-yrs) nd Army personnel hd the highest incidence rte of the Services (79.2 cses per 10,000 p-yrs). 6 Insomni ffects work performnce, socil functioning, nd qulity of life. It is lso significnt fctor in work-relted nd motor vehicle ccidents. 7 Individuls with insomni hve n incresed risk of developing depression, nxiety, nd drug buse. 8 Those with pulmonry disese, hert filure, or chronic pin re lso known to commonly suffer from insomni. 9,10 Recent studies hve investigted the ide tht insomni my be predisposing fctor for common chronic diseses such s dibetes, hypertension, obesity, nd coronry rtery disese Militry personnel re t high risk for insomni due to stressors such s frequent moves, deployments, nd rotting shifts or night shift work. 15,16 In militry setting, the consequences of work-relted ccidents cn be mgnified, given the nture nd demnds of militry opertions (e.g., ftigue is cited s the primry cuse of militry vition mishps). 6 Type II dibetes nd hypertension re chronic conditions tht led to lifelong disbility t potentilly significnt cost to the Deprtment of Defense. An updted report on insomni incidence nd n evlution of its potentil impct on these chronic conditions in n ctive duty popultion hve significnt militry relevnce. METHODS The surveillnce period ws 1 Jnury 1998 through 31 December The surveillnce popultion included ll individuls who served t ny time in the ctive component of the Army, Nvy, Air Force, Mrine Corps, or Cost Gurd. All dt used to determine incident dignoses of chronic insomni, hypertension, or type II dibetes were derived from records routinely mintined in the Defense Medicl Surveillnce System (DMSS). These records document both mbultory encounters nd hospitliztions of ctive component members of the Pge 6 MSMR Vol. 21 No. 10 October 2014

7 TABLE 1. ICD-9 codes for type II dibetes, hypertension, nd insomni Type II dibetes ICD-9 codes Dibetes mellitus without mention of compliction or Dibetes with ketocidosis or Dibetes with hyperosmolrity or Dibetes with other com or Dibetes with renl mnifesttions or Dibetes with ophthlmic mnifesttions or Dibetes with neurologicl mnifesttions or Dibetes with peripherl circultory disorders or Dibetes with other specifi ed mnifesttions or Dibetes with unspecifi ed complictions or Hypertension ICD-9 codes Mlignnt essentil hypertension 401 Benign essentil hypertension Unspecifi ed essentil hypertension Insomni ICD-9 codes Trnsient insomni Persistent insomni Orgnic insomnis Other orgnic insomnis Insomni, unspecified clculted by using the number of cses nd the number of person-yers of followup for ech cohort. Cox proportionl hzrds regression models were developed to ssess the influence of chronic insomni on the risk of developing hypertension or type II dibetes during the follow-up period. Hzrd rtios (HRs) were djusted for ge nd gender (to obtin strtified estimtes) s well s rce/ethnicity nd obesity. Individuls were clssified s obese if they hd received ny of the following ICD-9 dignoses in ny dignostic position in ny medicl encounter during the study period: , , V85.3x, or V85.4x (i.e., obesity, unspecified, morbid obesity, body mss index between 30 39, dult, body mss index 40 nd over, dult ). All nlyses were performed using SAS System for Windows, version 9.2. U.S. Armed Forces in fixed militry nd civilin (if reimbursed through the Militry Helth System) tretment fcilities. The study design ws retrospective cohort study with dynmic cohort (i.e., incident cses of chronic insomni t ny time during the surveillnce period were included in the nlysis). An incident cse of chronic insomni ws defined by records of two or more mbultory visits within 90 dys of ech other or hospitliztion with dignosis of insomni in ny dignostic position. 6 The ICD-9 codes used to define cse of chronic insomni re listed in Tble 1. Ech individul could be n incident cse once during the surveillnce period. Any individuls with chronic insomni dignosis prior to 1 Jnury 1998 were excluded. Incidence rtes of hypertension nd type II dibetes were compred between two groups of ctive component service members. The chronic insomni cohort comprised service members who met the surveillnce cse definition for chronic insomni s described in the preceding prgrph. Service members with dignosis of sleep pne t ny time during their militry service were excluded from the cohort. The control cohort consisted of service members who were not documented s hving chronic insomni. One control ws selected for ech cse from rndom smple of 25% of eligible controls. Ech control ws mtched on gender nd ge (within 1 yer of the mtched cse). A control must hve been in service t the time of the incident dte of insomni dignosis of their mtched cse. Service members with dignosis of sleep pne t ny time during their militry service were excluded from the cohort. The insomni cohort nd the control cohort were then followed from the dte of incident insomni dignosis until onset of hypertension or type II dibetes, seprtion from the militry, deth, or until 31 December The beginning of control s follow-up period begn on the sme dte s the incident insomni dignosis of the mtched cse. Incident cses of hypertension or type II dibetes were defined s two or more mbultory visits within 90 dys of ech other or hospitliztion with ny of the ICD-9 codes listed (Tble 1). Any cses of hypertension or type II dibetes tht occurred within 1 yer of dignosis of insomni were excluded from both groups. The two cohorts were compred for differences in rce nd obesity sttus using the chi-squre test. An incidence rte ws RESULTS During , there were 205,740 incident cses of chronic insomni mong ctive component service members with n overll crude incidence rte of 90.3 cses per 10,000 p-yrs. The highest nnul overll rte ws in 2012 with cses per 10,000 p-yrs (Figure, Tble 2). Since 2005, the Army hs consistently hd the highest rte of chronic insomni mong the Services, peking t cses per 10,000 p-yrs in This rte ws more thn 2.5 times the rte of ny of the other Services tht sme yer. A sevenfold increse in the rte mong men nd five-fold increse in the rte mong women occurred during the sme period ( ). The rte increse ws seen cross ll ge ctegories nd ll militry occuptions. The lrgest rte increse ws mong those in the occuptionl ctegories of infntry/ rtillery nd rmor/motor trnsport, from 33 per 10,000 p-yrs in 2005 to more thn 300 per 10,000 p-yrs in 2012 (Tble 2). Anlyses evluting the development of type II dibetes Of the 205,740 totl cses of incident insomni, 105,246 were excluded due to co-morbid dignosis of sleep pne or October 2014 Vol. 21 No. 10 MSMR Pge 7

8 TABLE 2. Numbers nd rtes of incident dignoses of insomni (one cse per individul in surveillnce period), ctive component, U.S. Armed Forces, Jnury 1998 December 2013 Totl No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte All 205, , , , , , Service Army 130, , , Nvy 19, Air Force 38, , , Mrine Corps 15, Cost Gurd 2, Sex Mle 164, , , , , Femle 41, , Rce/ethnicity White, non-hispnic 142, , , , , Blck, non-hispnic 39, Other 23, Age <20 7, , , , , , , , , Rnk E1 E4 95, , , , E5 E9 90, , , O1 O4 13, O5 O10 4, W1 W5 3, Mritl sttus Mrried 124, , , , Single 65, , , Other 16, Eduction High school or less 149, , , , , Some college 24, College 17, Advnced degree 10, Unknown 4, Militry occuption Infntry/rtillery 34, Armor/motor trnsport 11, Pilot/ircrew 2, Repir/engineering 48, , Comm/intel 49, , Helth cre 25, Other 33, Rte per 10,000 person-yers Pge 8 MSMR Vol. 21 No. 10 October 2014

9 TABLE 2 (cont.) Numbers nd rtes of incident dignoses of insomni (one cse per individul in surveillnce period), ctive component, U.S. Armed Forces, Jnury 1998 December No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte No. Rte* All 8, , , , , , , , Service Army 4, , , , , , , , Nvy , , , , , , , Air Force 2, , , , , , , , Mrine Corps , , , , , , Cost Gurd Sex Mle 6, , , , , , , , Femle 2, , , , , , , , Rce/ethnicity White, non-hispnic 6, , , , , , , , Blck, non-hispnic 1, , , , , , , , Other 1, , , , , , , , Age < , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , Rnk E1 E4 4, , , , , , , , E5 E9 3, , , , , , , , O1 O , , , , , , O5 O W1 W Mritl sttus Mrried 4, , , , , , , , Single 3, , , , , , , , Other , , , , , , Eduction High school or less 6, , , , , , , , Some college , , , , , , , College , , , , , , , Advnced degree , , , , , Unknown Militry occuption Infntry/rtillery 1, , , , , , , , Armor/motor trnsport , , , , , , Pilot/ircrew Repir/engineering 2, , , , , , , , Comm/intel 2, , , , , , , , Helth cre 1, , , , , , , , Other 1, , , , , , , , Rte per 10,000 person-yers October 2014 Vol. 21 No. 10 MSMR Pge 9

10 pre-existing type II dibetes. The remining 100,494 mde up the study cohort nd the control cohort comprised n equl number of mtched individuls. Chrcteristics of the study nd comprison cohorts re shown in Tble 3. The verge follow-up time ws 3.09 yers in the study cohort nd 3.42 yers in the control cohort. Individuls in the study cohort were more likely to be blck, non-hispnic (18.9%) or obese (12.4%) thn in the control cohort (16.5% nd 5.9%, respectively) (Tble 3). Overll, the crude incidence rte of type II dibetes ws 15.4 cses per 10,000 p-yrs in the study cohort nd 6.2 cses per 10,000 p-yrs in the control cohort (Tble 4). Those in the insomni cohort were t higher risk of developing type II dibetes (djusted HR, 2.17 [95% CI, ]). The elevted risk of type II dibetes in the insomni cohort ws significnt mong both those younger thn 30 yers of ge (djusted HR, 2.53 [95% CI, ]) nd those ged 30 yers or older (djusted HR, 2.12 [95% C, ]. The risk of type II dibetes mong those with insomni compred to those without insomni ws elevted in ll rce ctegories nd in those with nd without dignoses of obesity. (Tble 4). Anlyses evluting the development of hypertension The study cohort for this nlysis included 94,162 service members; 111,578 individuls were excluded from the cohort of those with chronic insomni due to co-morbid dignosis of sleep pne or preexisting hypertension. The verge followup times were 3.06 yers nd 3.42 yers for the study nd control groups, respectively. Higher proportions of the insomni cohort were blck, non-hispnic or obese. (Tble 5) The crude incidence rte of hypertension ws 95.6 cses per 10,000 p-yrs in the insomni cohort nd 46.2 per 10,000 p-yrs in the control cohort, with n overll djusted HR of 2.00 (95% CI, ) (Tble 6). The ssocition of chronic insomni with hypertension ws seen both in those younger thn 30 yers of ge nd in those ged 30 yers or older, with djusted HRs of 2.32 (95% CI, ) nd 1.94 (95% CI, ), respectively. Strtifiction by gender demonstrted tht FIGURE. Annul incidence rtes of dignoses of chronic insomni, ctive component, U.S. Armed Forces, Jnury 1998 December 2013 Rte (cses per 10,000 person yers) TABLE 3. Type II dibetes study: demogrphics nd comorbidities compring ptients with nd without chronic insomni Insomni Yes No No. % No. % P vlue Totl 100, , Age <30 58, , , , Sex Mle 79, , Femle 21, , Rce White, non-hispnic 69, , < Blck, non-hispnic 18, , Other 12, , Obesity No 88, , < Yes 12, , chronic insomni hd stronger ssocition with hypertension in men (djusted HR, 2.17 [95% CI, ]) thn in women (djusted HR, 1.59 [95% CI, ]). The HR ws lso greter for whites (djusted HR, 2.26 [95% CI, ]) thn for blck, non-hispnics (djusted HR, 1.72 [95% CI ]). Finlly, both obese subjects nd non-obese subjects hd significntly incresed risk of hypertension relted to insomni with djusted HRs of 2.09 (95% CI, ) nd 1.86 (95% CI, ), respectively. Mentl helth dignoses were evluted s possible confounders of the ssocition between insomni nd hypertension. Previously published cse definitions of four mentl helth disorders (post-trumtic stress disorder [PTSD], depression, nxiety, or lcohol buse or dependence) were used to clssify individuls into ctegories indicting the presence or bsence of these conditions. 17 These vribles were included in multivrite nlyses to evlute whether they were significnt independent predictors of the development of hypertension or if they modified the reltionship between insomni nd development of Pge 10 MSMR Vol. 21 No. 10 October 2014

11 TABLE 4. Type II dibetes risk ssocited with insomni Vribles Insomni Hzrd rtios Yes No Chronic insomni vs. those without chronic insomni Cse P-Y IR Cse P-Y IR Crude HR CI Adj HR b CI Overll , , ( ) 2.17 ( ) Age < , , ( ) 2.53 ( ) , , ( ) 2.12 ( ) Sex Mle , , ( ) 2.20 ( ) Femle 95 68, , ( ) 2.34 ( ) Rce White, non-hispnic , , ( ) 2.23 ( ) Blck, non-hispnic , , ( ) 2.13 ( ) Other 97 39, , ( ) 2.32 ( ) Obesity No , , ( ) 2.35 ( ) Yes , , ( ) 1.85 ( ) Rte per 10,000 person-yers b Adjusted for ge, sex, rce, obesity P-Y=person-yers; IR=incidence rte; CI=confidence intervl hypertension. Inclusion of vribles indicting the presence/bsence of depression, nxiety, PTSD, or lcohol dependence/ buse in the model resulted in n overll djusted HR of 1.91 (95% CI, ), vlue essentilly unchnged from the crude HR. EDITORIAL COMMENT This report exmines the incidence of chronic insomni in ctive component militry members nd its ssocition with subsequent dignoses of hypertension nd type II dibetes. Chronic insomni incidence rose shrply from 2004 through 2012, with pek crude incidence rte of 252 cses per 10,000 p-yrs. After yers of stedy increse, the rte leveled out in 2013 (IR: 248 cses per 10,000 p-yrs). Severl possible mechnisms of ction hve been proposed to explin why insomni might increse the risk of developing hypertension or type II dibetes. Ptients with insomni hve dysregultion of the hypothlmic pituitry drenl xis resulting in hypercortisolemi High levels of serum cortisol increse the risk of mny chronic diseses including TABLE 5. Hypertension study: demogrphics nd comorbidities compring ptients with nd without chronic insomni Insomni Yes No No. % No. % P vlue Totl 94, , Age <30 57, , , , Sex Mle 74, , Femle 20, , Rce White, non-hispnic 66, , < Blck, non-hispnic 16, , Other 11, , Obesity No 83, , <0.001 Yes 10, , hypertension nd metbolic syndrome. By extension, metbolic syndrome is known precursor to type II dibetes. In ddition, sleep loss my impct the neuroendocrine control of ppetite, cusing individuls to overet nd resulting in n incresed risk of metbolic syndrome nd type II dibetes. 21 Chronic insomni ws ssocited with two-fold incresed risk of hypertension nd type II dibetes (p<0.0001). Previous studies on this topic hve provided conflicting results, with some demonstrting strong ssocition 11,22,23 nd others finding miniml to no ssocition. 13,24 26 There re severl resons why the findings of this study my not be directly comprble to studies in civilin popultions. Militry members re generlly younger nd October 2014 Vol. 21 No. 10 MSMR Pge 11

12 TABLE 6. Hypertension risk ssocited with insomni Vribles Insomni Hzrd rtios Yes No Chronic insomni vs. those without chronic insomni Cse P-Y IR Cse P-Y IR Crude HR CI Adj HR b CI Overll 2, , , , ( ) 2.00 ( ) Age < , , ( ) 2.32 ( ) 30 1, , , , ( ) 1.94 ( ) Sex Mle 2, , , , ( ) 2.17 ( ) Femle , , ( ) 1.59 ( ) Rce White, non-hispnic 1, , , ( ) 2.26 ( ) Blck, non-hispnic , , ( ) 1.72 ( ) Other , , ( ) 2.00 ( ) Obesity No 2, , , , ( ) 2.09 ( ) Yes , , ( ) 1.86 ( ) Rte per 10,000 person-yers b Adjusted for ge, sex, rce, obesity P-Y=person-yers hve fewer comorbidities thn their civilin counterprts. Although this inquiry employed longitudinl study design tht llowed for follow-up of individuls over time, mny previous studies evluting these ssocitions employed cross-sectionl designs which do not llow for the evlution of temporl reltionship between n exposure (i.e., insomni) nd n outcome (i.e., hypertension, dibetes). In ddition, insomni symptoms (rther thn physicin-ssigned dignoses) were used to define exposure in mny of these studies; such vrition in pproch my introduce differences in severity of symptoms. These methodologic differences might prtilly explin the incresed risks of hypertension nd dibetes ssocited with chronic insomni seen in this study but not seen in some other studies. There re severl limittions to this report tht should be considered when interpreting the results. First, despite djusting for confounding fctors there is the potentil for bis resulting from unmesured or unknown confounding fctors. Detiled ptient dt re not vilble in DMSS for smoking sttus, height, weight, diet, physicl ctivity, cholesterol level, lcohol use, blood pressure, supplement use, or fmily history of dibetes. Second, becuse the endpoints of nlysis were ICD-9 dignostic codes, there is the potentil for miscoding nd misclssifiction, which lso my result in bis. Ptient surveys or evidence from sleep studies were not obtined to verify the dignostic coding by physicins. Nevertheless, the strength of the study is its lrge popultion-bsed cohort with long follow-up period demonstrting the ssocition between chronic insomni nd the development of type II dibetes nd hypertension. In summry, the findings of this report suggest tht the incidence of dignoses of chronic insomni in ctive component service members hs been on the rise from 2004 through 2012, with trend towrd leveling off in A dignosis of chronic insomni ws strong predictor for subsequent incidence of type II dibetes nd hypertension, nd further study is wrrnted to determine the long-term impct of chronic insomni on service members helth. These findings highlight the importnce of sleep hygiene eduction for ctive duty members nd their supervisors, with referrl for tretment if symptoms do not improve fter 2 weeks. Disclimer: The views expressed re those of the uthors nd do not necessrily reflect the officil views of the Uniformed Services University of the Helth Sciences, U.S. Air Force, or Deprtment of Defense. Author ffilitions: Uniformed Services University of the Helth Sciences, Bethesd, MD (Dr. Lewis); Armed Forces Helth Surveillnce Center, Silver Spring, MD (Dr. Emselu, Dr. Rohrbeck, nd Ms. Hu). REFERENCES 1. Roth T. Insomni: definition, prevlence, etiology, nd consequences. J Clin Sleep Med. 2007; 15(5 Suppl): S7 S Shocht T, Umphress J, Isrel AG, Ancoli-Isrel S. Insomni in primry cre ptients. Sleep. 1999; 22(Suppl 2): S359 S Foley DJ, Monjn AA, Brown SL, Simonsick EM, Wllce RB, Blzer DG. Sleep complints mong elderly persons: n epidemiologic study of three communities. Sleep. 1995; 18(6): Ohyon MM. Epidemiology of insomni: wht we Pge 12 MSMR Vol. 21 No. 10 October 2014